GT V01 1960 187-190

The fuel situation since the war has changed from one of coal shortage to one of coal surplus;  the striving for efficiency, however, has not abated and the shift of emphasis from saving fuel for  national need to  one  of  reducing production costs  is discussed. Heat distribution  in  industrial furnaces and processes is compared and  the scope for waste-heat  recovery is considered. The importance of defining waste heat  correctly is emphasized, the term 'surplus' heat is often preferable. The importance of considering the furnace  primarily is stressed and every  effort should  be made to improve the inherent  design  before  considering waste-heat  recovery. Reference is made to the general problem of waste-heat  recovery in industry and the progress made and the importance of studying  the lessons learned in various fields. The advantages of waste-heat  recovery are endorsed but a plea is made for  considering the factory heat  demand as a  whole. Waste-heat boiler recovery  schemes  can materially reduce capital and running charges on other base load plants, but  at  the same time in certain conditions recovery of  waste heat can lead to difficulties in operation and cases are cited to illustrate this. The need for  exercising great  care at  the planning stage in waste-heat schemes is commented  on as the efficient operation of the primary furnace can be affected. Examples are given from various industries. Need for instrumentation and automatic control is discussed. 

H. B. Weston

GT V01 1960 121-129

Standard Telephones and Cables Ltd., Paignton Dimensional variation is at the root of many problems arising between producer and customer in the glass industry. It is usually possible for a designer to prepare a satisfactory specification by observing certain generally accepted quality standards, but in difficult situations (e.g. where special limits of accuracy are required) it is not always easy to decide what allowance should be made for the various sources of variation. A graphical representation of the random and systematic sources of variation enables the designer to see his problem more clearly, and by using statistical principles he can forecast the probability of success of his design. The problem of a tubular glass bulb is used for illustration and the effects of variation of mean diameter and mean wall thickness and of siding and ovality are dis- cussed. Limits of accuracy are calculated for a specific example

S. J. Morrison & M. Symons

GT V01 1960 144-154

The data presented here refer to 60 g melts of glass of composition in weight %, 73.5 SiO₂, 10.0 CaO, 16.5 Na₂O, founded in platinum in an electric furnace at 1400 and 1450" C. There was no significant convection of the melt. Melts were made at 1400°C, containing all possible combinations of 0, 0.10 and 0.50% As2O3 with 0, 1.00 and 5.00% Nus0 supplied as sodium peroxide. Four of these combinations were investigated in detail at 1400" C and two at 1450" C. By counting the seed remaining in 30 g slabs cast from melts founded for different times it was possible to construct graphs showing the decrease of total number (N) of seed per cm3 with time. This relation was approximately, but not exactly, exponential. The refining agents affected the number of seed present at the beginning of refining as well as the rate of refining; because of this it was concluded that melting and refining could not be considered as two independent processes. Measurements of seed size distributions allowed the decrease of the number of seed per cm3 (n), of given diameter (x), with time to be plotted. Most of the data were well fitted by an exponential relation between n and t. The refining agents altered the shapes of the seed size distributions and the rates at which seed disappeared. Small seed disappeared more slowly in melts containing 0.50% As203 and 5.0% Nu20 as Nu202 than in melts containing 0.50% Asz03 and 1.0% Nu20 as Na202. This suggested that it was possible to supply too much oxygen for any given addition of Asz03. Raising the temperature had effects similar to some combinations of refining agents, affecting seed size distributions and rates of disappearance of seed.

Michael Cable

GT V01 1960 115-120

An historical summary is presented of the development of photosensitive glasses and of materials associated with the Fotoceram and Pyroceram trade-marks. Various fundamental and technical problems encountered in their production are discussed. Properties of two currently standard varieties of glass-ceramic are compared with those of heat-resisting glasses and alumina ceramic, and some of the applications of glass-ceramics are mentioned.

H. R. Lillie

GT V01 1960 251-259

Although the basic  principles of Siemens original regenerative furnace have changed very little, i.e. heating the charge from above.  the  use of preheat  and firing alternatively from each end, outputs and fuel  economy have improved  considerably. Much of this improvement is due to increases in firing rate, but the use of superior fuels   and improvements in ,furnace  and  burner  design have also played  their part. The paper compares various fuels and  also examines the effect of preheat  level and the  use of combustion oxygen. Factors  affecting heat transfer are discussed  including bath coverage and flame emissivity and temperature. The paper concludes with a section on the effect of furnace  design on the flow of  the gases  within the chamber  and  their  consequent influence on overall heat  transfer. 

C. Holden

GT V01 1960 211-216

Seeking ways of obtaining further fuel  savings on already well-insulated tank furnaces, two new developments have recently been carried out in the insulating technique. I. Aluminium sheets have been suspended in front  of recuperator walls. The wall was not  insulated in the normal  way because of a provision of numerous plugged openings required for  periodic cleaning of spool checkers. The aluminium  sheets reduced losses from the uninsulated regenerator walls by about 13%. Tarnishing of the sheets had no appreciable effect on their  insulating properties. 2. An experiment was carried out with  insulation of full depth ZA C electrocast side- blocks in the melting  end. A panel  of four blocks was  insulated on one  side of the furnace while another  panel  was left without  insulation  on the other  side. After a campaign of twenty-nine months the insulated panel was no worse for wear than  the uninsulated one. Following the test, full depth ZAC walls have been   insulated  on  several tank furnaces,  with  appreciable saving of fuel. 

S. Kruszewski

GT V01 1960174-179

bedded in a refractory material provides an easy means of continuously recording chemical changes due to alkali penetration. It permits also the study of the kinetics of the refractory attack by molten glass. Great care must be exercised in conducting these experiments to exclude any change in the composition of the atmosphere because of the high sensitivity of these electrolytic circuits to the presence of reducing or oxidizing gases. Dissolution of refractory materials by molten glass does not appear to be an electrochemical process comparable to metal corrosion. Attempts to modify the course of the corrosion by injecting counter electro- motive forces into the refractory phase gave no positive results. Nevertheless, application of external electro- motive forces may enhance alkali ion migration and so accelerate the high temperature corrosion of the refractory surface. Such an electrolysis cannot extend very far from the electrode because of the low conductivity of the crystalline phases. At present only preliminary kinetic studies by the electrical method have been made but the interesting results appeared worthy of notice.

L. Leger, M. Bofe & E. Plumat

GT V02 1961005-010

The heating of air or other fluids by heat exchange from hot gases frequently sets severe problems  in  choice  of equipment and materials of construction  owing to the high  temperature of the hot gases  being  handled. Such difficulties as scaling of metals and loss of  strength  are aggravated by the possibility  of local overheating caused by stratification  in the hot gas or by uneven  cooling on the air side of the heat exchanger. The source  of  hot  gas may be  either furnace   waste gases, or, with  independently $red heaters, combustion products from gas or oil burners. This paper describes methods  employed on independently .fired heaters   to reduce  and  control  combustion  product   temperatures without  loss of efficiency. 

C. J. V. Denning

GT V01 1960 139-143

Some experiments already reported measuring the effect of sand grain size on the refining of a soda-lime-silica glass were repeated in a furnace that did not cause relatively rapid convection of the melt. The results in the two investigations for melts founded from one sand grain size differed, but showed that the exponential decrease in the number (n) of seed per cm3 of constant diameter (x) was not caused by the effect of convection on the motion of the bubbles. Therefore the rise of bubbles through the melt was not chiefly responsible for their removal. Four sand grain sizes were used, 0.22, 0.15, 0-114, and 0.077 mm. The seediness early in refining increased as the sand grain size decreased, but the rate of refining showed a maximum at 0.15 mm. Melts founded from 0.114 or 0.077 mm sand refined very slowly.

Michael Cable

GT V01 1960130-135

The scale of operations and production methods in the glass industry in the thirty or so countries visited by the author in the past three years is reviewed. The regions visited include Western Europe, North and South America, the Caribbean, East Africa, the Middle and Far East.

F. G. West-Oram

GT V01 1960 227-236

Techniques which  can assist in forming a complete picture  of  the heat distribution  throughout a glass  tank system are  described, and the practical difficulties discussed.  Illustrations of measurements  made  on large glass tanks are included. It is suggested that heat surveys can provide useful information for the tank designer.

R. A. Kirby

GT V31 1990 025-027

Low expansion glazes have been applied to the surface of glass containers using a flame spraying process. Four glass compositions have been characterised as potential glaze materials, a zinc borate glaze showing the most promise. This glaze was applied onto a large sample of containers which were then pressure tested. The bursting strength of the containers was significantly increased by the application of this glaze.

M. F. Barker

GT V43 2002 89-94

Within the scope of increasing the reliability, the controllability and the speed of glass container forming machines and thus improving the related forming process, servo-electric drives are increasingly replacing pneumatically driven mechanisms that still represent the standard in IS machines today. This contribution discusses the design and the results of an experimental unit
of a fully functional servo-electric plunger mechanism that has been developed by Emhart Glass SA. Since the control of the parison pressing action needs to satisfy a variety of different challenging demands, a thorough understanding of the system is indispensable, and therefore a physical model of the servo-drive and of the glass forming process has been developed. It allows one to
simulate the complete system in virtual reality and to iteratively optimise the electromechanical device, as well as the control algorithm. Trials with the experimental unit that has been built based on the theoretical results confirm that the accuracy of the models is remarkable. Extensive field tests were successfully carried out at Wiegand Glas, Germany, in a real production environment.
The encouraging results demonstrate the high potential regarding reliability, controllability of the forming process, performance and customer benefit of a servoelectric plunger mechanism.

GT V01 1960 204-208

A total fuel saving of nearly 7000 tons of coal has been achieved since the waste-heat boiler was installed at Pilkington Brothers (Queenborough) Limited, and taking all items into account, there has been a substantial  reduction in the cost of steam services to the works. The performance of the boiler on continuous service is  entirely satisfactory.

F. P. Lockwood  

GT V01 1960 017-024

The methods used for determining liquid contents and solubilities in binary and ternary systems are described in detail by reference to the SiO₂-Al₂O₃, CaO- Al₂O₃-SiO₂ and Na₂O-Al₂O₃-SiO₂ systems. The solubility data are used to assess the comparative resistances to glass of ,fused tank blocks. Having shown that a combination of liquid content and viscosity data from the Ca0-AIz03-Si02 system can be used to predict the factors affecting the life of steelworks ladle bricks, the effect of composition on porous tank block life is similarly studied by means of the Na₂O-Al₂O₃-SiO₂ system. The desirability of considering highly siliceous compositions is stressed.

W. F. Ford 

GT V01 1960 200-203

Although an experimental set-up about ten years ago had been used to give air preheats in excess of 600" C, the first applications of Austeel-Escher recuperators to glass furnaces were designed to give preheat temperatures of only about 500 to 550" C. As greater experience has been gained and the characteristics of high temperature steels have been better understood, together with their greater availability, recuperators have been used to give air preheats in the region of 700 to 800°C. Methods of obtaining these higher preheats are described and particular applications to glass fibre furnaces and unit melters are referred to. Austeel-Escher recuperators have been used extensively in the steel industry and reference is made to some installations for preheating both gas and air. 

H. Kay & G. E. Walker  

GT V02 1961 011-022

Manufactured gases used for  heating  industrial furnaces and equipment  range from producer gas to town's  gas and  are  selected  according to the  requirements  of the heating process. The range of  different  gases which  can now  be  produced  has  been greatly  extended with the development of processes for the gasification  of  liquid fuels. This paper  reviews the methods available for gas production and deals  more  particularly  with  two-stage producer gas (Motala System) and   the M.S. oil gaszfication process. Conventional  single-stage producers  have fallen   into disfavour recently, partly  for   economic  reasons, partly because of  the dirt and labour associated with them, and also because of the necessity for the periodic burning out of gas mains when gasifying bituminous coals. The over- cracking of volatile matter which causes the  latter  is avoided in two-stage producers by separating the  two stages of gaszfication and  distillation, by the  use  of a deep, fuel bed, and by regulating the quantity of hot gas used to distil off  the volatile matter of the coal. The raw gas can be  distributed hot over long distances or, if desired, cleaning and partial or complete cooling can be practised.  Producers  of this type are limited to a fairly narrow  range of coals,  and  the  areas where these  are produced  are  indicated. The use of refinery products for  gas making opens up considerable scope for  producing  gases of a much wider range of characteristics. It is possible  to utilize a number of the products  available from the refineries including the re-forming of liquefied petroleum gases to gases of a more suitable  composition. The plants  which  have  been  developed to gasify these products  are generally  clean and very automatic, so that  they need little supervision. The technical  and economic aspects of such  processes  are discussed and  the likely fields of application outlined.

E. J. Lawton 

GT V01 1960 085-093

The present work shows theoretically and experimentally the effect of slight ovality on the stresses produced in nominally circular pressure containers, such as glass bottles. In the theoretical treatment variations in the material thickness are taken into account. Five different "" characteristic cross-sections have been subjected to special investigation. By taking certain conditions into account, very short cylinders may be used, and the measuring equipment then becomes rather simple. With well-defined cross-sections, good agreement is obtained between theory and experimental results. Specially prepared steel cylinders are used in the experiments. Other experiments demonstrate how the results can be adapted to glass bottles. Here difficulties in defining the cross-sections and phenomena of enclosure produce complications. The results obtained will most likely create demands for roundness in bottles exposed to internal pressure. With the existing trend towards light-weight bottles, the matter will be of increasing importance.

A.O. Hougen 

GT V01 1960 155-161

 A highly siliceous unfired ('Rodur') tank block with an alumina content of about 11% has been successfu1ly used in a two-year furnace campaign for the melting of borosilicate glass. Reference is made to the working conditions under which a cord-free hard glass could be melted in the Rodur tank. The material underwent an intensive chemical, physical and x-ray examination after use and an explanation for the surprisingly good durability was found.

K. Konopicky, H. Gotthardt & E. Kohler 

GT V01 1960 080-084

The paper reviews present integration in Europe, briefly touching on history and current forms of official co- operation, and surveys the British position vis-a-vis European co-ordination and the associated problems. The attempts by Britain and others to increase the scope of integration are referred to, as well as the prospects for the future and the special situation of the British glass industry.

Dennis Rider

GT V01 1960 162-173

The quality of refractory  materials comprising the chequers in regenerators of a  tank furnace has an important  influence on  the efficiency and  the  length of life of the furnace. The principal factors which have an effect on the  brick  are  the  action  of condensed alkaline salts, the   cyclical   atmosphere   variations,   and  finally the periodical  temperature  changes.  Generally, basic bricks resist the effect of alkaline salts but  they have a tendency to disintegrate  under the influence of changes in the gas composition if they contain an appreciable  amount of iron oxide or iron silicate. This fact  has also been established by using forsterite or magnesite bricks. It is also equally important to avoid the contact of these brick-s or dust  with  siliceous material, contact  with  which  pro- motes corrosion in the  high  temperature  zones of the chequer-work by the formation of eutectics. Attention vas then  directed to materials  with lower iron  contents  but  richer in magnesia. Long trials are described  which show the good  durability  of  magnesia bricks. The difficulty of assessing the value of  chequer bricks is due to the length of furnace  campaigns. A  quick and  representative trial is necessary. The authors  developed a special type of furnace which permits trials at different temperatures  and  controlled  atmosphere. The trial furnace is  a channel in which the  temperature decreases. The test pieces are put in horizontally and the fIame from the burner  is charged with alkaline  vapours, the flame being alternatively  reducing and oxidizing. An attempt has been made to reproduce in this apparatus the conditions  which exist in  the  regenerators, und it has thus been possible to study  the destruction  process of chequer bricks and also to select  materials. 

P. Eloy, M. Jaupain & E. Plumat 

GT V9 1968 089-100

When glass is broken by impact, failure may be initiated either by thejlexural stresses or by the bearing stresses. The former may be calculated by the Tuckerman formula, as amended by Preston, the latter by the formulae of Hertz. Both types of stress have normally to be considered, and not merely one.

J. L. Glathart, F. W. Preston

GT V01 1960 011-016

Samples of twenty refractory materials have been sprayed with K₂CO₃, Pb₃₀e, and NazSOa at 1400°C. Whilst the corrosion was related to the chemical composition of the samples, it has been shown that the porosity of the sample had an important influence on the result.

T. S. Busby

GT V01 1960 025-050

Glass Technical Services Limited, London This retrospective study of the technological changes in glass bottle manufacture which have led to complete automation in the glass container industry today covers the slow change over many centuries from hand methods of bottle manufacture to the early attempts, towards the end of the nineteenth century, to introduce machines. There follows an account of the astonishingly rapid developments during the present century dating from the introduction of the first commercial Owens machine and the first stream feeder in 1903, leading to the modern suction and6gob'-fedmachines of today. Reference is made to the leading personalities and pioneers in this revolutionary process.

Edward Meigh 

GT V01 1960 069-079

In a previous paper (J. Soc. Glass Tech., 1958, 42, 227 T), the authors showed the necessity for removing a proportion of the alkali from the surface of a glass of high alkali oxide content before applying a silicone film to ensure resistance to atmospheric weathering. The durability of the silicone film in severe climatic conditions was tested using the surface conductivity of the glass at room temperature as a criterion. The present paper is concerned with the various factors which con- tribute to the measured values of this surface conductivity, for example the time of exposure to humidity, polarization due to continuous direct current, and the influence of leaching and baking procedures employed for the surface pre treatment. Four types of glasses were studied: a high alkali glass, a soda glass, a lead glass and a borosilicate glass. The initial conductivity of the untreated glass surfaces was dependent upon the alkali oxide content of the glass composition. The initial rate of change of conductivity with exposure time was also influenced by the alkali content in a similar manner. Siliconed samples of these glasses were subjected to an accelerated climatic test at 55OC and 95% relative humidity. In all cases the breakdown of the silicone film was shown to be inevitable in these severe conditions, but a leach-bake procedure definitely improved the performance of the soda glass and gave a marginal improvement with the lead glass, but produced a deleterious effect with the borosilicate glass.

J. Edge & L. F. Oldfield

GT_V3_1962_124-129

When glass is reheated gas bubbles may form within the glass; this phenomenon is often referred to as reboil. Some glasses are much more inclined than others to show this instability, and the object of this work has been to determine factors that affect the stability of glass towards this type of thermal decomposition. In all glasses the water content was found to be important. In reduced carbon-sulphur amber glasses an additional factor which controlled reboil stability was also found to be present.

S. M. Budd, V. H. Exelby & J. J. Kirwan

GT V01 1960 055-068

The reputed quart, equal to two-thirds of  an imperial quart, is a firmly established, but unofficial, 'standard' capacity, for wine- and spirit-bottles, not only in the U.K. but also in most other parts of the world. Up to now, no satisfactory explanation, for this has been found. In an attempt to solve the mystery, the author has studied the development of English glass wine-bottles from the time when the first ones were made early in the seventeenth century, and also the related history of weights and measures. It was found that the reputed quart has been in continuous use for about three hundred years, and it is suggested that it was originally an unofficial variation on the legal wine quart, which itself had arisen largely by accident, following mistakes in the wording of the early laws.

B. E. Moody

GT V02 1961 023-034

Changes in producer gas composition  and heat content resulting from preheating have been  determined  on a number of industrial  furnaces. With regenerators of rectangular brick packing in which preheat  temperatures of over 1000°C are  obtained in normal  operation,  it is shown that decomposition of tar   vapours   increases rapidly at preheat temperatures above 900°C, and decomposition of fixed hydrocarbon gases above 1050°C. The thermal input  to the furnace  increases continuously with increasing preheat temperature, the gain in sensible heat far  outweighing the loss of chemical energy through soot  deposition. The only reason for  limiting  the preheat temperature  below the maximum attainable may be  the increasing loss of luminants from the gas associated with increasing temperature. In regenerators which gave lower preheat  temperatures in  normal  operation,  decomposition of hydrocarbons occurred at lower temperatures. It is suggested that  the relation  between preheat  temperature  and gas  composition  is dependent  upon the manner in which gas flows within   the   regenerator,   and   the lower gas  preheat temperature   associated   with   regenerators of lower thermal efficiency does not  necessarily give any advantage in reduced hydrocarbon decomposition. 

G. A. Sparham & D. W. Gill

GT V1 1960 191-199

The trends of furnace development within one company are discussed, with particular reference to regenerator performance, and experimental results are presented which illustrate the progress made. Consideration of these results indicates that the introduction of a recuperator after the reversing valve might be more profitable than further development of the actual regenerators. The performance of a number of waste-heat boilers is summarized, and details are given of one installation which includes a back-pressure steam engine. The possibility of operating condensing power plant in conjunction with waste-heat boilers is examined, and it is shown that the economics are largely dependent upon the size of the plant. 

W. Scott

GT V01 1960 011-016

Although all-basic furnaces  were tried in Great Britain  even before the war, it  is only in the last few years they have proved economical. The swing having started is proving violent, some five  all-basic furnaces in 1956 having risen to fifty by the beginning of 1960. Attempts to justify such furnaces  purely on the basis of longer life have proved futile. They can only become  economical  with faster driving-involving greater fuel input  rates  with or without oxygen for  combustion and lancing. Increased outputs of 10-30% are now being achieved with roof lives two or three times that of silica in spite of  substantially  higher operating  temperatures. 

J. H. Chesters

GT V01 1960 243-250

Although all-basic furnaces  were tried in Great Britain  even before the war, it  is only in the last few years they have proved economical. The swing having started is proving violent, some five  all-basic furnaces in 1956 having risen to fifty by the beginning of 1960. Attempts to justify such furnaces  purely on the basis of longer life have proved futile. They can only become  economical  with faster driving-involving greater fuel input  rates  with or without oxygen for  combustion and lancing. Increased outputs of 10-30% are now being achieved with roof lives two or three times that of silica in spite of  substantially  higher operating  temperatures. 

J. H. Chesters

GT V01 1960237-242

H. Southern

GT V01 1960 260-267

A brief historical introduction surveys the improvements in gas producers and the development of two-stage generators during the last twenty years. The principles of the two-stage generators are discussed and the separation of the process of distillation of the volatile products from the gasiffcation of the coke by the steam/ air blast is described. The practical development of commercial plants for the production of three types of producer gas (hot raw gas, hot detarred gas and cold clean gas) is described. Typical operating data are given and the utilization of the gas is discussed. The methods of gas cleaning and treatment of by- products are described. The two-stage method of gasification enables the by-products to be obtained in easily handled form, thus enhancing their value, and at the same time reduces effluent to a minimum. The disposal of effluent is described. The types and range of brown and bituminous coals which can be gasified in the two-stage producer are described and illustrated by reference to the fuels now being gasz9ed in two-stage plants in various parts of the world. This leads to a discussion of thermal efficiency both in respect of the gas and the by-products. Reference is made to the particular advantages of the two-stage producer which are helpful in fulfilling the requirements of the Clean Air Act 1956 and which can- not be realized with conventional single-stage producers. In the last part of the paper attention is paid to the economic questions relating to the production and utilization of two-stage producer gas.

S. Kruszewski

A. H. O. Johnson 

GT V01 1960 209-210

When the  Harlow Works of Key Glassworks Ltd. was designed no provision was made for the  usual boiler plant  because it  was decided that  all  the steam for the factory could be provided by waste-heat boilers attached to furnaces. Auxiliary   oil firing for    the   waste-heat boilers was provided for flexibility and to guard  against the possibility of condensation. The completed  installa- tion is described andparticulars are given of its practical behaviour and of its thermal performance  during several years' service.

W. A. Moorshead

"A Study of the Rupture of Glass."  Discussion

F. W. Preston

JSGT 1917 V01 T213-T222

During the course of experimental work carried on in the writer's laboratory on the subject of the resistance of glass towards the action of water and chemical reagents, it has been found necessary to consult quite a number of previously published papers on the same matter. As the study of the corrosion of glass is by no means exhausted, and in view of the increased interest being taken in the subject in this country, it is considered likely that the following bibliography may be found of service to other investigators.

W. E. S. Turner

JSGT_1918_V02_T088_T089 (2 pages)

A very hot blowpipe flame, without having recourse to the rather expensive method of using oxygen, is so desirable that several people have at different times attempted to solve the problem by preheating the gas and air before burning them. The ordinary furnace methods of preheating the gas and air by the waste heat are impracticable in this case, but recently Mr Whatmough has described a successful method of preheating the gas and air mixture by passing it through a metal tube heated to redness by an auxiliary gas supply. The use of this type of burner, in which the gas and air are mixed before being burned, necessitates the employment of gas under a higher pressure than is obtainable directly from the mains. and some manufacturers hesitate to install a high pressure gas system in order to use such blowpipes. No such obstacle should be allowed to stand in the way of manufacturers who want to increase their output.

S. English

JSGT 1946 V30 T198-T200

A brief history of the Wear Glass Works, Greener & Company, taken over by James A. Jobling in 1875. A maker of cylindrical prismatic marine or “anchor” lights, pavement glasses, railway bull’s eye lenses, cans and tumblers, domestic ware, and marbled tiles. Changing it’s name in 1925 to James A. Jobling & Co. Ltd.

C. J. Purser-Hope

JSGT 1957 V41 T168-T184

The author reviews what is now known of the activities of Russian glassmakers during the eleventh to the thirteenth century, particularly at or in the neighbourhood of Kiev, Vyshgorod, Rogatchev, Minsk, Grodno, and Smolensk, and then presents a selection of the many analyses and results of other tests he has made during recent years of refractory materials and glasses found at the above sites. From the analyses of five crucibles and one furnace brick it was shown that clays of siliceous type were used at Kiev and could have been derived from a local source. Out of one hundred and seventeen analyses of glasses made by the author, he presents in this paper eight of variously coloured mosaic tesserae, six of coloured beads, seven of coloured bracelets, six of window glasses, six of hollow-ware, and five of unshaped factory melts. The striking feature of the analyses is that the majority of the glasses contained lead oxide as a major constituent, not only the mosaic glasses, in several of which the amount exceeded seventy per cent, but even four of the six specimens of hollow-ware and two of the six window glasses had lead oxide as a major component. Potash was the dominant alkali; and potash-lead oxide-silica glasses were shown to have been in use for hollow-ware, as well as for beads and bracelets, several centuries before the adoption of glass of this type in England.

M. A. Besborodov

JSGT 1936 V20 T338-344

A number of commercial soda–lime–silica glasses were analysed in two laboratories by two methods: (1) the standard procedure involving decomposition of 1 g by Na₂CO₃, two evaporations with HCl to remove SiO₂, removal of As by H₂S, precipitation of Al+Fe by NH₄OH, and of Ca as oxalate, and ignition to oxide; (2) decomposition of 1 g by HF and H₂SO₄, followed by analysis of the HCI solution of the residue, as in (1). Provided the HF decomposition was conducted with care, there was no appreciable difference in the results for Al₂O₃ and CaO by the two methods. The comparison was extended to glasses containing both CaO and MgO.

J. D. Cauwood & Violet Dimbleby

JSGT 1939 V23 T129-T140

Using the disappearance of batch as a criterion, a comparison of the relative rates of melting of a series of glasses has been made by two methods. Both methods gave substantially the same results. The glasses were made from soda ash, dolomite limestone, and silica sand. The composition range studied was 10 to 18 per cent Na₂O, 8 to 22 per cent (CaO + MgO), and 63 to 80 per cent. SiO₂.

J. C. Potts

JSGT 1937 V21 T350-T355

Constructional details are given for chambers designed to give accurate control of temperature and humidity, in investigations on certain properties of glasses, such as chemical durability and electrical resistivity.

E. Seddon

JSGT 1953 V37 T057-T060

An outline is given of some of the important features of the design of two-table blow-blow machines.

N. E. Densem

JSGT 1953 V37 T080-T087

The available feeder-fed bottle-making machines are classified and a detailed assessment is made of their usefulness for different purposes, and of their performance and maintenance.

W. A. Moorshead

JSGT 1919 V03 T027-T034 (8 pages)

Having listened with great interest to the paper read by Mr Mills I should like to explain a method of costing for glass bottle production which I adopted a few years ago after giving a great deal of consideration to the subject and testing various methods suggested. I rejected the weight unit: as a single unit, as not being trustworthy, and one which if employed wholly would be bound to lead to disastrous results if a change in the proportion of the various descriptions of bottles manufactured should be made.

F. Sweeting

JSGT 1922 V06 T030-T045

Many and various are the methods which have been proposed and employed from time to time for the comparison of the durability of glass, ranging from the rapid method of boiling glass powder in water to the slow one of exposing slabs or cubes to the action of atmospheric air for a period of a year or more. Since 1917, several new, or at least modified, methods have been employed, including the boiling of glass discs in the reagent, the passage of water vapour over polished cubes, and the heating of pieces of glass with water in sealed tubes. It is not the purpose of this paper to examine each of the various methods proposed, but rather those three or four which have been employed fairly generally. These methods involve (1) the testing of the surfaces of glass vessels, (2) the use of plates, discs or slabs of the glass, (3) the treatment of the glass in the form of powder, (4) the autoclave test. 

W. E. S. Turner

JSGT 1936 V20 T418-T427

A test for the durability of 4 oz medicine bottles was developed by the co-operation of six laboratories. It consists in determining the amount of alkali extracted from the inner surface of the bottle in five hours by boiling water. The test is examined in the light of information obtained by its continued use in a modified form, and the· possibility of its extension considered. The effects of glass composition, size and shape of container, conditions of manufacture, annealing and storage have been examined.

A. Cousen

JSGT_1927_V11_T124_T153

In the present paper it is proposed to examine and criticise the various methods in use for the analysis of glasses of the more common varieties. The glasses comprised in such a scheme normally contain silica, iron oxide, alumina, manganese oxide, magnesia, soda, and potash, with sulphate, chlorides and sometimes the oxides of arsenic and antimony in small proportions, and will include the soda-lime series of glasses. The scheme for the lead glasses will differ only from the above in having to consider the determination of lead oxide instead of lime.

W. H. Withey

JSGT 1934 V18 T112-T121

Present conditions are reviewed from the points of view of the buyer, the manufacturer and the designer. Factors governing the choice of closure are discussed. The designer should have personal contact with the glass-maker, as well as with the purchaser. The advantages of originality and novelty are discussed and the trick of deceptive capacity is condemned. Pottery shapes and ornamentation should not be reproduced by the glass designer. There should be stronger links between the glass manufacturer, the designer and the consumer.

Kenneth L. Graham

JSGT 1929 V13 T003-T015

This brief note is added as an extension of my former paper (wherein we traced the connection between fracture-systems and the stresses producing them) on account of a deceptive similarity between two sorts of radiating fissures which ought to be recognised and distinguished. If they are confused with one another, a works superintendent trying to cure the trouble will very likely do exactly the wrong thing and make his glass worse rather than better. I am referring, of course, to trouble experienced in annealing large flat sheets of plate or wire glass, which may break in subsequent operations, such as grinding or polishing or glazing.

Linked to JSGT 1926 T234-T269

F. W. Preston

A Further Study of the Performance of a Glass-Melting Tank Furnace Fired by Coke Oven Gas

JSGT 1946 V30 T013-T030

This paper is divided into four sections. Section 1 deals with questions raised in an earlier paper concerning the methods of making allowance for furnace temperature changes when fitting regressions to the observations. It is concluded (a) that the amounts of heat associated with the temperature changes which normally occur are so small that they can be ignored in a full analysis, thereby reducing arithmetical labour, (b) that using the data associated with the constant K_Tin the Furnace Performance Formula is not appropriate. Section 2. Further data are presented for the furnace discussed in the earlier paper. The utility of plotting Q, the Thermal Performance Index, is emphasised, and is shown to direct attention at the earliest moment to changes in performance which require a technical explanation. It is shown that the more formal analysis of the data leads to similar conclusions, and in particular that the lower values of Q obtained during the earlier weeks of the melting campaign now discussed, relative to the corresponding values of the previous campaign, are associated with a lower value of the quantity of gas required per day to maintain the new furnace at a temperature of 1400° with zero glass load. Section 3. Attention is drawn to the usefulness of the coefficient b of the formal analysis as an estimate of the amount of fuel needed to produce a ton of glass in the furnace concerned. If adjustment of fuel supply be possible, then this constant gives guidance as to the magnitude of the fuel change which should be made to accord with inevitable changes in the demand for glass. Section 4. A detailed analysis of data obtained daily for a period of six weeks shows that there are marked differences in the amount of glass melted which are not accompanied by corresponding changes in the fuel consumed. This suggests a possibility of further economies in fuel.

B.P. Dudding

JSGT 1954 V38 T457-T469

A pyrometer is described which has been developed for the measurement of the surface temperature of glass. The temperature range covered is 150°-550°. The theory underlying the design of the pyrometer, which is sensitive over a selected band of wavelengths in the infra-red, is outlined, and the errors which may be encountered in practice are discussed.

J. R. Beattie

JSGT 1959 V43 T059-T061

A rapid method for the gravimetric determination of silica in simple aluminosilicates has found widespread use in the ceramic industries. The sample is fused in sodium hydroxide and dissolved in hydrochloric acid. Ammonium molybdate is added to an aliquot to form silico-molybdic acid, which is then precipitated as quinoline silico-molybdate. This precipitate is then filtered off, dried, and weighed. The accuracy of the method is about +0·0%, −0·5% SiO₂, and the precision about ±0·25%. Borate and fluoride do not interfere.

H. Bennett

JSGT 1948 V32 T127-T146

A theory of the strength of glass based on the kinetic energy of the atoms has been developed on the assumption that, if the thermal energy of the atoms is distributed in accordance with the Maxwell–Boltzmann law, there is a finite probability of some of the atoms having sufficient energy to cause a local discontinuity with consequent local stress enhancement.

S. M. Cox

JSGT 1956 V40 T003-T017

The characteristics of blocks made from Indian Kyanite were well known and it was therefore decided that the most useful results would be obtained by making a direct comparison between the properties of blocks made with Indian and Kenya Kyanite. These materials, of course, are calcined before incorporation into the blocks and at the calcination stage they are converted into mullite and silica. In this report the term 'Kyanite' is used to refer to pre-calcined Kyanite, either from India or Kenya. As it was desirable to obtain preliminary information as soon as possible, it was arranged in the first instance to compare the blocks by means of laboratory tests only.

Refractories Committee

JSGT 1950 V34 T017-T024

A method is described allowing the determination of the porosity of substances in grain form, such as grog for refractories and other ceramic purposes, and of coke, ore and insulation aggregates. Data concerning the accuracy and limitations of the method are given. Some possibilities of application are described.

B. Eicheler

JSGT 1951 V35 T320-T331

The viscosity of highly viscous resins was examined using circular rods suspended in a manner similar to that of an horizontal beam fixed at one end. The rods were mounted in drill chucks capable of rotation about an horizontal axis. By continuous rotation, temperature equilibrium was reached without deformation, so that a series of experiments could be carried out successively on the same body. The rate of deformation for various rods of different sizes was determined, whilst the viscosity was measured by means of a Pochettino viscometer. Good agreement between the relation found by experimental research and the theoretically derived formula was proposed. This method, applied at higher temperatures, may prove suitable for application to glasses for the determination of the viscosities of the order of 10⁶ poises and above.

Finn Gronlund

JSGT 1951 V35 T490-T496

Apparatus has been designed, and is now described, for the determination of the viscosity and the flow characteristics, under known loads, of specimens of glass of fairly large size at temperatures below the softening point. For a glass of the approximate percentage composition SiO₂ 72, R₂O₃  l·5, CaO + MgO 12, Na₂O + K₂O 14·5, the viscosity at 525° was logη=14·35. Preliminary results for the temperatures 448° and 400° are quoted.

G. M. Blaimont, J. R. G. De Bast, P. G. Migeotte & H. P. C. Vandecapelle

JSGT 1932 V16 T284-T292

The author directs attention to fundamental weaknesses in the design of the existing typical tank furnace. He has designed a furnace in which the melting is carried out in a cylindrical chamber rotating uniformly at controlled speed, and claims for it a higher thermal efficiency and a more homogeneous glass product.

G. A. Zotos (Translated by W. Butterworth)

JSGT 1935 V19 T118-T124

Mention is made of the origin of sands and the impurities likely to affect their suitability for the manufacture of colourless glass with special reference to British sand deposits. Several chemical methods are referred to which have been tried in the past for treating sands to reduce their iron oxide content sufficiently to render possible their use for the manufacture of good quality colourless glass. A new process is outlined in which the sand to be treated is agitated for a short time at a temperature of 26° to 65°, with a solution of acid oxalate, together with small quantities of ferrous sulphate.

F. W. Adams

JSGT 1955 V39 T162-T172

The thermal efficiency of glass-melting is improved by infusing the fuel, air and gas which are previously compressed, into the mass of molten glass; and a flame-injection furnace incorporating this principle of firing is described. Importance is attached to conservation of heat, of which full use is made by feeding the batch in the form of a spray through the exhaust gases. A sandspray heat exchanger, reaching an efficiency of 60%, is described. Results are given of experimental plant trials.

E. Brichard

JSGT 1918 V02 T030-T032

In lampworking processes it is often necessary to change quickly from a large to a small flame or vice versa, and to do this it is essential to change the size of the air jet, as it is impossible to get a large and a small flame satisfactorily with the same jet. Attempts have been made at various times to construct a blowpipe which would allow a quick and easy change from one size of jet to another, but none of these blowpipes seemed to the writer to be entirely satisfactory. In the new variable jet blowpipe three separate jets are provided, and the whole instrument has been arranged to follow as nearly as possible the lines of the standard blowpipe.

S. English

JSGT 1936 V20 T416-T417

The powder is carefully prepared so as to ensure uniformity of grain size. Comparisons of solubility are made by taking standard quantities of glass and water and adding a standard quantity of a mixture of three indicators. The pH values of the solutions are taken by matching with standard buffer solutions.

A. R. Wood

JSGT 1951 V35 T407-T410

A survey has been made of the possibilities of glass formation in a restricted part of the potash–lead oxide–alumina–arsenic and of the potash–barium oxide–alumina–arsenic fields. The chemical compositions and physical properties of two of the glasses are recorded.

A. E. Badger & D. E. Sharp

JSGT 1938 V22 T038-T040

The principles are discussed on which the surface tension should be correctly derived from application of the fibre weight method. A new equation has been deduced by which the surface tension values agree with those derived from the bubble method.

J. Bailey

JSGT 1955 V39 T211-T214

It is a fact well known to makers of glass-metal seals that the low oxidation resistance of molybdenum and tungsten largely determines the techniques which must be used in making glass seals to these metals, and also limits the conditions under which the seals can be used. The layers of tungsten trioxide and molybdenum trioxide, which form on heating these metals in air, have very little protective action. The oxide layer produced on tungsten is porous and easily breaks away from the underlying metal, whilst molybdenum trioxide is volatile at temperatures above 700°, and so has no protective action whatever.

H. Rawson

JSGT 1948 V32 T378-T381

The author, in opening the discussion on papers read at the Saturday morning session, October 2nd, 1948, of the Symposium on Some Aspects of the Nature and Constitution of Glass, gave a resume of work carried out during recent years in Germany.

Adolf G. Smekal

JSGT 1927 V11_T023-T029

This note does not contain any material new to science: it frequently appears, however, that the subject is one of peculiar difficulty to most of those who would like to take an interest on the subject or to make use of photo-elastic phenomena in solving definite technical problems. The difficulty is apparently a composite of two difficulties: first, the difficulty of comprehending the exact relation between double-refraction and stress, which may be called the physics of the photo-elastic problem, and, secondly, the difficulty, more serious with most people, of visualising a whole stress system and recognising what is and what is not possible in such a system.

F. W. Preston

JSGT 1958 V42 T248-T252

A technique is described for the application of continuous coatings of a high expansion barium phosphate glass to uranium. The coatings were fired on at a temperature of 650°C in an atmosphere of purified argon. Best results were obtained when the coating was applied electro-phoretically and when metal of fine grainsize was used.

D. Hazelby & H. Rawson

JSGT 1934 V18 T071-T078

Investigations have been made of the temperature distribution in and around a pot during the arching period. Three such investigations are recorded, the points for temperature measurement being at the shoulder, inside and outside the base of the pot, respectively. Considerable differences in temperature have been found, which amount, in one case, to 170° between the first two positions. A gradient was found of 10° per inch across the pot bottom in each case. In two of the three tests reported there occurred frequent sudden fluctuations in temperature over the whole of the arching period.

W. Maskill

JSGT 1945 V29 T266-T267

The heavy mineral content in sixteen British sands which have been used in varying amounts for glass-making has been determined by separation in bromoform. It varied from 0·03 to 0·30%. The minerals identified were magnetite and limonite, ilmenite and leucoxene, tourmaline, kyanite, zircon, rutile, garnet and chromite.

W. Davies & W. J. Rees

JSGT 1935 V19 T040-T042

It sometimes happens that when one branches from the broad road of routine into a fork, the way is not always so smooth as the entrance might indicate. An example of this was experienced in the conversion of an ordinary soda–lime–silica glass into a glass in which some lime was replaced by barium oxide (to the extent of about 2–2½%) in order to give a softer glass without increasing the soda content. Since barium produces colourless salts, the way seemed clear and free from complications. It soon became evident, however, that all was not well with the colour of the glass. The glass steadily became more and more green despite efforts to adjust the selenium short of panic measures.

E. J. C. Bowmaker

JSGT 1933 V17 T093-T101

Measurements of the porosities and of the permeabilities of refractories to gases axe reported; it is shown that porosity measurements offer no guide to permeability except under such special circumstances as are not likely to occur in commercial usage. An apparatus is described which enables measurements of permeability to be made at elevated temperatures. The influence of temperature on permeability has been measured in the case of some experimental refractories.

A. Eric J. Vickers

JSGT 1924 V08 T162-T173

This brief note has been prepared at the request of your President, at very short notice, and it can necessarily, therefore, only touch the bare outlines of this subject. It may, however, serve to suggest to glass technologists a range of materials from which certain alloys may prove of value in glass-making operations.

W. R. Barclay

JSGT 1937 V21 T264-T269

The necessity for reliable data on the under load characteristics of refractory materials is stressed, and the information yielded by the various test methods briefly reviewed. The essential requirements of methods of testing are indicated, and new apparatus described which meets them, including means for eliminating the effect of expansion, other than that of the specimen.

H. C. Biggs

JSGT 1956 V40 T025-T031

Sodium silicate, approximately Na₂O.2SiO₂, has been saturated with water vapour at 1, 0·238, 0·117, and 0·064 atm partial pressure in nitrogen at 900, 1000, and 1100°. Samples were analysed by pumping off the water vapour in vacuum at 1200° for 3–10 h and measuring monometrically. The solubility is of the order of 3 N.cc/g at 1100° and 1 atm. From 900 to 1100° the solubility decreases by only 10%. Solubility as measured is not strictly proportional to (p_H2O)^1/2 as suggested by the relation H₂O + ≡Si–O–Si≡ ® ≡SiOH + HOSi≡ but falls below the expected figure at low partial pressures. This is thought to be due to the production of metallic sodium vapour by dissociation of the silicate melt in vacuum, which condenses in the cool parts of the apparatus and removes water vapour.

J. W. Tomlinson

JSGT 1947 V31 T147-T152

From many test results, correlation coefficients have been calculated between the porosities and the after-contractions for a group of cast sillimanite and a group of fireclay samples. For the sillimanite there is poor correlation and for clay good correlation. This divergence is explained as arising from a difference in the shapes in the two materials of the cavities which give rise to the porosity.

S. C. Waterton

JSGT 1953 V37 T302-T305

A. A. Griffith attributed the discrepancy between the theoretical and practical mechanical strength to flaws or cracks of varying severity. Because of the flaws which occur at the surface of a fibre, atmospheric attack would be expected to wield a further influence, a phenomenon which has been amply demonstrated. One of Griffith's principal arguments in favour of the flaw theory was a dependence of mechanical strength upon sample size. Thus, he found that the strength of his glass fibres rose rapidly with decreasing diameter, and indeed this has been demonstrated by others since that time.

S. Bateson

JSGT 1923 V07 T130-T131

The residues or tailings from the crushing of auriferous quartz for the extraction of gold are highly siliceous and as they are a waste product of considerable bulk, it appeared possible that they might be utilised as "sand" in the manufacture of common glassware in districts contiguous to ore treating plants where supplies of free sand are not cheaply obtainable.

W. J. Rees

JSGT 1958 V42 T119-T124

The use of the Senarmont method of photo-elastic measurement is advocated for the measurement of stresses in glass and glass-metal seals. Some of its advantages over more widely adopted methods are given.

H. Rawson 

JSGT 1952 V36 T297-T299

The writer believes that in this analysis lies a possible explanation for the existence of a maximum velocity of crack propagation in glass. This explanation may be stated quite simply as follows: the techniques used for measuring the velocity of crack propagation in glass are based on the assumption that the fractures are proceeding in straight lines. They are, in fact, incapable of measuring the velocities of cracks proceeding in any other manner. The positions of the fracture fronts in Barstow and Edgerton's experiments are determined by those fractures which are moving radially in straight lines as may be seen by examining their photographs. On this basis, then, the maximum measured velocity of crack propagation is equal to the maximum velocity at which a crack can continue to move in a straight line.

H. Rawson

JSGT 1934 V18 T222-T224

Fused transparent silica discs suffered no loss in weight when heated in air for 20 hours at 1400°. Experiments in which similar discs were subjected to the action of steam showed that in 20 hours at 1300° the loss in weight was negligible, but in 20 hours at 1400° and in 4 hours at 1500° the loss was small but appreciable. The surface in contact with the steam became slightly pitted and glossy.

Eric Preston & W. E. S. Turner

JSGT 1934 V18 T067-T068

As bearing on the recent paper of W. M. Hampton, this note directs attention to the stresses which may be set up in wire glass.

Link to JSGT 1933 T273-T286

Frank W. Preston

JSGT 1933 V17 T118-T121

The note comments upon the measurements of the rates of evaporation of BaO, SrO, and CaO, made by A. Claassen & C. F. Veenemans, deduces from their data certain values of the vapour tensions of these oxides, and draws the conclusion that volatilisation losses of CaO from molten glass will be extremely small.

Eric Preston

JSGT 1948 V32 T147-T153

Glasses are described which have been developed for use as “solder glasses” and some possible applications of those glasses are suggested.

A. E. Dale & J. E. Stanworth

JSGT 1945 V29 T124-T127

Two models of projection type strain viewers have been designed purposely for the examination of small glass objects, and particularly of seals between glasses and glasses, and glasses with metals. They are illustrated below, and one of them is described in detail.

J. Fortey

JSGT 1941 V25 T249-T269

From consideration of a basic heat-exchange equation a logical method has been derived for determining the performance figure of a tank furnace irrespective of the type of glass melted and the melting temperature. The following factors have been considered: the effect of temperature' on fuel consumption, the effect of the size of the furnace, the quality of the fuel used, the design of the furnace, and the rate of melting and rate of withdrawal. It is shown that insulation of the furnace refractories is likely to- give a greater saving of fuel than changes of design in the furnace str11cture. The· colour of the glass affects the performance figure, and an attempt has been made to take this factor into account in view of the differences in the heat-absorbing properties of colourless and coloured glasses.

W. M. Hampton, B.Sc., Ph.D., F.Inst.P., F.I.C., F.S.G.T.

JSGT 1919 V03 T249-T253 (5 pages)

On the occasion of the annual meeting of the British Laboratory Ware Association, Ltd., in January last, a request was made to one of us to undertake an investigation which would secure a much needed and ample supply of glass tubing of uniformly good quality of British manufacture for lamp-workers. This subject was brought before the committee on the standardisation of laboratory glassware of the Society of Chemical Industry, but as its terms of reference, did not include standardisation of glass tubing, no action was taken. A research committee of the Society of Glass Technology has been appointed, however, the scope of which includes lampworking glass as well as other types.

F. W. Branson & F. H. Branson

JSGT 1936 V20 T498-T510

Numerous experiments conducted in several laboratories have been carried out on glasses of various compositions, to ascertain how far small lengths of glass rod 5 cm long and 2–6 mm in diameter can be used to give a satisfactory test of thermal endurance, by heating in an electric furnace and quenching in water. The results have been subjected both to general and to statistical study, and encourage the hope that a general test can be formulated.

JSGT 1939 V23 T372-T377

By the addition of a small quantity of stannic oxide to quartz a glass, similar in its physical properties to pure fused quartz, is obtained which cuts off sharply all radiation below about 2800 Å and has a very steep transmission curve in this region. Its characteristics, methods of fusion, properties and uses are discussed.

A. J. Maddock

JSGT 1958 V42 T101-T108

Determination of the rate of throughput and an assessment of the degree of mixing occurring in a glass-melting tank furnace have been made using tantalum as a non-radioactive tracer. After addition of a small quantity of tantalum pentoxide to the glass batch the relative amounts of tantalum appearing in the product has been determined by irradiating samples of the glass in a high flux of neutrons followed by measurement of the activity induced in the tantalum present in the glass.

R. S. Barker

JSGT 1923 V07 T122 T129

Some time ago we were invited to express an opinion on tests outlined by L. Kroeber for the purpose of ascertaining the suitability of glass for making ampoules. For the purposes' of the test, the ampoule might be filled with one of the following solutions: (1) 1·0 to 2·0% of morphine hydrochloride: (2) 0·5% solution of strychnine nitrate; (3) 1·0% solution of mercuric chloride; (4) a dilute solution of phenolphthalein; (5) 0·1% solution of narcotine hydrochloride.

H. S. Blackmore, Violet Dimbleby & W. E. S. Turner

JSGT 1952 V36 T287-T296

The Adams and Williamson Law has been somewhat generally misinterpreted, in part because it has been generally extended over space instead of being confined to time. When restricted to what was actually observed, the law becomes intelligible and its seeming inconsistencies are straightened out. It is then seen that it is not “purely mechanical” but includes a term representing the “constitution” of the glass; it indicates that the constitution changes with time during the annealing process; and it gives a quantitative estimate of that change of constitution.

F. W. Preston

JSGT 1936 V20 T405-T415

The work is reviewed of a special Committee of the Glass Division of the American Ceramic Society on a powder method designated as the ACS No. 1 method, and on comparative results by the Sheffield method. The effects of different extraction liquids, the time and temperature of extraction are examined, and a test is described to ascertain the tendency to fogging.

W. C. Taylor

JSGT 1931 V15 T262-T306

The following paper is a review of some phases of the technical literature and researches on glass since the papers of E. Zschimmer. Prof. Zschimmer discussed the published work on the physical properties of glass, and covered quite completely the subjects of devitrification, viscosity, annealing phenomena, density, mechanical properties, thermal expansion, optical constants, and electrical conductivity. The main purpose of this work of Zschimmer, as the title indicated, was to show the relationships existing between physical properties and the composition of the glass. Since this work covered the field so completely, and would need little modification on the basis of more recent work the present paper will review more particularly the question of temperature effects on the physical properties of glasses, and will not discuss at any length the question of composition. Thermal effects cannot be entirely separated from composition, inasmuch as many phenomena which have been observed and ascribed to temperature effects have been later found to be due to composition changes, and of course during devitrification there are localised composition changes resulting from the thermal treatment of the glass.

Link to JSGT 1928 T082 and T333

J. T. Littleton

JSGT 1933 V17 T169-T238

The Report is written under the following headings, which furnish a summary of its contents:  

I. Introduction. 

II. Raw Materials. 

III Grog. 

IV. Preparation of the Raw Mixture. 

V. Methods of Manufacture: (l) Hand moulding of plastic mixtures; (2) Hand moulding-ramming method; (3) Mechanical ramming or beating processes; (4) Machine pressing-semi-dry process; (5) Extrusion; (6) Casting from slip; and (7) Casting from electrically fused material (electro-casting). 

VI. Drying and Firing. 

VII. Faults in Blocks. 

VIII. Methods of Testing. 

IX. Experience Gained from Practice: (1) Influence of batch composition; (2) Influence of furnace design, especially of the tank (a) Cooling and insulation and (b) Block size and joints; and (3) Influence of tank block quality. 

X. A Suggested Technique for Testing Tank Blocks. 

XI. A Tentative Specification for Tank Blocks.

J. H. Partridge

JSGT 1919 V03 T258-T260

IN a recent lecture at Stourbridge, Dr Turner directed attention once more to the, desirability of examining all  articles, especially heavy blanks, in order to see that they were well annealed before being sent forward to the  cutting shop, as it is not an infrequent experience that a glass dish or other .article is shattered on the wheel after a good deal of time has been spent in its partial decoration.

S. English

JSGT 1955 V39 T193-T204

A simple apparatus for the measurement of the DC conductivity of glasses at low temperatures is described. In order to make observations at low currents a pH meter is used, a range of measurement of specific resistances from 10⁸ ohm cm to 5×10¹² ohm cm being possible using the instrument in its original form. A modification of the pH meter permits this range to be extended up to a specific resistance of 10¹⁶ ohm cm. Typical results obtained with the apparatus are given.

H. E. Taylor

JSGT 1918 V02 T216-T219  (4 pages)

When proper care is taken, the calibration of a pipette by this method would not take less than fifteen minutes. Although this time is far too long to make the graduation of pipettes a paying proposition, any speeding up of the process results in errors due to the neglect of important details.

S. English

JSGT 1918 V02 T240-T241  (2 pages)

Communicated Discussion from J.D. Cauwood on English's paper on pipette calibration.

S. English

JSGT 1929 V13 T015-T016

The following simple test for the detection of small quantities of iron oxide in glass sands was evolved during the course of other work.

J. T. Randall & R. E. Leeds

JSGT 1949 V33 T077-T081

The measurement of the stresses in sandwich seals and bead seals of glasses and metals is in common use in valve and lamp manufacture for the routine control of the uniformity of glass- and metal-sealing properties. In such routine tests it is often appropriate to use the metal in the same sheet thickness or wire diameter as is employed in production. If, however, the aim is to make a general estimate of the degree of “match” between various glasses and metals, without reference to a particular single application, and when comparison of results between several laboratories is desirable, there is need for the seals to be of standardised dimensions, and also for the methods of determining the stresses to be standardised. The method now described provides these standards for reasonably matched combinations of glass and metal.

A report of the Society’s Committee on Glass: Manufacture and Properties

JSGT 1924 V08 T284-T285

Some time ago my attention was directed to a curious phenomenon which occurred during the emptying of a tank furnace used for melting bottle glass. The gas was kept on while the tank was being tapped, thereby keeping the melting end hot. As soon as the dog hole was exposed, the blocks surrounding it began to become vesicular and ultimately to flow down on the side adjacent to the melting end, the part of the block not exposed to the flame, namely, that abutting on the working side of the bridge, not behaving in this manner. When removed, after the tank had become sufficiently cool, the blocks attacked were found to have very spongy ends, as shown in the accompanying photograph. The body of the block was found to be thoroughly sound, so that the initial suggestion of faulty tank blocks was found to have no basis whatever.

W. E. S. Turner

JSGT 1935 V19 T125-T138

Thirteen specimens of cut crystal glass ware have been examined for general appearance, chemical composition and angle of cut. The decoloriser where employed appeared in all cases to be manganese: in three specimens, only arsenic was found and in two, traces of nickel were present in addition to manganese, but no other decolorisers were found. The angles of cutting differed greatly. British samples had the widest angle, partly accounting for the lack of high internal reflection and apparent brilliance such as was secured by the narrow angle of cut of a Belgian piece. Proofs taken from melts at various times during the working of a pot and examined for iron oxide content supported previous experience and results in showing an increase of the iron oxide content on standing. Examination of the cut and polished surfaces under the microscope made it possible to criticise the care with which the cutting and polishing processes had been carried on.

Eric Preston, A. J. Holland & W. E. S. Turner

JSGT 1948 V32 T309-T339

Changes of density with time in the transformation range have been studied in vitreous silica, a borosilicate glass and a soda–lime–silica glass. In previous papers attention has been drawn to the close analogy between the atomic structure of glasses and liquids, and the results of the present experiments are interpreted in the light of that analogy. As was expected this has necessitated consideration of the behaviour of more complex liquids. In the case of the borosilicate glass it was found that there are two processes, at least, causing volume changes. The slower one at a given temperature leads to an increase in volume as equilibrium from a higher temperature condition is approached, the other to a decrease. Possible interpretations of this behaviour are indicated in terms of the temperature variability of the miscibility of liquids and the behaviour of solutions in mixed electrolytes. In the case of fused silica the experiments suggest a negative coefficient of thermal expansion when equilibrium is attained. This is supported by other published work on vitreous silica and is now interpreted as indicating an increase of structural binding as the temperature decreases, that is, to the same underlying cause as the phenomenon of maximum density in water. No anomalies have been found yet in the case of soda–lime–silica glass, but in this case the time needed to reach equilibrium in the temperature range studied is so long that the work has not progressed very far.

R. W. Douglas & G. A. Jones

JSGT 1933 V17 T384-T410

Under free crushing conditions, whether the load be applied by simple crushing, shear or impact, it is found that homogeneous brittle solids break down with a constant fracture pattern which is independent of the size of the original specimen. This law must be interpreted statistically, since wide variations will be found in the behaviour of individual particles. It thus appears that the minimum work required to crush a powder depends on the product of:- (1) The constant for the material; (2) The weight of the material crushed; and (3) The logarithm of the total mean reduction (=log(The mean original size/The mean final size). To crush coal and anhydrite through a 10:1 reduction ratio involves an energy consumption of 0·05 and 0·15 kWh/T, respectively. Material of ½ in size crushed to 200 mesh involves a mean reduction ratio of about 100:1 and the work necessary is thus 0·1 kWh/T for coal and 0·3 kWh/T for anhydrite. Values of a similar order have been obtained for other materials, and indicate that industrial crushing requires about 100 times as much energy as that theoretically necessary. This means that the industrial methods of crushing are very wasteful in energy compared with the process adopted by a labourer crushing stones. Industrially this information is perhaps of little value, but it does give a basis on which to define the efficiency of other grinding processes in which the loading conditions of the individual particles are difficult to define.

W. F. Carey & C. H. Bosanquet

JSGT 1948 V32 T040-T045

The effect has been determined of adding varying amounts of CeO₂, V₂O₅ , TiO₂, Cr₂O₃, and of combinations of them (and also Co₂O₃) on the colour and spectral transmission of a soda–potash–lime–silica, and of several soda–lime–silica glasses. From the results, several possible commercial glasses were devised for protecting leather, plain and dyed, dyed silks and writing paper from change of tint due to the UV in sunlight. The glass regarded as best had the batch composition (parts by weight): sand 704, limestone 240, soda (with 15% water) 300, CeO₂ 19·5, CoO 0·05.

V. Ctyroky

JSGT 1934 V18 T225-T251

Determinations of the modulus of rupture have been made on 2300 specimens cut with a diamond under constant load, from 26 oz drawn sheet glass of composition: SiO₂ 72·22, Al₂O₃ + Fe₂O₃ 1·92, CaO 8·84, MgO 2·89, Na₂O 13·96%. The dimensions of the specimens were: length 10 cm (effective length between knife edges 7·6 cm) width 0·6–1·5 mm, and thickness 0·25–0·31. The modulus of rupture varied according to a variety of conditions, in particular with (I) the position of the diamond-cut edge according as it was in tension or in compression, (2) the state of annealing or re-annealing, (3) the interval of time after re-annealing, a steady state being attained after several days. (For practical purposes three day interval was adopted.) (4) the width of the specimen, (5) the condition of the edges, (a) as cut with the diamond, (b) sealed by edge-melting. (c) ground and mechanically polished, (d) ground and mechanically polished, followed by fire-finishing. For specimens tested 3 days after re-annealing, with the diamond-cut edge in tension, the modulus of rupture varied from 613 kg/cm²  for a width 0·6 cm to 486 for width 1·5 cm. Specimens with mechanically ground and polished edges had a modulus of 572 kg/cm² for widths 0·89–0·93 cm. Other values were: with diamond-cut edge in compression, 903 kg/cm² for widths 0·89–1·14 cm; diamond-cut edge fire-finished, 901 kg/cm² for widths 0·83–1·09 cm and 1023 kg/cm² for widths 0·72–0·93 cm; cut edges mechanically ground and polished and then fire-finished, 940 kg/cm² for width 1·02 cm and 1123 kg/cm² for width 0·8 cm. Specimens tested with the diamond-cut edge in compression gave values of the same order as those with fire-finished edges, and much greater than with cut edges in tension or edges mechanically ground and polished. On the other hand the two last-named types always gave the more consistent results. Three types of fracture, straight, V, and 1/2V shape, were found, dependent on width and on edge condition. Straight fractures were regularly obtained with specimens less than 0·9 cm wide, with cut edges or ground and polished edges. Specimens with fire-finished edges always gave a large proportion of V and 1/2V fractures whatever the width tested.

A. J. Holland & W. E. S. Turner

JSGT 1928 V12 T161-T164

In a previous paper on the casing of one glass by another, dealing with the special example of casing a colourless by a cobalt blue glass, we have laid stress on the fact that for successful welding or casing the thermal expansions of the glasses are not the only factors of importance to be considered.

[Link to JSGT_1928_T057-T074.pdf]

Francis Winks & W. E. S. Turner

JSGT 1928 V12 T057-T074

In 1924, an account was given by one of us at a meeting of the Society at Stourbridge of the properties of glasses containing lead oxide. At that same meeting, Mr John Northwood described the work involved in the reproduction of the Portland vase; and during the discussion which followed both papers the subject of the relationship between the thermal expansion of two glasses, one welded to or cased on the other, was referred to. It appeared to be a fairly general belief that the essential requirement for two glasses which are to be welded is that their coefficients of thermal expansion shall be as nearly as possible the same. Manufacturers of cased glasses have found from experience, however, that a coloured glass which has been coloured by the addition to the colourless base glass of a small amount of colouring matter cannot always be successfully welded to the base glass itself, and it is the usual practice, in order to obtain successful welding, to employ as coloured glass one which is relatively softer than the base glass. Hence the question has been raised by several practical workers as to whether or not the colouring agent brings about some marked modification of the thermal expansion as compared with the colourless glass.

W. E. S. Turner & Francis Winks

JSGT 1929 V13 T219-T229

On the basis of the investigation of a large number of samples of 70 different kinds of glasses subjected to heat treatment with the object of inducing crystallisation, the following conclusions may be made: (1) The presence of alumina (Al₂O₃) in the glass does not always tend to weaken the crystallising power. (2) An increased content of the alkalis in the glass, other conditions being equal, reduces the crystallising power. (3) Glasses containing a great number of components have a lower crystallising capability than the simpler glass. (4) The crystallising capacity of glasses decreases with the reduction of CaO content, and the content of CaO must not exceed 10%. (5) Glasses containing more than 11% of CaO, with an alkali content not exceeding 14% have a high crystallising capacity. (6) Glasses not containing any Al₂O₃ and having an equal percentage of CaO have the greatest crystallising capacity when the alkali content is lowest. (7) When MgO in the form of magnesite is substituted for CaO in standard glasses, it increases the crystallising capacity.

J. J. Kitaigorodsky

JSGT 1937 V21 T061-T081

Isothermal creep tests of stabilised fibres of a soda–lime–silica glass have been made in the range 475-525°. Elongation is due to viscous flow at constant rate plus completely reversible elastic distortion, instantaneous and delayed. For the latter, dl/dt=kl, l being the unstretched part at time t. Furthermore, log k=A' /T_abs+constant, where A' =8·05×10⁴. This relation shows no break at the “transformation point,” 515° (from thermal expansion), but above 515° the rates are rapid (large values of k)and the glass acts like an ordinary liquid. Young's modulus, based on the total elastic distortion, appears independent of temperature. The slope of the logη versus 1/T_abscurve for viscous flow is 7·04×10⁴. An ionic picture of the elastic process is given.

Nelson W. Taylor, Edward P. McNamara & Jack Sherman

JSGT 1937 V21 T450-T452

There is no contradiction between Stott's results (on an unstabilised glass) and those of Taylor, McNamara & Sherman (on a stabilised glass). Stott's data on the relaxation from torsion of a glass fibre may be represented by an equation of the Adams & Williamson type, namely, 1/l−1/l₀=k₂t, or −dl/dt=k₂l², plus an additional term which corresponds to −dl/dt=k₁l. A molecular theory of these processes is briefly outlined. Details are presented in an article by the writer entitled “The Law of Annealing of Glass-Quantitative Treatment and Molecular Interpretation,” published in the Journal of the American Ceramic Society.

 Nelson W. Taylor

JSGT 1937 V21 T356-T358

It is rather surprising that so little work has been done on the retarded elastic movements in glass, and any new contribution to knowledge in this respect is to be welcomed. Some of the results published in the above paper, however, appear to conflict with earlier work, and it seems desirable to indicate the differences.

 V. H. Stott

JSGT 1952 V36 T005-T055

The electrical properties of glasses offer, at first sight, a field of theoretical and experimental investigation capable of yielding important information about the nature and structure of glass. In the theoretical treatment, however, it is necessary to make simplifying assumptions leading to expressions which are only first approximations, and the experimental work presents difficulties which tend to limit seriously the accuracy of the experimental observations. Nevertheless, the measure of agreement between experimental results and the approximate expressions derived from the simplified theoretical treatment indicates that results of qualitative value can be obtained from a study of the electrical properties of glasses, though quantitative deductions drawn from such study would be unreliable. In the present paper a theoretical survey is made of the electrical conductivity, dielectric constant and power factors of glasses in relation to composition and structure. This is followed by a description of experimental observations of these and related properties on five series of glasses, and the experimental results are discussed in relation to the theoretical treatment. The discussion shows clearly that if the full measure of information is to be drawn from this important field of investigation there is need for a more exact theoretical treatment and also for experimental observations of greater accuracy than has hitherto been obtained.

H. Moore & R. C. De Silva

JSGT_1928_V12_T169_T190

The presence of boric oxide in glasses is accompanied by very definite characteristics, which are indicated by the fact that the oxide has been employed widely in modern chemical and heat-resisting glasses because of the reduction which it brings about both in the corrosion suffered when such glasses are treated with water and in their thermal expansion. In a series of papers from this Department, however, it has been shown that this beneficial use of boric oxide has very strict limitations. In some series, indeed, notably when silica is progressively displaced in Kavalier's glass by boric oxide, there is no appreciable improvement in chemical durability; indeed, beyond a certain basic oxide content, the glass is rapidly broken down by water and chemical reagents; whilst there is no diminution in thermal expansion but rather (again beyond a certain limit) an increase in value. The simpler series of soda–boric oxide–silica glasses, obtained by the progressive substitution of silica by boric oxide, are associated with maxima or minima in the curves which correlate the composition and a particular physical property.

[Related paper JSGT_1925_T389-T405.pdf]

Arnold Cousen & W. E. S. Turner

JSGT 1957 V41 T276-T282

In the Owens-Corning glass-wool process, glass fibres are produced by subjecting streams of molten glass to a blast of steam. It was clear that the rate of linear output of fibre from each stream was many times the velocity of the steam, but the mechanism of fiberisation had not been definitely established. The process has been studied by flash photography using a flash of 2 µsduration, and by cine photography at 3000 frames/s. Examples of the photographs and film sequences obtained are presented. It is concluded that the mechanism is neither the formation and attenuation of drops of glass nor the division of the stream of glass by branching, but is a looping and folding of the glass stream induced by turbulent flow of the steam blast. Subsequently, threads of glass in the form of inverted U shapes are detached, and many of these are drawn simultaneously into fibre.

A. de Dani & P. E. Jellyman

JSGT 1923 V07 T057-T073

Despite the importance which boric oxide has assumed in the glass industry, the volume of published researches concerning glasses containing it is not large. Undoubtedly the research workers attached to the large factories in Europe and America where borosilicate glasses have been made must have acquired data concerning its properties, but for the sake of general progress it seemed to us that some systematic investigations should be carried out and the results made available for the industry in general. Hence the present series of papers.

V. Dimbleby, F. W. Hodkin, M. Parkin & W. E. S. Turner

JSGT_1939_V23_T253_T264

Meltings have been made, mainly at 1400°, in platinum, of a glass of basic composition SiO₂ 75.0, CaO 9.2, Na₂O 15.8%, to which either 0.1 or 0.2% of iron oxide as Fe₂O₃ in the glass was added and the batch melted with the addition of one of the chlorides, ferric chloride, sodium, potassium or ammonium chloride, in amount equivalent, in terms of Cl, to 0.1, 0.2 or 0.4% of the glass. Analysis of the resultant glasses showed that during the melting process iron as well as chlorine were lost by volatilisation, probably as ferric chloride, the losses being very variable, but being, in general, lowest with sodium chloride and greatest with ammonium chloride. In no case did the loss of iron oxide reach 25% of the original amount, and, with sodium chloride, in no case did it reach 10%. Although ammonium chloride was the most effective of the reagents in eliminating iron oxide, its use in practice would be of doubtful benefit owing to its tendency to produce colouring effects varying from yellow to brown.

H. Marshall Bateson, B.Sc.Tech., and Professor W. E. S. Turner

JSGT 1953 V37 T269-T301

Additional work on reactions between silica and one or more metallic oxides was carried out with a view to investigating how the reaction rates are modified by: (a) introduction of lime as CaO and as CaC03, (b) the presence of alumina or zirconia, (c) the particle size of the silica; and (d) the forms in which the silica is introduced. Reaction rates between alumina, mullite or zirconia and one or more of the other metallic oxides were also investigated by the x-ray-diffraction method. The results obtained showed that: (1) The reaction rates in silica–lime–soda mixtures are independent of whether the lime is introduced as calcium oxide or carbonate. (2) The addition of alumina or zirconia to a silica–soda mixture decreases the reaction rates in the initial stages of firing, but after a certain period of firing at 1400° the reaction rates are higher. (3) The reaction rates in a silica–soda mixture increase with the diminishing size of the silica particles. (4) At 1200° and' 1300°, cristobalite reacts with soda at a higher rate than quartz, but at 1400° and 1450° the reaction rates are almost identical. (5) Zirconia reacts with the other metallic oxides more slowly than mullite or silica, and also reacts more slowly than alumina, except with the alkalis or boric oxide. (6) The various metallic oxides react with mullite at a much slower rate than with silica. (7) Each of the metallic oxides, when present alone, reacts with alumina at a much slower rate than with mullite or silica; the presence of silica gel or boric oxide considerably increases the rate of reaction of the metallic oxides with alumina, but has only a small effect on the rates of reaction with mullite or silica. (8) The various alkali-metal oxides and alkaline earths, arranged in decreasing order of their reaction rates with silica, alumina, mullite or zirconia, can be placed as follows: K₂O, Na₂O, Li₂O and BaO, CaO, ZnO and MgO respectively, when they are present in molecularly equivalent quantities.

Abd-El-Moneim Abou-El-Azm

JSGT_1934_V18_T032_T066

Nearly fifty glasses, belonging to the three series Na₂O–B₂O₃; Na₂SiO₃–B₂O₃; and NaBO₂–SiO₂ have been melted in the pure condition their general characteristics determined, and the annealing temperatures, densities, and thermal expansions systematically measured. In the majority of cases, maxima or minima occur on the physical property–composition curves. The general results have been correlated with other known data and attempts made, inconclusively, to establish a connection between these maxima or minima and the chemical constitution of the glass.

E. J. Gooding & W. E. S. Turner

JSGT_1933_V17_T273_T286

The paper gives the results of a mathematical investigation into the stress distribution present in a system consisting of two or more glasses fused together. The stress distribution is first worked out for the case of a doublet system where the coefficients of expansion of the two glasses are slightly dissimilar. The two cases are distinguished, first, where the strip is free to bend, and secondly, where it is rigidly held at the ends, as this approximates to the conditions for the manufacture of such articles as globes. The effect of varying the thickness of the portions of the sheet is discussed, and curves arc drawn for the stress distribution, showing close agreement with actual photographs taken by means of a Babinet compensator. By a modification of the equations a method is indicated whereby an estimate of the difference in coefficient of expansion of two glasses can be obtained from a measurement of the curvature of two rods which are drawn together. In the second portion of the paper, the distribution is worked out for the case of three glasses fused together; and subsequently the practical case where the two outer glasses are of one composition, and the inner glass of another, is worked out in detail. Again, typical instances are computed and the results compared with actual photographs.

W. M. Hampton

JSGT_1933_V17_T122_T144

The volatilisation of K₂O from potash–silica glasses of composition ranging from 89·88% SiO₂ 10·12% K₂O to 50·68% SiO₂ 49·32% K₂O has been studied at temperatures ranging from 1100° to 1400° and in five cases to 1450°. Within the temperature range studied the silica is non-volatile. At 1400° the loss of K₂O during the first 20 hours varied from 5·88 mg per hour per sq. cm. of surface for the glass containing 49·32% K₂O to 0·27 mg for that containing 19·61% K₂O; at 1100°, for the 49·32% K₂O glass the loss was 0·56 mg. The rate of volatilisation could be represented, as is the case with the Na₂O–SiO₂ glasses, by the formula log(A−2x)/A=kt, where A is the amount of K₂O initially present and x the loss after t hours; whilst a linear relationship holds good for each glass between log (initial rate of loss) and 1/T° (Abs.). The vapour tension of the 49·32% K₂O glass ranged from 3·10 mm Hg at 1300° to 0·90 mm at 1100°; for the 43·46% K₂O glass from 1·82 mm to 0·50 mm; and for these two glasses the heat of volatilisation of K₂O was calculated as −37,000 cal. per gm.-mol. On plotting total volatilisation loss at different times against K₂O content in the glass, there was evidence of a break at the composition corresponding to the compound K₂O.4SiO₂, although less sharp than for Na₂O.2SiO₂ in the soda–silica glasses. Several other relationships however, gave clear evidence of the existence in the molten glasses, of the compound K₂O.4SiO₂ and indicated the possible presence also of the compound K₂O.2SiO₂ at the appropriate concentration.

Eric Preston & W. E. S. Turner

JSGT_1935_V19_T296_T311

The mixtures ranged in composition from 53·96 to 99·80% PbO (24 to practically 100 mol%) and by rapid cooling could be obtained as glasses up to 90·05% PbO. The initial rates of volatilisation in mg/cm²hr ranged from 0·10 to 11·6 at 900°, 0·20to 80·0 at 1000°, 1·1 to 250at 1100°, and 2·0 to 254 at 1200°. The high volatility of the last two members of the series precluded any measurements beyond 1100° and the first two members were devitrified at 900°. The temperature–volatility relationships approximated to a vapour pressure law. The relationship between loss by volatilisation and concentration indicated the existence of the silicates PbO.SiO₂, 2PbO.SiO₂, 4PbO.SiO₂, and further relationships involving the progress of volatilisation with time confirmed this view. The combined heats of decomposition and evaporation were, in the order given, −36,000 −49,000 and −30,000 cals. per gram molecule. respectively. Density effects in the melts marred the ideal conditions for the volatilisation, but the results obtained were a general confirmation of earlier work on alkali–silica glasses.  At 1200° the calculated vapour pressures of PbO ranged from 0·7 to 72 mm. Hg for the first twelve glasses. At 1100° the vapour pressure of pure PbO was calculated to be 70 mm. Hg.

Eric Preston & W. E. S. Turner

JSGT_1936_V20_T127_T138

Four soda–lime–silica glasses containing 16% Na₂O and 0·31-0·83% SO₃ were prepared and samples heated at temperatures from 1300° to 1450° in an oxidising atmosphere. The order of volatility was the order of the sulphate contents, and the loss of sulphate was approximately proportional to the initial sulphate concentration. After 100 hours heating at 1350°, 1400° and 1450° the sulphate contents were reduced respectively to about 2/3, ½ and 1/3 of their initial values. The variation of total volatilisation loss with temperature obeyed a vapour pressure law. Comparison of the total volatilisation losses with the sulphate losses determined analytically showed that the presence of SO₃ caused an increase in the loss of alkali, approximately to the extent of twice the alkali volatilisation experienced by a pure glass of the same composition containing no sulphate.

Eric Preston, W. E. S. Turner & Horace Laithwaite

JSGT 1957 V41 T259-T275

The pattern of currents in glass contained in tank furnaces has been studied using transparent 1:20 scale models. Endeavours were made to maintain similarity, using the Gallileo Number as a criterion, glass being represented by a liquid phenolic polymer. A concept of a 'refining factor', which combines viscosity and time, has been introduced to judge the effects of operation and design factors on the pattern of currents and their velocity. Examples are given of the effects of the thermal barrier position and of the location of the throat.

S. Kruszewski

JSGT_1920_V04_T107_T111 (7 pages including 2 of plates)

The object aimed at in the author's device, is to enable the interior surface of the pot to be worked upon when in a hard and toughened condition. It is known that the life of the pot is improved if this can be done, and it is the ordinary practice on the Continent with pots which are made in moulds to beat the inside surface after it has become hard with a beater in order to produce a dense and tough skin.

C. W. Thomas

JSGT 1943 V27 T053-T059

This account of the examination after failure of open pots made of a sillimanite-ball clay mixture is given to encourage the wider adoption of photographic and petrological methods in the examination and recording of the performance and failure of glass-melting pots. It is illustrated by reference to studies of the performance of pots employed in the founding of a hard borosilicate glass at a temperature of 1470–1480°. Long pot life is not expected when melting this type of glass, but failure seemed to be due in the cases examined to shot-holing of the pot wall, pointing to local inhomogeneity, possibly due to large iron spots, or to inadequate pugging and mixing, or to slight differences in texture. A feature of the defective parts was the pronounced mullite development, and appeared due to the effect of B₂O₃ vapour on alumina-silica refractories; for it was noted also on gathering hoods at 1350° and on the crown of an electric furnace melting a very high B₂O₃ content glass at 1250°.

Eric Preston

JSGT_1935_V19_T189_T207

The paper records the results of measurements of the surface temperature at many points in the structure throughout the operating period of a glass tank furnace. An attempt has been made to correlate the progressive changes in surface temperature with changes in tank block thickness due to corrosion, and with other variable conditions internal and external to the furnace.

H. S. Y. Gill & N. A. Nichols

JSGT_1927_V11_T415_T424

As a step in the direction of improving the quality of glasshouse pots, it was suggested to the author by a manufacturer about four years ago that a review of pot-making conditions in this country might serve a useful purpose. The suggestion was accepted as a very useful one and statistics have gradually been collected. It seemed opportune that a meeting at Stourbridge should be made the occasion to set out the information obtained.

W. E. S. Turner

JSGT 1956 V40 T319-T337

When an article made of glass, or other transparent material, is viewed in air, the observed image of its inside shape is not true, because refraction of the light as it leaves the outer glass surface causes distortion of the images of shapes inside the article. This refraction, and the resulting distortion, can be prevented by immersing the article in a liquid which has the same refractive index as the glass, thus enabling accurate measurements of the cavity dimensions and of wall thickness to be made by projecting the image on to a screen. Liquids are described which are suitable for workshop use. When a single liquid does not match a particular glass, mixtures of liquids can be employed. Simple tests can be used when mixing, to ensure that a match is obtained. Small differences of refractive index between the glass and the liquid are always present, and will cause distortion. It is shown, however, that the accuracy obtained is satisfactory for the purposes required in glass container manufacture, and simple formulae are given to determine the errors and to enable corrections to be applied when necessary. These formulae also establish the sizes of glassware which may be examined without introducing errors exceeding specified limits. Other applications of the immersion technique are briefly described.

W. R. B. Frank & L. Green

JSGT 1944 V28 T053-T091

The account of the test which follows is divided into two parts. In Part I, two of the authors describe the experimental work, the data resulting from it and their own deductions. In Part II the third author subjects the data to detailed statistical examination and presents his conclusions.

L. Moxon, F. Winks & B. P. Dudding

JSGT 1956 V40 T509-T512

 
A method has been developed for the determination of silica in glasses, sands, and refractories, by precipitation of potassium silico-fluoride followed by alkalimetric titration of the precipitate. Starting with the ground and dried sample, the total time for carrying out duplicate determinations is less than three hours. Apart from its general use for speeding up silica determinations, it is completely applicable to fluorine containing glasses.

J. M. R. Maxwell & S. M. Budd

JSGT_1934_V18_T307

This early Norman lancet window in Kirdford Church, Sussex, has been glazed with fragments of glass dug up on the sites of four mediaeval glass furnaces in Kirdford, and representing the industry from the fourteenth to the sixteenth century.

S. E. Winbolt

JSGT 1952 V36 T181-T195

The need for conserving selenium supplies and yet for continuing the production of decolourised glass is stressed. The iron and chromium content of the raw materials, and that resulting from transport, mixing, conveying and melting, are compared with the iron content of the finished glass. The effects on the quantity of selenium necessary for decolourising of additions of arsenic, sodium nitrate, sulphates and cobalt, and of the melting and working properties are discussed in detail. The economy in the use of selenium achieved in the full scale production from 11 tank furnaces was considerable. In tests on two furnaces it was found that selenites could replace the equivalent amount of selenium without affecting the colour of the glass.

H. W. Howes

JSGT_1917_V01_T144_T147

The action of caustic soda on different types of glassware is described. No earlier data are available for a comparison of the effect of 2N-caustic soda with that found by us, but it is worthy of note that with 2N-caustic soda solution the extent of its action on glass, according to our tests, is greater than in other previously published reports. The data for comparison are few and confined to Jena glass. For 2N caustic soda, acting at 100° for three hours, the losses found at the National Physical Laboratory on six different new Jena flasks.varied from 60 to 79 mg per sq. dcm. In our own determinations, the loss (first period) was observed to be 112 mg. It is possible that some of the discrepancy may be accounted for by a slight difference in method. Although in the report cited, no details are given, the practice frequently followed is to heat up the caustic alkali solution to 100° in a silver flask and to transfer it then to the glass vessel under test. In our own experiments, it was essential to avoid this transference, and the flask and alkali were heated as quickly as possible on a hot plate to 100° prior to being placed in the thermostat. This means an additional ten minutes' contact between alkali and glass; but seeing that the period in the thermostat is three hours and the corrosive action of alkali below 80° is slight, the difference of procedure cannot account for much of the margin between the figures quoted and our own.

G. W. Westcott Way, J. D. Cauwood & W. E. S. Turner

JSGT_1929_V13_T270_T279

Shortly after the publication by us of the previous paper* on this subject it was observed by one of us that the originally quoted method t for the removal of P₂O₅ by means of silver carbonate and silver nitrate was not satisfactory in the case of a complex glass containing P₂O₅ in amount, say, of 5% or more. Accordingly, the method previously given has been more fully investigated in the laboratories at Sheffield and Canning Town in order to test the conditions under which it breaks down and to work out a procedure of more general application.

J. D. Cauwood, J. H. Davidson & Violet Dimbleby

JSGT 1946 V30 T318-T326

Values have been computed for the energies of formation from their component ions in the gaseous state, of a large number of oxygen-containing inorganic crystals and glasses. These energies of formation can be computed approximately by the addition of energy constants characteristic of the component positive "ions" (elements other than oxygen). These energy constants depend primarily on the ionic charge, but also on the inter-atomic distance between positive and negative ions, on the co-ordination numbers, and other factors.

Minor deviations from constancy of the energy constants for phosphorus and silicon can be related to variations in the phosphorus–oxygen or silicon–oxygen ratio and so to the proportion of oxygen atoms in the PO₄ or SiO₄ groups which are shared with other phosphorus or silicon atoms. SiO₄ or PO₄ groups differing by more than one in the number of shared oxygen atoms tend to react together and so probably exist together only rarely, if at all, in the same substance.

Maurice L. Huggins & Kuan-Han Sun

JSGT_1924_V08_T150_T162

The author has been asked to give a short account of various alloys specially resistant to heat or corrosion and likely to be of interest to the glass industry. He is not very familiar with the details of glass manufacture, but believes that there are ample opportunities for the use of alloys which resist oxidation at high temperatures, corrosion under atmospheric or other conditions, or which, at fairly high temperatures, retain the greater part of the strength they possess at normal atmospheric temperatures.

J. H. G. Monypenny

JSGT_1924_V08_T277_T283

The effect of the presence of sillimanite (Al₂O₃,SiO₂) in fired glasshouse pots and tank-blocks in increasing their resistance to corrosion has been frequently commented on. It was thought of interest to examine some used glasshouse pots and used and unused tank-blocks made from British clays.

W. J. Rees

JSGT 1948 V32 T113-T121

The author records his impressions of a visit in 1946 to observe American methods used in the manufacture of domestic glassware, more especially of the hand-made types.

W. Maskill

JSGT_1918_V02_T003_T004

The idea of applying to the blowpipe flame the principle of preheating the fuel mixture is due to Mr George Keith, of the James Keith & Blackman Co., who made the first burner of this type in the pre-war days of 1914. The burner was resurrected and hastily reconstructed in 1916 as a twin burner with two opposed inclined jets and put into regular commission to meet large and urgent Army demands for “anti-gas”  capsules of crushable glass containing alcoholic ammonia. The burners were designed, made, and at work in a few days, with consequent doubling of output of the firm (John Tye and Son, 457, Caledonian Road, London, N.), with which the author is associated. The burners now described are the result of eighteen months' collaboration by Mr Keith and the author. The flame temperature of the small single-jet burner described below is, in the author's opinion, the highest obtainable by burning ordinary coal gas with air.

W. A. Whatmough

 JSGT_1924_V08_T038_T045

The method of calibration used was that of transferring successive quantities of water from the blank burette tube to a standard pipette, a method suggested by Ostwald for testing burettes. The apparatus is similar to an arrangement which I saw recently in the graduating department of Messrs. A. Gallenkamp & Co., Ltd. Any equally convenient method of connecting the burette tube and pipette may be used.

Verney Stott

JSGT_1935_V19_T151_T155

An apparatus for determining the resistance of refractory materials to the action of slags at high temperatures is described. It consists of a circular furnace built up of the materials under test, the furnace is heated by a rotating blowpipe, fed with gas and air under pressure, the slag is injected through the blowpipe flame

 A. Eric J. Vickers & R. A. Bell

JSGT_1919_V03_T034_T037 (6 pages including 1 page of plates)

The principle of the method described recently for the calibration of pipettes can very easily, accurately, and fairly quickly be applied to the calibration of tubes for burettes. Before describing the method. it may be an advantage to restate two conditions which should be observed in the calibration of volumetric apparatus and of burettes in particular.

S. English

JSGT_1937_V21_T281_T284

The detection of defects within the body and on the surface of glassware, by a contrasting field method, is described. Photographic examples of the visible effects so produced by cords, stones, seed, and surface unevenness are shown.

E. Seddon

JSGT 1948 V32 T340-T365

From an argument based on the randomness of the ionic arrangement in glass, and from kinetic considerations, the conclusion was reached that configurational changes are bound to take place in glass as the temperature is changed. The distribution function associated with the randomness is assumed to be approximated by a two-cell histogram, and on this assumption approximate formulae for the thermal, viscous and elastic properties of silica-rich glasses are derived. The formulae were applied to experimental results on Pyrex brand glass, and it is shown that there is a quantitative inter-relation between the variations of thermal expansion, electrical resistivity and viscosity with temperature.. Notes on the rate of growth of cristobalite and on the release of stress in the same glass are included.

S. M. Cox

JSGT_1923_V07_T020_T024

IN reading carefully through the recent literature dealing with the allied questions of the detection of strain in glassware, and of the annealing of strained glass, a number of statements have been met with which need either qualification or correction. In most of these cases no figures or evidence are given to support the statements, and one is compelled to conclude that in some cases generalisations have been drawn from insufficient data. In order to set on a more sure basis the methods for the detection of strain and its removal
from glass, some of these statements need examination.

S English

JSGT_1950_V34_T274_T304

A study has been made of the corrosion resistance of refractories of high purity when exposed to molten glass of ordinary colourless bottle glass composition. The refractory mixtures were made mostly from kaolin, alumina and silica in various proportions; but in one or two cases burnt kaolin (Molochite) was used in place of some of the raw kaolin, and in one series sillimanite was used as the “grog.” The compositions were varied to give alumina/silica ratios ranging from 30:70 to 80:20. For most of the experimental refractories the grog was prepared by firing a suitable mixture of the raw materials, the grog being subsequently bonded by a similar mixture. The proportions of grog and bonding mixture were varied, the proportion of kaolin in some of the mixtures being 40%, and in the others 35%. The shrinkages during firing and the apparent porosities of all the fired refractories were determined; the effects on porosity produced by firing to different temperatures and by holding the mixtures at particular temperatures for longer or shorter times were also studied. The corrosion resistances were compared by determining the extent to which cylindrical test-specimens were corroded after partial immersion in molten glass at 1400° and 1500° for periods of 10 h, the specimens being moved slowly through the glass. To determine the effect of grog size on corrosion resistance, tests were made using mixtures containing different grog-gradings but otherwise identical. The results indicate that corrosion resistance cannot be correlated in any direct and generally applicable way with overall chemical composition, or apparent porosity, or, again, permeability to molten glass. With certain provisos, however, ·the conclusions can be drawn that corrosion resistance is enhanced by the use of fine-grained grog and of lower proportions of raw clay to grog, also by firing the grog and the refractory mixture at a sufficiently high temperature for a sufficient time. The difficulty of formulating any useful specification other than a “performance specification” is pointed out, and it is suggested that the corrosion test used in the investigation should be given full consideration in the event of such a specification being put forward.

H. Moore & R. Heeley

JSGT_1933_V17_T320_T323

Our improved method consists in standing the test-piece on three steel balls on a steel plane table. A steel ball carried at the top of the test-piece bears against the plunger of an Ames dial gauge, and the gauge is standardised against steel rods of known lengths. An accuracy of one unit in the second decimal place is claimed.

A. Eric J. Vickers & J. A. Sugden

JSGT 1941 V25 T164-T220

Determinations of the transmission the visual spectral have been made at temperatures of 20°, 100°, 200°, 300°, 400°, or near, the approximate temperature in the ease of 27 glasses, save that for a copper glass the upper was restricted because of permanent colour and transmission change occurring on heating at higher temperatures. For one heat-resisting glass the measurements were carried to 800°. The determinations were made for rising and falling temperatures and the original 20° measurements finally re-checked to ensure absence of permanent change on heating. Of the 27 glasses used in the form of optically polished discs 19mm in diameter, a soda-lime-silica and a soda-boric oxide-silica glass were melted in platinum from pure materials and served as a basis of reference, the 25 others being commercial glasses covering a wide range, including chemical and heat-resisting, plate and decolourised bottle, lead-containing glasses for domestic coloured glasses for bottles, for heat-protecting purposes colour filters. The complete composition of glasses is set out, including all the minor constituents likely to exercise selective absorption. Two standards of thickness, 2mm or 10mm, were employed, according as the glass was colourless (or nearly so) or coloured. The range of wavelength was from 4400 to 7200A, measurements being carried out at each intermediate 100A interval. Thorough tests showed that the whole of the light incident on the glass specimen could be accounted for in the experiments, and, in particular, it was demonstrated that no change in the loss by reflection with rise or fall of temperature took place other than that dependent on the variation with temperature of the refractive index of the glass. As a general rule, reduction in the transmission and a shift of the maximum transmission towards the long wavelength end of the occurred progressively with rise of temperature. This was the case with colourless and only faintly coloured glass, although the effect of temperature rise affected the transmission of such glasses differently at different wavelengths, and particularly for the extreme ends of the visual spectrum. The least effect was observed for the colourless standard glasses free from any but traces of selective absorbing agents (Fe₂O₃ and Pt), the soda-lime-silica glass having its overall transmission for illuminant C reduced from 91 to 86 per cent between 20° and 600°, the corresponding values for a soda-boric oxide-silica glass being 90 and 88 per cent, respectively. In the case of glasses coloured by Fe₂O₃, FeO, CoO, NiO, Cr₂O₃, Mn₂O₃, “Di₂O₃,” S, Se, Cu and Au, the following observations are especially noteworthy:

(1) The dominant effects of certain oxides, particularly Cr₂O₃ and CoO even in extremely small amounts, whether in coloured or faintly coloured glasses

(2) the sharper "cut-off" in the violet end produced by relatively small amounts of sulphur (with carbon and low iron oxide) than by considerable amounts (1-2%) each of Fe₂O₃ and MnO

(3) The presence of points of intersection in the temperature-transmission curves for glasses containing Cr₂O₃, Fe₂O₃-MnO, and CoO, whereby in certain wavelength regions transmission fell with rise of temperature and showed inverse behaviour in others

(4) the presence of points of intersection in quite a number of glasses in the graphs obtained by plotting temperature against wavelength for a specific transmission, these points to the transformation temperature (T_g) and others to additional points of the glass. A potash-lead oxide (31%)-silica glass alone of the colourless glasses showed this phenomenon. For the coloured and three colourless glasses the effect of temperature on the colour was calculated by the method of A. C. Hardy for the C.I.E. coefficients, x, y, and z, and the results tabulated and also set out graphically.

A. J. Holland, M.Sc., Ph.D., A.R.C.S., D.I.C., & Professor W. E. S. Turner

An investigation of ancient opal glasses with special reference to the Portland

JSGT 1959 V43 T285-T288

The nature of the opacifier in a fragment of opal glass from the Portland Vase and in other specimens of cameo glass of the same historical period has been investigated. Chemical analysis shows the base glass to be a soda–lime–silica type containing more than 8% of antimony oxide and x-ray diffraction proves that a precipitate consisting of the calcium pyroantimonate Ca₂Sb₂O₇ is responsible for opacification. The crystal structure of Ca₂Sb₂O₇, although of relatively low symmetry, has affinities with that of the cubic pyrochlore class of compound.

H. P. Rooksby

An Investigation of Cracks and Stress Waves in Glass and Plastics by High-speed Photography

JSGT_1952_V36_T074_T089

A high-speed spark camera is described which will take a series of nine photographs at rates of 20,000 to 200,000 frames per second. The camera has been used to photograph the propagation of stress waves and the formation of cracks in sheets of glass and “Perspex.” The illustrated results show the effects of applying pressure pulses of from 2 to 40 µs duration to small plates of the materials. The reflection of the stress waves from the free edges of the plate is discussed with particular reference to the formation of fractures. Finally, the type of hackle structure appearing on some of the fracture surfaces is discussed in relation to the method of formation of the cracks.

D. G. Christie

JSGT_1935_V19_T043_T103

Following a critical survey of the literature on selenium decolorising, an investigation of the decolorising of a soda–lime–silica glass by selenium has been made. The colours have been expressed quantitatively on the Lovibond scale. Standardised melting and cooling conditions were worked out and a series of melts were obtained from which the effects of the addition of minor constituents on selenium colours were recorded. A number of the glasses were subjected to reheating and sunlight treatment. As a result of these experiments a theory of the reactions occurring is proposed. The theory is as follows: the iron in glass is present as ferric oxide which decomposes with increasing temperature to ferrosoferric oxide with the loss of oxygen. The reaction is irreversible unless oxidising agents remain in the glass. Selenium reacts with the ferrosoferric oxide forming a ferrous selenide. Arsenic combines with ferrosoferric oxide forming ferrous arsenate, which is ionised by the action of sunlight. Sunlight has no effect on selenium or selenium compounds in glass. At high temperatures arsenic also probably combines with selenium to form a complex which decomposes at about 700°.

E. J. Gooding & J. B. Murgatroyd

JSGT 1958 V42 T020-T031

Sixty gram melts of glass of composition (weight per cent) 73·5 SiO₂, 10·0 CaO, 16·5 Na₂O, were founded for varying times at 1400°C using four different sand grain sizes from 0·34 to 0·095 mm. Discs were cast from the melts and measurements of the total number of seed and their size distribution made on photographic enlargements of the annealed discs. The results showed that the total number of seed decreased exponentially with time. As the sand grain size decreased the seediness at the batch-free time increased, but the rate of refining also increased, and the most homogeneous and least seedy glass was produced most quickly by using the finest sand. The data enabled the rates of disappearance of particular sizes of seed to be derived and again the relation between number and time was exponential. The refining was therefore not controlled by the rise of bubbles through the melt. Relatively rapid convection currents were shown to exist in the melt but later work (not fully reported here) showed that the elimination of convection did not affect the refining. It was concluded that the exponential relation was not due to convection but to some other process that caused small seed to disappear much more rapidly than predicted from Stokes' law. The size of sand grains used affected the rate of disappearance of a particular size of bubble as well as the number and size distribution at the batch-free time. The results for the series using the finest sand showed an anomaly that could not be accounted for solely by the change in homogeneity of the melts.

Michael Cable

JSGT 1941 V25 T297 -T309

We have received and print below with pleasure the Twenty Fifth Annual Report of the Glass Delegacy.

W. E. S. Turner

JSGT 1941 V25 T310-T327

The temperature gradients set up within refractory blocks during their firing have been determined with reference to three blocks of one foot cube in size. One hand-made block was of fireclay containing 50 per cent grog, another hand-made block of sillimanite and the third slip-cast of sillimanite containing 75 per cent grog. Platinum-platinum rhodium thermocouples were inserted in the blocks, which were then individually subjected to a firing treatment comparable with commercial practice for some 200 to 300 hours in a gas-fired furnace having a total combustion space of 22 inches diameter by 31 inches depth. The time-temperature curves obtained showed that when heat is applied intermittently there is neither a smooth temperature rise, nor uniformity of temperature gradient between different parts of a block in the firing process; and these parts finally attain a temperature controlled by their location within the furnace, so that a natural temperature gradient becomes set up which prolonged soaking cannot remove. In the blocks for which data are given, the centre lagged behind the outer parts by about 6 to 8 hours in attaining a certain temperature during heating up; but at any one setting of the burners some 20 to 30 hours were necessary for stable temperature distribution to be reached, the natural gradients finally set up being 8° to 10° per inch from top to bottom of the block. The effect of these differences in thermal treatment of different parts of the block was ascertained by cutting sections from selected locations and determining their porosities and after-contractions.

Eric Preston, Ph.D., D.Sc., F.I.C., F.Inst.P., and Professor W. E. S. Turner

JSGT 1958 V42 T261-T270

The results are reported of an investigation into the effects of fluorine and alumina on the production efficiency of opal glass. As a result of the investigation, radical alterations were made to the founding procedure of the glass. Corrections to the batch were applied to allow for the considerable loss of fluorine from the glass by volatilization during founding and for the pick-up from the aluminous china clay tank blocks. Stable production efficiencies were obtained as control of alumina and fluorine was established.

R.W. Ellis & D. G. Osborne

JSGT_1929_V13_T019_T025

The equation referred to is the one devised by M. Le Chatelier to express the variation of viscosity of glass with temperature, published in Comptes rendus in 1924 (179, pp 517, 718) and in the Journal of the Society of Glass Technology in 1925 (8, pp. 12-24). [With a Reply by H. Le Chatelier from page 24.]

Linked to JSGT_1925_T012-T024

F. W. Preston

JSGT_1929_V13_290_T296

The constitution and structure of glass have long been a puzzling problem. The accepted opinion now is that glass is a supercooled liquid, a solution of various metal oxides in silica. Data on specific heat, surface tension, melting point, viscosity, and dielectric strength indicate that the material is of high molecular weight or highly associated. For a more thorough discussion of the current information on the· constitution of glass the reader is referred to a collection of papers entitled, “Constitution of Glass” from the Journal.

G. L. Clark & C. R. Amberg

JSGT_1939_V23_T338_T346

X-ray studies indicate that there are three varieties of mullite distinguishable by the differences in axial ratios of the unit cell. They are: α-mullite made from pure materials corresponding to the formula 3Al₂O₃,2SiO₂ and containing 28.2% SiO₂, 71.8 Al₂O₃; β-mullite containing excess alumina in solution; and γ-mullite containing small proportions of ferric oxide and/or titania in solution. All three varieties have been prepared artificially, but so far only the β- and γ- varieties have been found as minerals. β-mullite can contain as high as 78% of alumina, although there is a progressive change in the lattice from the α- to the β-variety as the alumina content rises above 72%. γ-mullite is obtained when comparatively small proportions of iron or titanium oxides are present; with larger proportions a general expansion of the lattice occurs. {3-mullite containing 78% of alumina corresponds to the formula 2Al₂O₃,SiO_2. It is clear that the replacement of silicon by aluminium is not limited to the change from sillimanite to α-mullite, but can be carried a stage further until the composition of β-mullite is reached.

H. P. Rooksby, B.Sc., F.Inst.P., and J. H. Partridge, Ph.D

JSGT_1922_V06_T045_T068

This paper is an account of some apparatus for the control of annealing and for the determination of annealing temperature.

F. Twyman

JSGT_1928_V12_T138_T140

The development of automatic glass-blowing machines using the suction method of gathering has made it imperative that the level of the molten glass in the melting and working sections of the tank should be maintained as nearly constant as practicable.

J. W. Ball & C. G. Eden

JSGT_1933_V17_T239_T259

Utilising data obtained in different laboratories from thermal endurance tests on glass rods of various diameters, some of the simpler statistical methods of analysis are illustrated. The principles underlying the determination of limits to be used in quality control charts are considered, and the adaptability of these charts to the scrutiny of routine test data emphasised. Reference is made to methods of testing the significance of two estimates of a correlation coefficient.

B. P. Dudding & I. M. Baker

JSGT_1934_V18_T019_T031

This paper deals with questions raised during the discussion of a former paper. The factors which should be considered in deciding on the “population,” the logging of test data, the number of specimens in the test sample and the determination of rejection limits are discussed in detail.

B. P. Dudding & I. M. Baker

JSGT 1957V41 T185-T219

In glass technology it is often difficult to achieve precision in small-scale experimental tests, because of the characteristic variability of certain attributes of the material, such as strength and dimensions. By employing statistical analysis to measure the experimental error the risk of drawing false conclusions can be covered. In addition to the experimental field, applied statistics can also be used in such things as the study of glass-working processes, and the engineering design of glass vacuum envelopes. Examples of applied statistics in glass technology include the following: (a) the study of dimensional variation, (b) the measurement of high values of electrical insulation resistance of glass–metal seals, (c) thermal shock tests, (d) photoelastic stress measurement, (e) the study of variability in glass engineering design, (f) the determination of risk of failure in glass engineering design, (g) the analysis of gauging machines. Statistical techniques which are illustrated include: (a) calculation of mean and variance, (b) prediction (of proportions within a range), (c) the logarithmic transformation, (d) significance tests, (e) analysis of variance, (f) sampling of attributes, (g) Monte Carlo experiments, (h) standard error of a function of variables, (i) methods of graphical analysis.

S. J. Morrison

JSGT 1957 V41 T185-T219

In glass technology it is often difficult to achieve precision in small-scale experimental tests, because of the characteristic variability of certain attributes of the material, such as strength and dimensions. By employing statistical analysis to measure the experimental error the risk of drawing false conclusions can be covered. In addition to the experimental field, applied statistics can also be used in such things as the study of glass-working processes, and the engineering design of glass vacuum envelopes. Examples of applied statistics in glass technology include the following: (a) the study of dimensional variation, (b) the measurement of high values of electrical insulation resistance of glass–metal seals, (c) thermal shock tests, (d) photoelastic stress measurement, (e) the study of variability in glass engineering design, (f) the determination of risk of failure in glass engineering design, (g) the analysis of gauging machines. Statistical techniques which are illustrated include: (a) calculation of mean and variance, (b) prediction (of proportions within a range), (c) the logarithmic transformation, (d) significance tests, (e) analysis of variance, (f) sampling of attributes, (g) Monte Carlo experiments, (h) standard error of a function of variables, (i) methods of graphical analysis.

S. J. Morrison

JSGT_1934_V18_T096_T105

It is necessary to study the technique of glass manufacture before attempting to design the form of any article. A shape obtained in a simple and straightforward manner is best. Moulded glass presents special difficulties, notably in obtaining equal distribution of the glass. Pressed glassware has a great future as a mass-produced article. During the eighteenth century British glassmakers developed a beautiful and individual design, based on principles of simplicity and utility. It was destroyed by the desire to imitate foreign patterns. Shapes and forms based on typically British style should be encouraged.

James H. Hogan

JSGT 1951 V35 T260-T269

A survey is made of the development of astronomy from observations made by Eratosthenes in 250 B.C. up to the discovery of the refracting telescope by Lippershey in 1609. The development of refracting telescopes is described, and reference is made to the invention of reflecting telescopes. The difficulty of figuring mirrors to an aspherical surface and its consequences on the development of the telescope is discussed. The factors leading to the eventual triumph of the reflecting telescope and the vital contribution which glass makes to the science of astronomy are described. Passing into modern times, the problem of the telescope mounting is briefly discussed, together with a more detailed description of the method by which the smooth sidereal motion is imparted to the polar axis of the Mount Stromlo 74-inch reflecting telescope. Finally, the latest development in telescopic optical systems due to Bernhard Schmidt is discussed and reference made to the proposals for the Isaac Newton telescope. This, the largest of its kind in the world, is to be installed at the new Greenwich Royal Observatory, Herstmonceux Castle.

G. M. Sissox

JSGT 1955 V39 T016-T036

In this paper earlier work on the low-temperature viscosity of simple glasses of widely-different chemical composition is reviewed, and the results are explained on the basis of the properties of ions and their interaction. It was found that a physical picture of the viscous flow of glasses can be obtained only if the interaction of the atoms to a considerable depth is taken into consideration. Among the current approaches to the nature of chemical binding forces, only the concepts of K. Fajans lend themselves to the development of a physical picture of such a "depth-action." This result can be generalised with respect to other mechanical properties of substances as determined by their chemical compositions. The response of matter to mechanical forces must change drastically if the volume under stress decreases below a certain critical value. Viscosity and strength should increase in the same fashion as the surface energy or the adhesion properties if the volume under test becomes too small for a co-operation of atoms over large volumes. This conclusion may be expressed in another form. If a condensed system is subjected to mechanical forces which tend to sever the chemical binding forces between two hypothetical planes of atoms, a large number of adjacent atoms co-operate with the mechanical forces by developing dipole moments, the forces of which aid the separation of the two hypothetical planes.

Evelyn C. Marboe & W. A. Weyl

JSGT_1917_V01_T153_T202

It was the necessities and the demands of the industrial chemist which gave rise to the British chemical ware industry. Laboratory glassware is one of the essentials in the production of munitions of war; for, without rigid chemical control, steel and explosives, as well as a host of other substances, cannot be effectively manufactured or their standard of quality maintained. The staffs of the large laboratories attached to the Sheffield steelworks speedily realised the situation, and, in co-operation with other chemists, made known their needs to the Board of Trade. The problem of substituting the German glass in bulk was not an easy one to solve, for not only had the chemical ware supplied by Germany attained a high standard of quality, but scarcely a glass manufacturer in this country was accustomed to working glass of the type required. The chemist was certainly in the position to

specify the properties which the glassware desired should posses, and, thanks to the published results of systematic investigations in Germany, the dependence of these properties on the composition was already tolerably well known.

J. D. Cauwood, S English & W. E. S. Turner

JSGT 195 V40 T388-T404

Factors governing the suitability of a glassmaking material are stated and illustrated by considering alternative materials for the introduction of alumina and of lithia into glass. Other batch materials, such as limestone, fluorspar, barium compounds, zinc oxide and borates are briefly considered. A table giving typical analyses, bulk densities, specific gravities and melting points of the common batch materials is given. Thermal conductivities of powdered batch materials have been determined and a description is given of the apparatus.

C. E. Ramsden & S. H. Ryder

JSGT 1933 V17 T324-T347

Measurements of the thermal expansion from room temperature to the softening point in each case have been made on seven glasses differing widely in composition and in viscosity and other properties. The glasses include fused boric oxide, three soda–silica glasses of soda content ranging between 19·2 and 39·29%, two soda–lime–silica glasses corresponding approximately to types used in commerce, one actually a commercial window glass, the other a pure synthetic glass, and finally an extra light flint optical glass of comparatively complex composition. Determinations of the linear thermal expansion were made when the specimens, of standard length and approximately the same diameter, were heated at the rate of 1° per minute throughout the range and also when heated at the rate of 1/6° per minute. Prior to the determination at 1° per minute, the specimen was annealed according to a special schedule described, involving cooling from the softening temperature at a rate of not more than 1° per minute. Prior to the determination at the very slow rate of heating, the annealing schedule involved cooling from the softening point at a rate not exceeding 1/6° per minute. The results show that with but special exceptions the extreme difference in linear thermal expansion coefficient of the glasses heated at the two rates did not exceed 4%. Of the seven glasses, with five the transformation point was reduced at the slow rate of heating by an amount which did not exceed 8°. Under similar conditions the softening point was also lowered, but in no case could the softening point be reduced to coincide with the transformation point. Transition points at temperatures below 500° were even more noticeable at the slow rate of heating than at the quicker rate. In several glasses also heat absorption effects caused retrogression in the curve just below the transformation point. Four successive slow beatings of a pure soda–lime–silica glass produced thermal expansion coefficient results differing by not more than 4% over the range 0–400°. The conclusion is drawn that whilst the determinations at very slow rates of heating are of special interest in studying the constitution of glass, there is no need, for practical purposes, for a specimen of 5 mm thickness to be heated at a rate slower than 1° per minute.

Eric Seddon & W. E. S. Turner

JSGT 1957 V41 T347-T349

It has been found that bismuth trioxide itself does not form a glass. However, glasses of high bismuth trioxide content can be made by melting the oxide with silica, boric oxide or phosphorus pentoxide. Glasses have also been made in the systems bismuth trioxide-vanadium pentoxide and bismuth trioxide-tellurium dioxide. It is pointed out that bismuth oxide behaves in a very similar manner to lead oxide in glass-forming systems.

M. S. R. Heynes & H. Rawson

JSGT 1919 V03 T037-T049 (13 pages)

Although there are various phases of the subject of bottle glass and bottle manufacture which all need discussion, the author considered that in the present paper he might usefully make his contribution under the two headings of quality and quantity, or durability and production. The author has found, in connection with the various types of automatic feed devices now coming into use, a great tendency to use very soft batches. The nature of the device, indeed, calls for a soft batch; but the engineer appears in many cases to have been working without collaboration with the chemist, and to have come perilously near, in some instances, to making a non-durable glass.

By W. E. S. Turner

JSGT 1936 V20 T586-T594

A study was made of the principal features of bricks of artificial mullite, produced in an electric furnace, reduced to grog and agglomerated by a ceramic process. The results are given of industrial experiments which emphasise the resistance of these bricks to the corrosion caused by vitrifiable substances blown over into the regenerator chambers.

Pierre Bergeron

JSGT 1918 V02 T035-T071 (39 pages including 2 of plates)

Potash-felspar has three chief uses in industry. In the manufacture of certain kinds of glass it is of considerable value as a source of both potassium and aluminium. Its use in this connection is dealt with more fully hereafter. The mineral is used in the pottery industries principally in the making of glaze, but also in the body of such ware as that used for electric insulators, where compactness and some desire of vitrification are required. In glazes the use of potash is preferred to that of soda, on account of the former being less liable to crystallise.

P. G. H. Boswell

JSGT 1932 V16 T022-T033

The paper demonstrates the importance of sample size and sampling procedure for tests to determine the probable performance in service of glass containers. Various tests, under conditions which gave positive breakages (5–10%) with satisfactory articles, are quoted and the results set out to show the effect. on the accuracy of the result, of sample size and of (described) sampling procedures. The results indicate that (1) appreciable errors are possible with too small samples, (2) for research and similar tests, the sample should consist of at least500 systematically selected articles, and (3) for routine tests, samples need not be so large as for research tests, but with “patchy” or, “border-line” quality articles larger samples are required than with regularly “good” or “bad” articles to give trustworthy indications of quality.

E. J. C. Bowmaker & J. D. Cauwood

JSGT 1932 V16 T022-T033

The paper demonstrates the importance of sample size and sampling procedure for tests to determine the probable performance in service of glass containers. Various tests, under conditions which gave positive breakages (5–10%) with satisfactory articles, are quoted and the results set out to show the effect. on the accuracy of the result, of sample size and of (described) sampling procedures. The results indicate that (1) appreciable errors are possible with too small samples, (2) for research and similar tests, the sample should consist of at least500 systematically selected articles, and (3) for routine tests, samples need not be so large as for research tests, but with “patchy” or, “border-line” quality articles larger samples are required than with regularly “good” or “bad” articles to give trustworthy indications of quality.

E. J. C. Bowmaker & J. D. Cauwood

JSGT_1936_V20_T680_T694

Formulae are suggested for the average specific heats. These formulae have been established for vitreous silica, lime and alumina in view of the fact that the largest number of measurements have been carried out on these substances by different authors. The values calculated in this way correspond very accurately to most of the values observed at the temperatures between −273° and 2600°.

Andre Thuret

JSGT 1936 V20 T319-T323

The first part describes briefly how the substances to be examined are best dissolved. The second part explains the calorimetric analysis of iron in accordance with the well-known method with potassium thiocyanate. The standard iron solution, the comparison solution, the comparison vessels, the tests and the limits of accuracy are specially discussed. A standard method is proposed.

L. Springer

JSGT V12 T075-T085

 When the glasses were prepared and drawn out into rod form it was quickly noticed that there was a difference in working properties between the colourless glass and those containing cobalt oxide. The difference was enhanced as the amount of cobalt oxide was increased, the general effect being a diminution of the working range, a quicker setting at the end of the drawing process, a more pronounced chilling of the parison when it was rolled on the marver, and a greater capacity for reheating (in the sense that the reheating took place more quickly) when the parison was warmed in the mouth of the furnace.

S. English & W. E. S. Turner

JSGT 1936 V20 T045-T053

 A proposed specification of the composition of refined pig iron suitable for glass bottle moulds was, total carbon 3·1% (maximum), sulphur 0·07 (maximum), phosphorus 0·45 (maximum), manganese 0·8–1·5, the silicon content varying according to the thickness of the casting. Two such pig irons of different silicon contents (e.g. 1·59 and 2·68%) covered, by appropriate mixtures, all casting thicknesses from 3/8” to 1 ¾”. Little alteration of composition occurred if the raw materials were melted in crucibles, but allowance was made for the increased carbon and decreased silicon contents brought about by cupola melting. The molten metal was partly desulphurised and refined in the ladle by treatment with sodium carbonate.

N. L. Evans, W. Goacher & J. E. Hurst

JSGT 1936 V20 T063-T071

N. L. Evans, W. Goacher & J. E. Hurst

JSGT 1931 V15 T219-T227

Within the last few years research among North Staffordshire parish registers has proved the existence of an early glass-house at Eccleshall, started by the families of Tyttery and Tyzack. The foreigners (perhaps after their cool reception by the inhabitants and commons of Warwickshire ) set up works in the extensive woodlands in that district. Peregrine Henzey was baptised at Eccleshall in 1586. In February 1600 Stephen Botting “of the glass-house, glass-maker,” was married, and two years later James Leggeeye, Frenchman, glassmaker, was married to one Judith Tyzacke of the same place. In the same year, Edward Henzey (of Wisborough, Sussex) married Sara Tetrye of Eccleshall (undoubtedly a form of Tyttery): in the next year, George Henzey, glass-maker, died. So here in Eccleshall we find quite a flourishing little glass colony, the existence of which had been forgotten.

D. N. Sandilands

JSGT 1937 V21 T215-T234

An account is given of the most desirable properties of coals for use in gas producers, illustrated by nineteen coals from seven different localities. Dull, hard, non-caking coals are preferred, of low moisture, ash, and volatile matter contents, the fusion point of the ash in a reducing atmosphere being about 1300°. It is recommended that the properties of different coals should always be compared by the same coal sizes, and that the selection of a coal should finally depend on the calorific value of the coal as received, Some sacrifice may be made with regard to the bulk density of the ash, provided attention be paid to those parts of the producer and furnace structure likely to be affected by ash deposition. High chloride contents should not be tolerated because of their deleterious effect on the furnace refractories. The permissible sulphur content depends on the type of glass melting operations involved.

D. Sismey

JSGT 1943 V27 T133-T206

Coloured Glasses. A Monograph Prepared For The Society of Glass Technology Under The Provisions of the Glass Research Association Trust. Part I.  The Constitution of Coloured Glasses.

Woldemar A. Weyl

JSGT 1943 V27 T265-T295

Woldemar A. Weyl

JSGT 1944 V28 T158-T266

Woldemar A. Weyl

JSGT 1944 V28 T267-T354

Woldemar A. Weyl          

JSGT 1945 V29 T289-T390

Woldemar A. Weyl

JSGT 1946 V30 T090-T172

Woldemar A. Weyl

 JSGT 1922 V06 T156-T160

An interesting example of superinduced prismatic structure has recently been observed in sandstone blocks used in the construction of a glass tank furnace. The formation is identical in every respect with that which takes place in Nature, particularly in the case of a columnar basalt, and is, in fact, a reproduction of contact metamorphism on a small scale. The tank in question was used for the manufacture of green bottle glass. The melting compart156ment measured 19 ft by 15 1/2 ft and was separated from the rounded working end by a fixed bridge. The dead weight capacity of the tank was approximately 84 tons, when filled to a depth of 3 ft. The construction of the tank was on the ordinary lines, the bottom being of fireclay binders built into a series of six arches. Sandstone blocks formed the lowest course of the sides, and these were surmounted by two courses of fireclay blocks.

John Currie

JSGT 1954 V38 T285-T296

A detailed study is made of theoretical and practical concepts regarding the compression factor in different types of glass-to-metal seals.

H. Adam

JSGT 1935 V19 T208-T216

A spectroscopic study of glasses containing iron and manganese oxides, both separately and in conjunction, supported by a consideration of the colours produced by oxides and salts of manganese in aqueous, hydrochloric acid and vitreous solutions, as well as in solution in organic solvents, has led to the conclusions that (l) the proportion of the purple colouring Mn₂O₃in a manganese-containing glass increases with increasing concentration, but in any case is very small compared with that of the faintly colouring manganous oxide; (2) the Mn₂O₃ is very sensitive to the reducing action of FeO present; (3) glasses containing 2%iron oxide and manganese oxides up to 4 or 5% are relatively transparent to light by reason of the reduction of purple Mn₂O₃ by FeO to relatively colourless MnO, and the corresponding oxidation of bluish-green FeO to the golden-brown and more transparent Fe₂O₃ (4) in mixed iron and manganese oxide-containing glasses the ultimate colour depends on the total Fe₂O₃ and MnO present.

W. E. S. Turner & W. Weyl

JSGT 1924 V08 T105-T109

A successful veteran American glass man has often made the comment, when called on at conventions of glass manufacturers, to express his views as to directions in which one may look for progress in the glass business, that he has two specific recommendations to offer: the more widespread employment of chemists and technical men throughout the industry and the more extensive use of pyrometers and heat-measuring devices and apparatus. Possibly the two suggestions are more or less correlated, as scientifically trained men are likely to demand more exact information of heat conditions than it has hitherto been the practice to obtain. Furthermore, the modern tendency is to demand more exact records of performance in all branches of the glass industry, largely resulting from the adoption of costing systems by progressive manufacturers. Details of the business are much more closely inquired into and recorded, material, labour, fuel, and furnace costs segregated, and the final cost of the product is more or less accurately determined. Practically every manufacturer would like to know more about his business, but the ways and means of so doing often appear too expensive or complicated, throwing some doubt on their accuracy, especially if the resulting figures are arrived at by complex processes.

William M. Clark

JSGT 1936 V20 T530-T535

An account is given of investigations carried out in the USA on methods for determining the mechanical strength of glass. The properties examined are the compressive and transverse breaking strengths, the former on cylinders, the latter on glass rods.

L. Navias

JSGT 1942 V26 T177-T196

A comparison has been made of the direct shadow, polariscopic and photo-optical methods of estimating the degree of cordiness of glass by the examination of numerous samples of cordy glass obtained from eight experimental melts made under controlled conditions. For routine comparison and control the direct shadow method is an extremely sensitive one, and very satisfactory estimates of cord can be made by it and a “figure of merit” obtained. The polariscope method is less sensitive for direct inspection, and its most useful employment is in the examination of ring-sections, by means of which the distribution of glass throughout the article can be checked. A photo-optical method has been developed which gives a measure of the cords present, depending only on the accuracy and sensitivity of the apparatus employed. By its means the direct shadow prints may be graded. A variant of the thermostatically controlled centrifuge method of Turnbull has been devised and built in order to separate, from finely powdered samples, glass of differing densities. An estimate of inhomogeneity may be obtained in this way by giving the total density spread of a sample, or the mass-density distribution in the form of a grading curve. A general comparison of surface defects can be made by a shadow method, but optical or photographic methods are to be preferred. Examples are given of the general surface condition of pressed, blown, and blown and spun ware made by automatic machinery. A numerical value for surface condition is more difficult to obtain even than for cord, but satisfactory ”figures of merit” have been derived from photographs of the surface at magnifications between 50 and 100 under conditions of oblique illumination. The dependence of surface condition on that of the mould is illustrated by photographs taken of samples of a small panel bottle taken at regular intervals throughout the mould life. It is suggested that the methods described should be more widely adopted in the routine control of glass quality.

A. J. Holland, M.Sc., Ph.D., A.R.C.S., Eric Preston, Ph.D., D.Sc., F.I.C., F.Inst.P., and Professor W. E. S. Turner.

JSGT 1941 V25 T100 T120

Being the Presidential Address delivered to the Society at the Annual General Meeting, at Sheffield, April 23rd, 1941

S. B. Bagley, LL.D., F.S.G.T.

JSGT 1952 V36 T105-T114

Measurements have been made of the creep and recovery of a soda–lime–silica glass over the temperature range 20–80°C. The deformation of the glass can be divided into three parts:1. An instantaneous elastic strain; 2.A Newtonian viscous flow at constant rate, which is not recoverable. 3. A delayed elastic strain which develops slowly according to a decay function of time and which is completely recoverable.

S. Pearson

JSGT_1922_V06_T181-T204

About three years ago, the Provisional Specifications were prepared by a Committee of this Society which had the benefit of consultation with representative manufacturers of refractory materials for the glass industry nominated by the Employers' National Council for the Clay Industries. These specifications, it was hoped, would tend towards standardisation and uniformity in the production and use of the refractory materials dealt with. The provisional specifications referred to had, therefore, been prepared based upon a combination of existing scientific data with the best practical experience, and it was hoped, in view of the fact that failures due to unsatisfactory refractories were sufficiently numerous, that all manufacturers of refractories for use in connection with the glass industry would take such steps as would enable them to comply with the specifications framed, which, although admittedly somewhat severe, seemed to be necessary if Great Britain was to keep pace with the progress of an important and rapidly reviving industry, namely, the production of glassware. Astonishingly small use has, up to the present, been made of these specifications either by the refractory materials manufacturers or by the glass manufacturers, and it is pertinent to inquire why this is so. Are they adequate and if not, in what direction should they be amended. Possibly the neglect of them is due to a recognition of the fact that, at best, the provisional specifications were bound to be tentative, and a great deal of data bearing upon the points at issue had yet to be collected before really definite standard specifications could be prepared which would be acceptable and useful both to the users and the makers of glasshouse refractories.

W. J. Rees

JSGT 1934 V18 T336-T390

 The problem of permeability is examined in a critical manner, the significance of permeability measurements discussed, and methods of determining it examined. The assumptions involved in the calculation of permeability are reviewed, and the necessity stressed of recording the volume of gas passing through the sample as the volume at the mean absolute pressure of the test. A definition of permeability which takes account of this fact is proposed. Data are quoted on the relation of permeability to refractory type, and to the method of manufacture.

The temperature–permeability relationships are discussed on the basis of a fundamental concept of permeability, and the influence on permeability of firing temperature, the amount and kind of grog in the refractory, and of the amount of bond present. Various avenues of research are indicated throughout the paper.

Eric Preston

JSGT 1929 V13 T129-T130

In a study of opal glass many fundamental problems arise, and much further work is still required to obtain essential physicochemical data on the equilibrium relations of the glassy and crystalline phases.

Link to JSGT_1927_T362-T374.pdf

J.F. Hyslop

JSGT_ 1931 V15  T012-T029

 From the work of O. Stumm, W. König and H. Jebsen-Marwedel, it is known that temperature gradients in glass tanks cause convection currents in the molten glass. The currents are sometimes of considerable strength (with velocities up to 12metres per hour) and they can influence appreciably the melting and refining processes as well as the homogeneity of the glass. König first gave the right explanation, which is that glass flows from regions of high to regions of lower temperature, because it has a greater density in the latter regions.

G. Gehlhoff, with W. Schneekloth & M. Thomas (Translated by H. W. Howes)

JSGT 1951 V35T206-T225

In continuation of previous work on the configuration changes of glass in the transformation range, the density changes occurring in fused silica at 1000–1500 have been studied. The equilibrium density curve has been determined and the time taken to reach equilibrium found for various temperatures. The coefficient of expansion at temperatures below 1000° has been found to alter when the configuration changes; the density changes measured at room temperature are corrected for change of expansion in order to give the changes of density at the temperatures of heat treatment. The equilibrium density of fused silica is found to increase with increasing temperature, and this is attributed to the tendency of the open random network structure to break down at higher temperatures. The final section gives preliminary results on measurements of the volume of fused silica as a function of pressure in addition to time and temperature.

R. W. Douglas & J. O. Isard

JSGT 1922 V06 T228-T231

We have shown in previous papers that whereas, in the trisilicate glasses of general formula 6SiO₂.xRO.yNa₂O, (where x+y=2) substitution of the soda by lime brings about an increase in density, similar replacement by magnesia produces a specifically lighter glass. The glasses containing the mixed bases lime and magnesia would be expected to have densities intermediate between those containing only the single base.

S. English & W. E. S. Turner

JSGT 1958 V42 T084-T087

To determine the total quantity of selenium left in a glass after addition of a decolourizer, samples of the glass with and without the selenium additive were irradiated together in a nuclear reactor. By comparing the resulting y-radiations with a differential type of scintillation spectrometer, selenium contents of 10 parts per million were readily determined. This use of inactive tracers revealed by radioactivation analysis is of wide potential application in production processes where radioactivity in the product has to be avoided.

J. L. Putman & W. H. Taylor

JSGT 1936 V20 T139-T143

The viscosity of quartz glass in the softening range was determined by the elongating rod method from 1100° to 1450°, The viscosity varied from 5×10¹³ to 9×10¹¹ poises. The temperature–log viscosity curves changed direction at about 10^12·5 to 10¹³ poises corresponding to a lower limit of the softening range, T₀, of 1220° to 1240°. Samples of silica glass from different sources gave slightly different results, due to the varying amounts of impurities present.

M. P. Volarovich & A. A. Leontieva

JSGT 1934 V18 T079-T087

The transformation temperature of seven different optical glasses of very different composition and physical properties have been determined at Jena, at Berlin and at Sheffield, using in the first-named laboratory the thermal expansion method, and in the two others an electrical resistance method in different forms. The results are all in such good agreement as to indicate, in general, that the transformation point, when the glass is heated at a rate of approximately 4° per minute, can be determined with an error not exceeding ±5°.

E. Berger, M. Thomas & W. E. S. Turner

JSGT 1950 V34 T273

Appended to original file

Correction to the paper in the August issue, pp. 153-172

J. E. Stanworth

JSGT 1950 V34 T153-T172

The development is described of “soft” glasses with good transmission to 2537 Å and suitable for the construction of efficient sources of bactericidal radiation (“germicidal” lamps). A satisfactory glass has the percentage composition SiO₂ 65, B₂O₃ 2, Na₂O 5·5, K₂O 9·5, BaO 18 and contains about 0·01% iron oxide expressed as Fe₂O₃ the iron being reduced to the ferrous state by addition of aluminium powder to the batch. Sulphates present as impurities in the raw materials (for example, in the barium carbonate) or deliberately added for refining purposes reduced the ultraviolet transmission of the glass; and small amounts of titania, present as an impurity in many sands, were also harmful. A soda–lime–silica glass of percentage composition SiO₂ 75, CaO 10, Na₂O 15 also transmitted well at 2537 Å even when it contained about 0·01% iron oxide, in contrast to the earlier findings of D. Starkie & W. E. S. Turner, who failed to obtain transmission at such short wavelengths in a glass of the same composition. Changes of basicity or of soda–potash ratio within the limits studied had a negligible effect on ultraviolet transmission. Melts in the series Na₂O–B₂O₃ showed that the transmission to 2537 Å decreased with increase in Na₂O content, but the transmission of any particular glass depended markedly on the type of raw materials used in its preparation. Melts made using borax were superior to those using sodium carbonate as the source of Na₂O, and the addition of a small percentage of aluminium powder to the batch also improved the transmission. Lithium borate glasses were slightly better and potassium borate glasses a little worse than corresponding sodium borate glasses; but nothing absolute is claimed for these results, owing to unavoidable impurities even in the purest available raw materials. It was shown that small percentages of iodine in a sodium borate glass caused marked absorption of 2537 Å radiation. Sodium borate glasses containing a small percentage of sodium sulphate, melted under reducing conditions, may be colourless, blue or yellow, depending upon the Na₂O content, the ultraviolet transmission being reduced in all cases by the presence of sulphur. Inclusion of 0·5% ZnO or CdO in a sodium borate suppressed the sulphur blue colour and increased the ultraviolet transmission; a similar inclusion of CaO, MgO or BeO failed to suppress the blue colour, although the addition of BeO improved the 2537 Å transmission.

J. E. Stanworth

JSGT 1954 V38 T371-T382

A completely automatic apparatus for differential thermal analysis is described in which the controlled heating rate is variable up to 20°/minute. Both reference and differential temperatures are plotted simultaneously on a standard 0–20 mV recorder.

F. W. Wilburn

JSGT 1939 V23 T171-T196

Preliminary measurements of the cooling rates of, the different glasses showed that the coloured glasses examined had a higher radiation power than a colourless glass of similar composition. The emissivity of these glasses depends not only on their colour, but also on the concentration of the colouring agent present in the glass.

R. Halle, B.Sc.Tech., Eric Preston, D.Sc., Ph.D., and Professor W. E. S. Turner

JSGT 1946 V30 T356-T363

The effect of cobalt oxide on the temperature distribution of molten soda–lime–silica glass was studied in a laboratory scale tank furnace. To the parent glass of percentage composition SiO₂ 73·5; Al₂O₃ 1·0; CaO 9·0; Na₂O 16·5, cobalt oxide was added in amounts varying from 0·0022 to 0·44 per cent., care being taken to maintain a constant concentration of iron oxide present in the glass as impurity. For the series of ten glasses obtained, the vertical temperature gradient at the centre of the mass of glass was determined. Up to about 0·025 per cent. cobalt oxide, a concentration bestowing a distinct coloration on the glass, the temperature gradients, varied from 7° to 12° per inch depth, these values being considerably lower than those found in amber or green glasses. Only when the amount of cobalt oxide exceeded 0·1 per cent., was there an appreciable increase in the temperature gradient. Similarly, a commercial cobalt· blue glass had a temperature gradient of 36° per inch as compared with 45° and 85° per inch for an amber and a dark-green glass, respectively. The investigation showed that the colour and transmission of cobalt-blue glasses at room temperature cannot be taken as an indication of their heat-transmitting properties in the molten state.

 R. S. Allison, R. Halls & W. E. S. Turner

JSGT 1957 V41 T229-T244

In the decoration of glass, Dutch artists were supreme with the diamond-point during the seventeenth and eighteenth centuries, and reached a high standard with wheel-engraving in the late seventeenth and the eighteenth centuries. It has long been recognised that many seventeenth-century glasses with English connections are likely to have been decorated by Dutch artists, and this article attempts to indicate which glasses are English but decorated by Netherlands artists, and which were made and decorated in the Netherlands. In the eighteenth century, English lead glass was much admired on the Continent, and formed a favoured vehicle for the Dutch engravers, whether with diamond-point or with wheel. Constant attempts, however, were made on the Continent to reproduce English crystal and to make glasses in the English style. It is therefore necessary to be alive to the possibility that an English-looking glass is in fact Netherlands-made, and the article discusses the centres where this glass might have been produced, as well as the style of engraving and cutting with which both English and Netherlands glasses were decorated.

R. J. Charleson

JSGT 1958 V42 T214-T226

The work described was undertaken in collaboration with the United Glass Bottle Manufacturers Ltd. A suitable bottle was chosen and a protective silicone film was deposited by spraying a silicone fluid into the leer. The process was carried out on a production scale. The siliconed bottles were packed into several types of case and corresponding quantities of non-siliconed bottles were packed similarly, as controls. The bottles were carefully examined under standard conditions for damage of any kind, both on arrival and after being washed in either water or 5% caustic soda solution. The siliconed bottles were shown to have sustained far less damage than the controls under similar conditions. In particular, the degree of scuffing had been reduced to a remarkable extent. It appeared that the protective property of the silicone film was not appreciably reduced by washing the bottle once either in water or in 5 % caustic soda solution. The bottles were successfully subjected to normal packeting operations, including labelling. The bottle contents were not demonstrably affected by the silicone coating on the interior surface. Condensation of water on the interior surface of the silicone bottles above the liquid contents precluded the use of these coated containers for pharmaceutical aqueous liquids. Further trials were carried out on containers used for non-aqueous preparations and again a substantial reduction in wastage was demonstrated as a result of silicone treatment. It is considered that it has been demonstrated that the treatment of glassware by silicones substantially reduces the incidence of breakage and other surface damage. The development of condensed moisture on the interior surfaces of siliconed containers containing aqueous liquids may limit the application of the method for pharmaceutical products but nevertheless the method shows considerable promise and justifies further investigation.

D. A. Hughes

JSGT 1943 V27 T113-T132

A contribution to the general history of optical glass manufacture. Edmond Feil was a great-grandson of Pierre Louis Guinand (1748–1824), the inventor of the modern process of manufacturing optical glass. He spent the earlier part of his working life in Paris in the optical glass works of his father, Charles Feil; had associations with Chance Brothers and Company for a brief period, commencing March, 1887; and later went to USA where, in 1893, he was engaged to superintend the optical glass factory of Macbeth and Company of Pittsburgh, Pa. Between 1897 and 1902 he was employed by the Manhattan Optical Company at Cresskill, N.J. An account of his optical glass formulas and procedures, as communicated to Chance Brothers and Company, is given.

W. H. S. Chance

JSGT 1944 V28 T355-T367

JSGT 1929 V13 T248-T262

Up till quite recently the use of substantial proportions of boric oxide in glasses has been confined to those needed for chemical and heat-resisting purposes and for types of optical glass. It is quite true that both in America and in England glass batches containing borax have at times been employed in tank furnaces melting bottle glass, but the amounts have been very limited and no investigation had been vigorously carried out to ascertain if, and how far, boric oxide was of advantage in such glasses. The reason was partly economic. The cost of borax and boric acid has until comparatively recently been too high to permit of its use in mass production glass. The recent very substantial fall in price of boric oxide compounds (the price of borax now being only very approximately one-half of that prevailing 10 years ago) has changed the aspect of the problem, and during the past 18 months or more experiments have been proceeding on the large scale in America to determine the value of borax in melting bottle glass, particularly colourless bottle glass.

Violet Dimbleby, Michael Parkin, W. E. S. Turner & Francis Winks

JSGT 1923 V07 T025-45

Among the fundamental physical properties of glass, one of the most important is its viscosity, because the ease with which a glass flows at various temperatures determines to a very large extent its melting and working properties, and these, in turn, whether or not glassware can be manufactured successfully. Besides influencing these properties, the viscosity of the glass in its nearly rigid condition also determines the temperature at which annealing is most suitably carried out and the range of temperature through which it is necessary to regulate carefully the cooling of the glass.

S. English

JSGT 1929 V13 T037-T047

In the case of potash–lead oxide–silica glasses, with which the present report deals, the question of the influence of the cullet added is still more important in glassmaking practice in the sense that a much larger proportion is used than is customary in melting the soda–lime–silica glasses.

S. English, G. A. Green, F. W. Hodkin & W. E. S. Turner

JSGT 1921 V05 T121-T123

SILICA after fusion is noteworthy by reason of its very low thermal coefficient of expansion, which makes it extremely valuable for conversion into articles which have to res st sudden changes of temperature. This property of low thermal expansion it also bestows on sllicate gla.sses, although whether to the same extent
as in the free state is not yet clear. In previous attempts to calculate the thermal coefficient of expansion of glasses, * we have assumed that silica takes \vith it into the glasses this same low expansion rate. This certainly is an assumption , although the agreement between the calculated and the experimental results
does at any rate not contradict it.

S English and W E S Turner

JSGT 1929 V13 T296-T303

In previous papers,* we have described the results of investigations on the melting of soda–lime–silica and potash–lead oxide–silica glasses from mixtures made up of varying proportions of batch and cullet. The influence of the proportion of cullet was ascertained on the rate of melting and refining as well as on the working properties, and the conclusion was reached in all cases examined that there was a shortening of the working range of the glass as the proportion of cullet to batch was increased. This phenomenon could not clearly be ascribed to continued loss of alkali content, or to other appreciable changes of composition such as increase in iron oxide content.

Links to this issue papers, pages beginning: T025, T037, T047, T064

A. A. Childs, V. Dimbleby, H. W. Howes & W. E. S. Turner

JSGT 1933 V17 T102-T117

The paper demonstrates a method for improving heat exchange in regenerators by increasing the effectiveness of the heat exchange in the lower zones of regenerators. The design of the filler bricks is such that the regenerator is divided into three zones, the lower of which have successively increased restriction to the path of the gases and maintains the velocities of the gases by compensating for their reduced volume at the lower temperatures. Turbulence is also provided in these lower zones to improve the effectiveness of the heat exchange. The bricks required are of a comparatively simple design, and a special feature of it is that, although the velocities are maintained in the lower zones of the regenerator, it is so arranged that the gases do not pass through the regenerator any more quickly than would be the case when passing through straight-walled vertical passages of the same height with reduced velocities in the lower zones

W. B. Mitchell & H. D. Bennie

JSGT 1923 V07 T228-T237

It has been well known in the glass industry that salt-cake, especially when used in tank furnaces, is corrosive of the fire clay blocks with which the furnace is lined. In many instances the corrosion has been due especially to the existence of a floating layer of the saltcake, and the most severe corrosion was therefore experienced on the flux line blocks, these requiring most frequent repair.

Edith M. Firth, F. W. Hodkin & W. E. S. Turner

JSGT 1953 V37 T003-T006

The use of a commutator to prevent electrode polarisation during d.c. conductivity measurements on glass is discussed, and the advantages of a high speed relay system are given. The achievement of efficient commutation and a balanced alternating output wave-form is described.

P. L. Kirby

JSGT 1944 V28 T463-T468

Energies of dissociation into the component ions have been computed for many simple and complex oxides. From these, 51 energy constants have been obtained, which may be used to compute approximate ionic dissociation energies in crystals and glasses in which each “metallic” atom is surrounded by oxygens. The relation of these constants to the Periodic Table is shown and briefly discussed. Slight deviations from additivity in silicates and phosphates are related to the dependence of the energy on the number of oxygens in SiO₄ and PO₄ groups which are shared with other Si or P atoms.

Maurice L. Huggins & Kuan-Han Sun

JSGT 1956 V40 T520-T541

This paper reviews three different types of toroidal glass vacuum chambers which have been developed for use in betatrons operating at 16 and 20 MeV. General problems of design are dealt with; in particular, the problem of minimum wall thickness in relation to collapse under vacuum loading. Design tests on models and on sections are described. The relative merits of the three alternative methods of construction are discussed. The development of special processes and equipment for the 20 MeV toroid is described, and limits of accuracy are quoted.

S. J. Morrison

JSGT 1950 V34 T063-T068

Heat-treated glass fibres are shown experimentally to undergo contraction equally in the longitudinal and cross-sectional directions of the fibre, thus showing them to consist of isotopic material and without appreciably stronger longitudinal than transverse bonds. Determination of fibre strength yielded approximately the same results whether made by tensile or torsion tests, that is, under stresses applied longitudinally or at 45° to the length; and this being so, the so-called Griffith flaws cannot be oriented in any one direction.

W. H. Otto & F. W. Preston

JSGT 1945 V29 T250-T257

The heavy mineral percentage of a number of British and foreign sands used for the manufacture of colourless glass has been determined. A rapid analytical method is given for the determination of their chromium content and also that of the sand itself. Possible methods for the elimination of heavy minerals from sand are discussed. An indication is given of the amounts of iron oxide and chromium oxide permissible in soda–lime–silica glass melted in a tank furnace so that satisfactory decolourising may be effected.

F. W. Adams

JSGT 1928 V12 T190-T204

In the production of good clean glass and in the maintenance of uniformity and continuity of production there is no more important factor than that of the durability of refractories.

J. F. Hyslop, R. F. Proctor & H . C. Biggs

JSGT 1959 V43 T094-T099

The paper describes the progress made in the application of a flame photometric method in a glassworks laboratory. The equipment used is described and a method is given for the determination of the alkali metals in glasses and glass-making materials. Factors influencing the accuracy of the determinations are discussed and results are given of a number of analyses carried out on a variety of samples.

G. A. Hedgecock

JSGT 1952 V36 T242-T265

The effects of small additions of soda on the rate of recrystallisation of alumino­silicate mixtures during firing, and on the properties of the fired materials, have been studied. The mixtures consisted of calcined alumina and a specially fine china clay, in proportions corresponding with sillimanite (Al₂O₃.SiO₂) and mullite (3Al₂O₃.2SiO₂), and sodium silicate was added in amounts to give up to 0·5% of soda in the “sillimanite” mixtures and up to 3% in the “mullite” mixtures. Two “sillimanite” mixtures and one “mullite” mixture were made with fired china clay (Malachite) replacing some of the raw china clay. Grogs were prepared from each mixture by firing for 20 hours at 1400°, 1450°, 1510° and 1590° respectively, the linear shrinkages during firing and the apparent porosities of the fired grogs being determined in each case; the true porosities of a number of the grogs were also determined. By X-ray methods determinations were made of the proportions of mullite, corundum and non-crystalline material in a selection of the fired grogs. Refractory mixtures were made by mixing portions of each grog with a suitable proportion of the corresponding “raw mix,” and corrosion tests were carried out on extruded specimens of each mixture after firing for 20 hours at the temperature used in the original firing of the grogs.

R. Heeley & H. Moore

JSGT 1929 V13 T064-T070

The experiments which form the subject of this report arose out of our experiences in the preparation of soda–lime–silica glasses under varying conditions, in particular (a) from moist batches, and (b) from a batch in which the limestone had been used in the form of rather coarse-grained particles, that is to say, coarser than 30 mesh; whilst the investigation is also linked up with our experiments on the influence of cullet on the working properties of glass.

G. A. Green, F. W. Hodkin & W. E. S. Turner

JSGT 1936 V20 T429-31

Numerous physical properties of glass are additive functions of the composition, and under the same experimental conditions the influence of composition on the mechanical properties is evident, but is greatly over-weighted by the effect of experimental conditions, such as size of test piece, time of loading, temperature and even the kind of fracture. A study of these factors is a necessary part of the development of glass as a structural material.

E. Albrecht

JSGT V15 T212-T218

In 1824 the Royal Society appointed a committee of its Fellows and the Members of the Board of Longitude to consider the improvement of optical glass, especially such as would be suitable for telescope objectives. In 1825 a sub-committee was formed, consisting of Faraday, who was to make the glasses, Sir J. F. W. Herschel, who was to determine the refractive indices, and G. Dollond, who was to make up the glasses into lenses. A lengthy report was drawn up by Faraday at the end of 1829 and constituted the first of his Bakerian Lectures, and so long was it that it took three days to read, November 19th, December 3rd, and December 10th.

Arthur Marshall

JSGT 1947 V31 T259 -T263

Experiments on the large scale showed that finely-crushed quartz used in a soda–lime–silica glass batch caused the latter to melt and become refined appreciably faster than when Belgian sand was employed. The carry-over of dust into the regenerator chambers was remarkably low.

Finn Gjesmoe

JSGT 1925 V09 T028-T032

 ''A Striking Instance of Fireclay Corrosion through the Action of Salt-cake." Discussion

W. E. S. Turner

[Link to JSGT_1924_T284-T285]

JSGT 1936 V20 T570-T573

The quality of tank blocks is determined by the properties of the grog used in the fireclay mixture. The properties and uses of different types of grog for glass refractories are outlined.

R. Barta

JSGT 1931 V15 T119-T128

Several times during the past four years tests on the flow of glass have been made on small reheating tanks used in the manufacture of glass tubing and bulbs. The indicator employed for tracing the currents of glass in colourless glass was either of cobalt-blue glass or cobalt oxide. Although the temperature distribution was not measured, this colour indicator proved very useful and for this reason the records are now published.

A. L. Marden

JSGT 1957 V41 T331-T336

My knowledge of men who make glass has, in many instances, been first-hand, but I must confess that my knowledge of glass itself in relation to its fabrication and manipulation has been entirely second-hand. To me 'forehearth' might indicate the location of the hearth-rug, a 'taker-in' be a form of rascal peculiar to the glass industry, and a 'lehr' merely an awkward word whose anomalous spelling provided a challenge to the erudition of Professor Turner! For forty years I have been beside you, forty years in which my interest in the industry and its personalities has steadily grown and has been so warmly reciprocated that, although I may never claim to be 'inside', I certainly no longer feel 'outside' the glass industry.

C. E. Ramsden

Fracture and Cavitation in Glassy Materials

JSGT 1955 V39 T394-T403

The phenomena of cavitation in liquids and of the incipience of cracks in glass-like solids are shown to have many features in common. In both cases mechanical energy stored in the system is converted into surface energy, and both" bubbles" and cracks grow at inhomogeneities in the material under the effects of applied tensile stress. The paper describes how these phenomena may be studied in solids and liquids by the use of small explosive charges which produce stress pulses of only a few microseconds duration. Transient cavitation effects obtained in water from such pulses are described, as are also similar fracture patterns found in "Perspex" cylinders. The transition from liquid to glass-like behaviour has been studied by observing fracture effects in an organic glass. This material softens from a brittle solid to a viscous liquid in the temperature range 10°–50°.

H. Kolsky

JSGT 1957 V41 T157-T267

Cone cracks have been measured as a means of studying the effect of liquids and gases on the fracture of glass. A series of experiments was carried out from which the fracture energy of plate glass in contact with seventeen different liquids and four gases was calculated. Five optical glasses were cracked under distilled water and the results compared. The fracture energy of plate glass varied from 2900 dyn/cm in contact with water to 7800 dyn/cm in contact with carbon dioxide.

C. J. Clilf

JSGT 1955 V39 T134-T144

Two glasses, a borosilicate ("Pyrex") and a soda-lime, have been studied. Series of fibres were prepared, each fibre being drawn under a constant axial load which was retained during cooling. Determinations of dimension changes as a result of heat treatment were made. Changes in length were determined by direct observations, while changes in radius were calculated from observed changes in density. The results show that, superimposed on the uniform compacting usually associated with the heat treatment of chilled glass, there is a relaxation of a "frozen" elastic deformation which is proportional to the stress applied, this stress being calculated from the load and the diameter of the fibre after drawing. Each glass was thus found to exhibit a modulus, analogous to Young's Modulus, being the ratio of stress to "frozen" longitudinal strain, and a ratio: of "frozen" lateral strain to "frozen" longitudinal strain, analogous to Poisson's Ratio. In addition, the fibres as drawn show optical activity proportional to the stress applied in drawing the fibre. It is therefore concluded that glass fibres as drawn cannot be isotropic, since some stress must be applied in drawing, and consequently an elastic deformation "frozen in." A short discussion of the significance of the results in relation to theories of glass structure is given. An elementary mathematical treatment, in which it is assumed that the glass, when being drawn into fibre form, behaves as a number of elastic elements and a number of fluid elements, gives a linear equation connecting the observed changes in length with the applied stress.

J. F. Stirling

JSGT 1927 V11 T205-T214

The observations recorded in papers dealing with earlier sections of this investigation were made for melting temperatures ranging from 1300° up to 1400°, and for the founding period of about 22 hours. Moreover, we have shown that similar glasses, melted in tank furnaces of various sizes at temperatures from 1350° up to at least 1400°, and maintained in the furnace for periods up to 5-7 days, also contain arsenic varying in amount from 55 to nearly 100% of that added. In these cases also, the residual arsenic is present largely in the form of arsenate.

[Part II above]

Edith M. Firth, F. W. Hodkin, W. E. S. Turner & F. Winks

JSGT 1947 V31 T245-T253

The fundamental condition for glass formation is the existence of strongly bonded large networks or long chains of atoms in the liquid. Such chains or networks will not only prevent or hinder the orientation or arrangement of atoms in the chains or networks themselves but also that of other atoms not in the networks in the liquid near the melting point or liquidus temperature. Thus, a liquid structure will persist even after the liquid has been cooled below the melting point or liquidus temperature and indeed farther down to such temperatures that free long-range movement of atoms becomes impossible. In order to obtain such long chains or large networks, it is necessary that (1) the bond strength of atoms in the chains or networks be very strong; (2) small ring formation of these strongly bonded atoms be at a minimum; and (3) the relative numbers of various atoms in the chains or networks be such that at least a continuous chain is configuratively and structurally possible. At the same time, the co-ordination numbers of the glass-forming atoms must be as small as possible to keep the bond strong. The general treatment is applied to all kinds of glasses, particularly oxide glasses. For the latter, the bond strengths of all M–O single-bond linkages have been calculated and are here tabulated for the first time. It was found that the bond strength of all the glass-formers is greater than 80 k.cal. per Avogadro bond, that of the intermediates between 50 and 80, and that of the modifiers below 50. The transition is continuous and the division into glass-formers, intermediates and modifiers is arbitrary. The table of bond strengths is particularly useful for the synthesis of new glasses and for the study of glass structure.

Kuan-Han Sun

JSGT 1939 V23 T154-T170

The variable factors in gas producer operation are enumerated and discussed. For the automatic control of a gas producer it is desirable to have a potentiometer recording controller for controlling the coal feed and gas temperature, a pressure recording controller to control gas pressure or blast volume, and a vapour pressure bulb recording controller to control blast saturation temperature. The operation of such devices is described. A study was made of the influence of the fuel bed on producer control. Local variations of the porosity of the fuel bed between 30 and 50% result in corresponding variations in velocity up to 2.5:1, whilst variations in particle size between 1 and 2 lead to a change in velocity o£ 1.4:1. Experiments were carried out on an 8ft Wellman producer comprising a study of the overall pressure drop with different sizes of fuel and thickness of fuel bed, and the exit velocity of the gas from the top of the fuel bed employing a Pitot tube measuring velocities of about 1 ft./sec., and later a "Velometer" measuring to 0·5 ft./sec.

J. G. Bennett

JSGT 1936 V20 T257-T278

The question to be answered is, how can the numerical values for the susceptibility to attack obtained by present methods be replaced by scientific constants. The time laws (parabolic and logarithmic laws) always contain two constants. In principle, therefore, two points for a given temperature and a reactive liquid are introduced into the time curve. The basis of measurement is the amount of decomposed glass per unit surface area or the thickness of the film removed. To obtain scientifically accurate constants, the greatest care must be exercised that the liquid is stirred at a sufficient rate that the glass is in an untreated condition and the surface area known.

Edwin Berger

JSGT 1941 V25 T241-T248

The relative melting rates and the devitrification temperatures have been determined of two series of glasses of parent composition 54% SiO₂, 34 PbO, 12 K₂O in which PbO was progressively replaced by 0-6% BaO, and by 0-6% CaO. Both these oxides resulted in easier melting and refining, the devitrification temperatures being reduced from 1020° to 920° for the 6% BaO glass, and to 915° for the 6% CaO glass. A comparison was made with nine commercial compositions known to be in use for the manufacture of domestic ware, the quicker melting rate of the majority of these glasses being due to the relatively high alkali contents, ranging from 12.56 to 18.2%. Based on one of these compositions, the melting rates and devitrification temperatures were determined of a series of glasses of composition 64% SiO₂, 20 PbO, 1.5 BaO, 1·5 CaO, (12 - x)K₂O and xNa₂O. Of this series, the glass containing 4% K₂O and 8% Na₂O appeared to be of greatest promise. It gave an excellent colourless glass, the cutting qualities of which compared favourably with those of the lead crystal glass commonly used in Great Britain, was rather more easily melted than that glass, and should be a durable glass.

 
Eric Preston, Ph.D., D.Sc., F.I.C., F.Inst.P., and Professor W. E. S. Turner

JSGT 1940 V24 T124-T138

A study has been made of the rate of melting in small platinum elliptical capsules heated in an electric furnace of a batch mixture, of approximately 2-5g in amount, yielding a glass of percentage composition 73.5 SiO₂, 10 CaO, 16.5 Na₂O. Pure soda-ash and calcium carbonate, both of controlled, constant particle size, were employed together with silica grains in one of two forms, the first derived from pure Brazilian quartz, the second from a sieved Dutch sand. Quenched Brazilian quartz was found to break down with peculiar characteristics leading to grains mainly of columnar or needle shape. As the result of a series of experiments at temperatures ranging from 1100° to 1450° it was shown that the rate of glass melting was approximately doubled for each 50° rise in temperature. The rate at any one temperature was found to be proportional to the area of the grains of silica. The violent chemical reactions were found to occupy 10 per cent of the total time of glass melting, and the solution of the residual silica 90 per cent under the conditions of experiment, the solution of the residual silica was at the rate of 0.003mm per minute of the grain diameter.

Eric Preston, Ph.D., D.Sc., F.I.C., F.Inst.P., and Professor W. E. S. Turner

JSGT 1941 V25 T136 T149

The rate of glass formation in the PbO-SiO₂ system has been studied employing 0·5gm charges in the temperature range 1000-1375°, the compositions ranging from 60 to 90 per cent PbO. Outside this range glass formation was limited by the devitrification of compositions containing more than 90 per cent PbO, and by the formation of a scum of undissolved sand grains, which prolonged heating did not remove, in those containing more than 40 per cent SiO₂. When, in the temperature range stated, glass formation was possible, solution of quartz was comparatively very rapid, the 80 per cent PbO mixture melting to a clear liquid in 4 mins.at 1375°, 8 mins at 1300°, 14 mins at 1200°, 18 mins at 1100°, and 30 mins at 1000°. The reaction between red lead and quartz, and litharge and quartz, was studied by microscopical and X-ray means. Pb₃O₄ was rapidly decomposed to PbO on heating, and the total loss in weight of the mixture comprised this oxygen loss, of approximately 2 per cent of the weight of Pb₃O₄, and a further slight loss due to volatilisation of PbO. In the litharge-quartz mixtures the only loss in weight was the small amount of PbO lost by volatilisation. As heating proceeded, the Pb₃O₄ (or PbO) lying between the sand grains progressively disappeared and was clearly transferred, by molecular attraction, and presumably in the vapour phase, to the surface of the quartz grain where reaction proceeded of a character dependent on the temperature. At temperatures above the melting point of PbO, a lead oxide-rich glass was formed on the surface of the sand grains, the PbO gradually penetrating them so that the amount of residual quartz gradually decreased with increased time and temperature of treatment. This was indicated microscopically and by the X-ray spectrograms, which also showed that no solid solution was formed. Immersion of the crushed product in bromoform showed also that the percentage floating on this liquid progressively decreased with longer times and higher temperatures of treatment. Below the eutectic temperature, 730°, between PbO,SiO₂, and SiO₂, a solid reaction appeared to proceed in the red lead-quartz and litharge-quartz mixtures, the X-ray spectrograms showing the formation of 2PbO,SiO₂, though this could not be detected microscopically.

Eric Preston, Ph.D., D.Sc., F.I.C., F.Inst.P., and Professor W. E. S. Turner

JSGT 1924 V08 T173-T181

 It was of considerable interest to extend the observations on alumina in glasses by a study of glasses analogous in composition to the boric oxide glasses, that is, to glasses in which silica is progressively replaced by alumina weight for weight. The results with the two previous series of glasses tested had given no indication of abnormality, such as is exhibited by the boric oxide glasses, but a strict comparison could not be said to have been made.

Violet Dimbleby, S. English, F. W. Hodkin & W. E. S. Turner

JSGT 1919 V03 T129-T131 (3 pages)

No examination had been made of two types of American glassware, namely, that of the Macbeth Evans Glass Co. and of the Libbey Glass Co. although P. H. Walker & F. W. Smither had tested them. When the results of the tests on the Macbeth Evans glass are compared, it will be seen that for every reagent tested the glass easily falls within the first class, so that the authors' tests have revealed now seven glasses which are good all round. The composition it will be noted, is that of a zinc borosilicate. Alumina is present only to a small extent, and is therefore, in this case, not the component which imparts resistance.

 Constance M. M. Muirhead & W. E. S. Turner

JSGT 1954 V38 T297-T364

A thorough-going investigation, having regard to present-day practice, has been carried out to re-assess the effects on the melting and refining rates of progressively introducing boric oxide as hydrated borax into batches for the production of six different container glasses and of one sheet glass, all of the soda–lime–silica type with variants containing one or more of the constituents AI₂O₃, MgO, BaO and F. The range of B₂O₃ in the glasses was 0–2·5%, substituted overall. In all, twenty-nine glasses were studied. The meltings were carried out in platinum–rhodium crucibles (to eliminate corrosion effects) in a town-gas-fired furnace fitted with arrangements for the strict control of the gas and air supplies. Measurements of melting and refining rates (the latter in terms of visible bubbles per 100 c.c.) were carried out at 1400°, 1450° and 1500°, with additional meltings at 1425° and 1475° in special cases. Measurements of the viscosity of all the glasses were carried out over the temperature ranges 525°–700° and 900°–1400°. Chemical composition control was exercised, not only by precise analysis of all the glasses, but also of volatilisation during the drawing of glass fibres and during the high temperature viscosity measurements. The rate of melting of the batches at all temperatures was significantly increased by progressive addition of B₂O₃, except in the single case of the sheet-glass composition. The rate of refining at all temperatures progressively increased with B₂O₃ addition in all cases. The rate of increase varied according to the character and amounts of the major constituents of the batches; but even in the presence of the fluxing constituents BaO and F, boric oxide still exercised a marked effect. Coefficients were worked out relating the concentration of B₂O₃ to the increases in melting and refining rate. The effect of increased temperature alone was also calculated. Increase of B₂O₃ within the range 0–2·5% progressively lowered the viscosity of all glasses within the range from 1400° down to a point between 6020 and 6250, dependent on the basic composition, below which introduction of B₂O₃ progressively increased the viscosity. Accordingly, the annealing temperatures were also raised by the addition of B₂O₃; the Littleton softening point, however, was progressively lowered. From the viscosity data it was shown that a given stage of shaping a glass article could be carried out at a somewhat lower temperature as addition of B₂O₃ was increased.

R. S. Allison & W. E. S. Turner

JSGT 1946 V30 T243-T277

Data of the weekly glass production and weekly fuel consumption for a total of eight melting campaigns involving four different furnaces are studied in detail. The furnaces are: (a) An oil-fired furnace the melting area of which was 83 sq. ft. fur two campaigns and 93 sq. ft. for a third campaign; (b) a producer gas-fired furnace of 580 and 560 sq. ft. melting , area in two separate campaigns; (c) a producer gas-fired furnace of 590 sq. ft. melting area; (d) a furnace of 171 sq. ft. melting area, fired with coke oven gas for two campaigns. The importance is emphasised of plotting some measure of furnace performance against statistically determined limits, in order that changes in performance shall be detected at the earliest moment. In Section II the variations of the thermal efficiency index Q are discussed in some detail, with a view to providing evidence of the technical value of calculating this index weekly and of plotting it on a “control chart.” A formal analysis of regressions connecting the weekly fuel consumption with the age of the furnace and with the weekly glass production is given in Section III. Some statistical methods not used in an earlier paper are illustrated. A general discussion of the data in Section IV provides evidence that some of the variability in furnace efficiency arises from the fact that changes in glass output from the furnace and fuel changes are not synchronised.

For all practical purposes the calculated value of Q is as good a measure by which to estimate the thermal performance of a furnace as the more involved and formal analysis dealt with in Section III. It is also shown that the correction for furnace area recommended by the Furnace Committee is a good practical approximation. In spite of the wide divergence of size of the furnaces studied and of the fuel used, it is possible to make a reasonable estimate of the magnitude of the proportion of additional fuel required each week to maintain the furnace at the required melting temperatures and of the amount of fuel which is associated with the production of one pound of glass under the present conditions of operation. A table is given of values for the construction of provisional control charts. These charts would serve as a guide to reasonably good practice until data are available for any particular furnace and method of operation, in order to assist those wishing to introduce the control chart technique for watching furnace performance.

B. P. Dudding

JSGT 1937 V21 T463-T475

Tests have been carried out to determine the difference in the thermal endurance values obtained when test pieces of glass rod are heated, under conditions otherwise similar in a furnace in the horizontal and vertical positions respectively. The average thermal shock values for fifty specimens, obtained for glasses of low thermal endurance, were, for the two furnace positions, within 2°, but for “Pyrex” rods there was a difference of 15°, the tests in the vertical furnace giving the higher values. The thermal shock values decreased with increasing diameter of rod, the relationship appearing to be hyperbolic and tending to a finite positive value as the diameter increased. The results of previous workers were shown to fit similar curves. A suggested “standard procedure” for thermal endurance tests is given.

R. J. Waller

JSGT 1940 V24 T041-T045

Glasses of the basic percentage composition SiO₂ 73.5, Na₂O 16.5, CaO 10.0 and containing 0.08% Fe₂O₃ were melted in platinum at 1500° in a gas-fired furnace both without and with the addition of sodium chloride and certain other volatile materials. The amount of sodium chloride added in all cases was equivalent to 0.2% Cl in the glass. The other volatile materials added with the sodium chloride and their amounts in different series were: (a) selenium, 0.4 oz. per 1000 lb. of sand, with cobalt oxide, in amount appropriate to decolourise the iron oxide, (b) borax, to yield 2.0% in the glass, (c) ammonium sulphate, 7.5 parts per 1000 of sand. Analysis of the resulting glasses showed that the losses of iron oxide in the presence of sodium chloride and other volatile constituents were: (a) sodium chloride alone, 11.3%, (b) with selenium 15.0%, (c) with boric oxide 17.5%, (d) with ammonium sulphate 28.8%.

R. Halle, B.Sc.Tech., and Professor W. E. S. Turner

JSGT 1931 V15 T255-T262

The solvent action of hydrofluoric acid on fireclay refractory material, especially tank blocks, as a basis for ascertaining the probable behaviour of the material when used in the glass-containing walls of a tank furnace was investigated and described by one of us in a paper read to the Society in 1929. The other of the two authors subsequently applied it extensively, and came to the conclusion that it was a very useful test. The paper reports on further experiences.

Links to JSGT 1929 T130 and 1930 T063

E. J. C. Bowmaker & .J. H. Partridge

JSGT 1917 V01 T223-238

Gas firing is an advance on the direct application of coal for heating purposes in which many of the industries are rightly interested, including the glass industry. Since the war began the use of high-grade gas for all such purposes has increased rapidly on account of the ease and rapidity of installation and the exact control and freedom from dust and dirt in any form which can be secured in operations. This progress is likely to be steadily continuous in a number of directions when the war is over, for reasons which I have attempted to specify, and it will be hastened if we can only secure that the new standards which must inevitably be adopted for public gas supplies are placed upon a more rational and scientific basis

J. W. Cobb

JSGT 1927 V11 T293-T300

The object of the present paper is to examine, as far as it relates to gas-fired furnaces, the proposition that "one British Thermal Unit is as good as another," or, in other words, that nothing matters except the calorific value of the gas consumed.

R. Wigginton

JSGT 1929 V13 T098-T111

It is obvious that the large quantities of gases evolved during the melting down of the batch are produced by the decomposition of the individual constituents of the batch, e.g,CO₂, SO₂, and O₂, from carbonates, sulphates, and the higher oxides. It is debatable, on the other hand, whether the fine bubbles which appear during the refining of glass or when cold glass is remelted, are due to gas dissolved in the chemical or in physical slate. It is possible to conceive of the solution of these gases in a glass in a physical manner similar to the solution of gases in liquids. A chemical solution may be conceived by imagining the gases to be present in the form of compounds, such as carbonates, sulphates, and oxides, which have escaped destruction during the melting process. The evolution of these gases during remelting may then be conceived to be a kind of secondary refining. The decision as to which point of view, the physical or the chemical, is correct, however, must be settled by experiment.

A. Becker & H. Salmang (Translated by H. W. Howes)

JSGT 1959 V43 T359-T403

It appears that in the case of optical glass manufacture the main sources of blisters are: (a) gases liberated from the pores in refractory walls; (b) gases which saturate the molten glass and which evolve as blisters on mechanical stirring. Rate of melting, kinetics of refining actions, the exhaustion of refining agents, play an important part in blister occurrence. There are many problems which require further study, e.g. interaction between various gases (particularly carbon dioxide) and molten glass, the glass-forming processes and the refining action.

V. T. Slavyanski, translated by S. Kruszewski

JSGT 1936 V20 T475-T497

Experiments on the sudden cooling of glass rods indicated an increased thermal endurance with decreasing diameter of the rods. The difference of temperature at which the first cracks appear is denoted as “technical thermal endurance” and the temperature difference at which the first transverse cracks occur as “true thermal endurance.” The latter is slightly higher and shows a narrower spread of the results. When annealed rods were tested there was not a great deal of difference whether the ends had been fire polished or not. For unannealed rods the “thermal endurance” is about 40° higher, with a much greater dispersion. For ordinary industrial glasses the coefficient of expansion gives a reliable indication of the value of thermal endurance. A method of determining the “thermal endurance” is described, in which the temperature difference is not increased, but the tests repeated until fracture occurs.

H. Schönborn

JSGT 1940 V24 T159-T175

On the basis of the X-ray patterns and assumptions as to the coordination between the oxygen atoms and those of silicon or boron, and with the ions Na⁺ or Ca²⁺ theoretical principles are discussed governing glass versus crystal formation; miscibility or immiscibility in glasses and anomalies in certain physical properties of glasses. The simple glasses soda-silica, lime-silica, and soda-boric oxide serve as illustrations in the views expressed.

Professor B. E. Warren

JSGT 1927 V11 T010-T023

If a block of glass, say a cube, is allowed to cool from the hot condition, the outer surface very quickly attains the rigid condition, owing to the rapidity with which hot glass loses its heat to the surrounding air. This question of the rate of loss of heat to the air is one regarding which we have practically no trustworthy data. Anyone who has experienced the difficulties incurred in the measurement of heat dissipation from bodies at a high temperature will understand the reason for the lack of information. There are so many practical difficulties involved in a research of this kind.

James Weir French

JSGT 1935 V19 T167-T170

The article reviews the main features of the exhibits of glassware shown at the Exhibition of Industrial Art held in January 1935, and discusses some of the lessons to be learned.

James H. Hogan

JSGT 1924 V08 T085-T092

The subject of this short paper is identified with what is, without doubt, the most famous piece of glass which students of glass-and we may include earthenware-know. Since the discovery of the original vase in the sarcophagus of the Emperor Severus near Rome in the seventeenth century it has been considered an example of the finest craftsmanship by both of these industries. This ancient vase, which is carefully preserved in the gem room of the British Museum, is made of glass, but up to the time when Josiah Wedgwood borrowed it from the Duke of Port, land for the purpose of imitating it in his jasper ware, it was thought to be of stone or earthenware. The error was excusable, as the dull appearance of the vase from age obscured the translucency or transparency which one naturally looks for in glass.

John Northwood

JSGT 1924 V08 T092-T093

The writer in this instance acts merely as a recorder; for he believes there should be placed on record in the Journal a brief account of the work of Mr John North wood, jun., whose modesty causes him to shrink from undertaking the task. The writer saw the plaque in the show-room of Messrs. Stevens & Williams at the Brierley Hill Glassworks, decided to extract what information he could about it, and to set it down in the Society's records, both as an acknowledgment of exceedingly fine work and as an encouragement to young and ambitious craftsmen in this country.

W. E. S. Turner

JSGT 1936 V20 T758-T763

The glass industry of Bohemia became famous in the 17th and 18th centuries. Means of competing in the 19th century with the lead crystal glass were found by improving the Bohemian potash-containing glass and glass colouring processes. The development of cased glass and the application of various finishing processes created a new popular art.

F. X. Jirík

JSGT 1919 V03 T253-T256  (4 pages)

In May last the Society had appointed a committee to deal with the problem of table-working glass, and a considerable amount of information had been collected but experimental work had been delayed pending the formation of the Research Association. The author might however be permitted to express his own views on the subject. The problem was largely one of furnace practice, for British glass tubing too often bore, evidence of the fact that the glass had not been founded at a sufficiently high temperature. He would urge manufacturers to' adopt small tanks and gas-fired furnaces for the manufacture of table-working glass tubing. Most direct-fired and semi-gas-fired furnaces could be converted for gas-firing without difficulty. In this respect our manufacturers must get abreast of the times.

M. W. Travers

JSGT 1933 V17 T050-T067

The paper discusses the operations passed through during the formation of a glass bottle in the moulds of a forming machine and discusses in some detail the relationship between the volume of the parison mould and the volume of glass in the article to be made. This relationship has been shown to be linear on the basis of 42 moulds operating on a blow and blow machine making articles varying from approximately 3to 30 oz in weight. The contour of the parison is also discussed and special consideration given to the case of wide-mouth containers and bottles of irregular shape. The qualities which are needed in moulds are enumerated and an account is given of the result of experiments on a number of moulds made from special alloys.

Thomas Wardley

JSGT 1936 V20 T717-T723

The different kinds of abrasive wheels employed in glass working are considered. Factors governing the choice of material are: the effect desired, the fineness of grinding, and the time and area of contact between tool and work. The grinding processes considered comprise facet grinding, grinding of jar and mouth edges, spectacle lens edging, crystal “cutting” and slitting.

A. Guilleaume

JSGT 1917 V01 T137-T139

The following remarks apply to the refractories used in tank furnaces. The comparatively low temperatures employed are favourable to long life" and there is an abundance of siliceous material available sufficiently refractory, so far as resistance to fusion is concerned. A high fusion point is, however, only one of the factors to be considered. Fusibility depends on the character of the atmosphere to which the material is subjected while highly heated. The nature of the charge in glass tank furnaces is such that at the temperatures employed, alkaline dust or vapours are a significant constituent of the furnace gases. It is therefore not merely a question of a reducing or oxidising atmosphere, but of one charged with vapours or particles with a distinct corrosive action on acid materials. The effect would be most marked at the exit ports, where the products of combustion were being swept away at a high velocity through narrow openings but, as the furnaces are reversible, the port blocks would each in turn be subjected to the corroding effect of the furnace gases. The obvious remedy would be to employ, in the portion of the furnace attacked, materials that would not suffer corrosion or the rate of corrosion of which would be such that the efficient life of the furnace would not be affected. But the first requisite for the selection of materials with the required properties would be an accurate knowledge of the composition of the furnace gases and the temperature, at which they are brought into contact with the ports of the furnace under discussion.

Cosmo Johns

JSGT 1958 V42 T062-T069

This paper discusses the special advantages of gamma-radiation techniques for level measurement under conditions where it is undesirable to have access to the interior of the vessel. Owing to the high temperatures inside glass furnaces and feeders, these are typical vessels where the above considerations apply. The limitations of the method with regard to accuracy, range, and radiation safety are discussed and it is suggested that the measuring apparatus could be adapted to form part of an automatic control loop. Brief reference is made to other possible applications of penetrating radiation in the manufacture of glass containers

E. W. Jones

JSGT 1918 V02 T005-T012

The graceful outlines of the older drinking glasses which had their origin at Bolsterstone are highly suggestive of an Egyptian influence, The bell-shaped bowls supported by stalk-like stems, though not so delicately formed as those of Murano, or Venetian drinking-vessels of the sixteenth century are so wonderfully like those of the lotus flower on its stalk that it is just possible English makers may have copied some early Phoenician example found in ancient store or burial mound. The fact that this shape of bowl is said to conserve the full flavour of the wine conveyed therein is another proof that the nearer we approach to Nature the more easily do we attain to that simple beauty which is wedded to objects of real utility.

Glass-making at Bolsterstone, near Sheffield, from about A.D. 1650 to 1750

Joseph Kenworthy

JSGT 1939 V23 T268-T280

Glasses containing 5% boric oxide plus phosphoric oxide, 20-35% alumina, and 15% of the basic oxides, lime, baryta, magnesia, and zinc oxide, the remainder being silica, have been studied with a view to finding suitable hard glasses for the inner tubes of high-pressure mercury discharge lamps. Glasses of this type containing more than approximately 27.5% alumina or more than approximately 7.5% lime, bubble too much during bench working in an oxy-coal gas flame. On the other hand, if the lime and alumina contents are too low, the glasses are difficult to found and work at the furnace. A glass of theoretical composition, SiO₂56, Al₂O₃ 24, CaO 7.5, BaO 7.5, B₂0₃ 5% has been melted on a semi-commercial scale in a small tank furnace constructed from china clayball clay blocks. Satisfactory lamps may be made using bulbs blown from this glass, and in addition to having excellent lumen maintenance can be run horizontally without magnetic control.

J. E. Stanworth, M.Sc.Tech.

JSGT 1947 V31 T114-T121

The various expressions which have been derived to determine whether an element acts as a network-former in glass, or as a stabiliser or modifier are discussed. On the basis of the author's theory of glass structure, in which such atoms as sodium are considered as playing a more or less important part in the maintenance of the continuous network, a covalency factor (ϕ) is assigned to the various elements. It is emphasised that very strong bonds tend to produce regular crystalline atomic arrangements, and that normally weak bonds may be constrained by the rest of the structure to have some additional covalent character. The arbitrary classification of elements into glass formers, stabilisers or intermediates, and modifiers cannot be supported.

H. Cole

JSGT 1929 V13 T183-T194

To those who are interested in glass it should not be uninteresting to spend a little time tracing in the course of centuries and in many countries the history and development of the various modes of decorating it.

Walter Butterworth

JSGT 1939 V23 T347-T371

Two years ago, or even less, I should have objected strongly to giving a lecture with this title, because I dislike the term "organic glass" I still consider it unfortunate that the word "glass" should be used as a general description for all types of amorphous solids or pseudo-solids, instead of being restricted to solid nanocrystalline inorganic materials. In view of the almost universal acceptance of this meaning nowadays, however, I must content myself with merely recording a protest and fall into line. I remember the term "flexible glass" being used to describe celluloid when it was introduced during my early youth, but the term "organic glass" does not appear to have come prominently into use until about 1924, when it was used to describe "Pollopas", the urea-formaldehyde resin developed by Pollak, which attracted so much attention about that time. The possibilities of this new material as a substitute for glass in particular applications were perhaps somewhat over-rated, partly as a result of the claims made by its inventor, and partly owing to imaginative articles contributed to both technical and non-technical sections of the press. They could not be entirely disregarded by the glass interests, and in 1926 a useful survey of the position from the "glass angle" was made in the Glass Review under the title "Substitutes for Glass and a Newly Advertised Invention." The survey was not confined to a consideration of the properties of Pollopas as compared with glass, but was extended to include mica, rock crystal, cellulose nitrate, and cellulose acetate. At that time there was not much direct evidence of the nature of glass, nor was the nature of the "organic glasses" entirely established; nevertheless the indirect evidence was considerable, and the author of the article was bold enough to draw a parallel between cellulose and silica glass, describing both materials as complexes of simple molecules "clotted" together.

Harry Moore, D.Sc.

JSGT 1924 V08 T268-T277

The object of this paper is to set out as briefly as possible facts that have been discovered about the making of glass in the Leeds area, from the end of the seventeenth century to the middle of last century.

Francis Buckley

JSGT 1944 V28 T105-T112

Types of furnaces are described for the sm92all-scale melting of glasses of low thermal expansion coefficient and requiring very high temperatures to melt them. The types described include two with tungsten grid-heating elements and two small glass-tank furnaces. The first of the tank furnaces was heated by oxy -coal gas with a burner of special design, the body of the furnace being built of zircon blocks. In the second and improved type of small tank furnace, heat was supplied by tungsten-rod electrodes immersed in the glass as well as by a single oxycoal gas burner. The temperature of operation ranged between 1800°C and 2000°C, the upper limit being controlled by the stability of the refractory zircon blocks of which the tank was built.

J. H. Partridge & O. Adams

JSGT 1931 V15 T107-T119

The origin of glass-making in the Duchy of Lorraine is obscure, but it is generally conceded that the craft was introduced there some years before the close of the fourteenth century, the first glassmakers being natives of Bohemia and of Saxony. The date and circumstances of their migration are not known, but there is reason to believe that one of the successors of Ferry IV, Duke of Lorraine, namely, Raoul (1328–1346) or John (1346–1391), induced these foreign glass-makers to settle in the Duchy. Throughout this period friendly relations existed between France, Lorraine, and Bohemia, and the conditions of affairs were not unfavourable to the development of trade and industry. Thus, Raoul continued the policy of his immediate predecessors, founded on a close alliance with France at a, time when John of Luxembourg, the blind King of Bohemia, was living at the French Court. Both these rulers fell fighting as allies of France at the Battle of Crecy.

E. Graham Clark

JSGT 1945 V29 T048-T076

Photoelastic measurements have been made on standard sandwich seals over a wide temperature range by means of the stress-optical bench. Between room temperature and the annealing range for a seal the stress varied considerably and usually changed its sign. The stress at any temperature could be altered considerably by a change in the thermal treatment of the seal. Change of stress with temperature depended mainly on differential free contraction between glass and metal, but in the annealing range of the glass the stress was modified by viscous flow. Where the metal had a low yield point (e.g.copper), the stress in the glass was much lower than that anticipated from differential free contraction. Axial stresses (p₀) were determined for bead seals at room temperature using the method of A. W. Hull & E. E. Burger; and the theoretical relationship between p_zand the ratio of the radii of the glass and metal components was verified by measurement. Hence the determination of axial stresses in seals could be used for rapid routine checks on the thermal expansion properties of glasses. It was also concluded that the permissible value of the stresses in bead seals at room temperature depends on their actual design and application, and in some cases can safely be considerably in excess of 1 kg/mm²tension.

G. D. Redston & J. E. Stanworth

JSGT 1945 V29 T092-T110

It is shown experimentally for a seal of practical dimensions that the stress due to differential expansion of metal and glass approximates that derived by A. W. Hull & E. E. Burger for a very long seal. Further, the stress in a glass-metal seal arising from differential expansion is shown to be considerably modified both by the thermal history of the seal during the manufacture and the operating conditions of the device in which· the seal is used. A procedure is described and illustrated by which a suitable compromise in the dimensions of the glass-metal seal and its heat treatment during manufacture can be determined to ensure absence of extreme stresses at any time during manufacture or subsequently.

R. W. Douglas

JSGT 1937 V21 T310-T349

Machines catering for the production of table glass and similar ware are required for the following purposes:-Cracking-off; Edge-grinding; Re-melting and Glazing; Decorative Grinding: (a) Bottom Grinding.(b) General Decoration; and Acid-Etching. It is proposed to deal with these machines in the above order, describing their principal characteristics, and incorporating a discussion of those features in their design which experience has indicated to be of importance.

E. Venis

JSGT 1949 V33 T089-T094

Grid gas made by the Lurgi process is now produced on a large scale at a plant near Teplice and is piped under pressure to a large number of factories, including glass plants. The purified gas contains approximately 19% CO, 61% H₂ and 17% CH₄.

M. Fanderlik

JSGT 1953 V37 T032-T040

Describes the Hartford type "K" forehearth; the Hartford single- and double-orifice feeder; the Lynch 10, 16B and J.P. machines, and the Hartford IS machine.

T. W. Wardley

JSGT 1927 V11 T037-T039

A conveyor of the continuous type, which is the one most generally used in modern lehrs, has the disadvantage of great weight and, being in continuous travel through great heat down to room temperature, carries away a substantial quantity of heat through the metal from the hot chamber of the lehr; in turn, the conveyor has again to be heated to the temperature of the glassware on returning to the feeding end.

F. A. Hurlbut

JSGT 1936 V20 T695-T716

The properties required in heat resisting materials for medium and high temperature operations are reviewed, with reference to the following: composition, thermal conductivity, changes occurring on heating, microstructure, strength at high temperatures, resistance to scaling, the effects of repeated heating and cooling, and the bearing of structural changes. An account is given of the performance of heat resisting alloys in service. New experimental methods of investigation are described relating to resistance to the effects of repeated heating and cooling, and the attack by molten glass on materials suitable for the tools and equipment used.

R. J. Sarjant & T. H. Middleham

JSGT 1956 V40 T260-T276

It is important to know how the temperature inside a refractory block changes when the block is fired or reheated, particularly in the case of large tank blocks for glass furnaces, but existing methods of calculation are almost impossibly tedious. The methods now described can be applied quickly and accurately to any practical case, and do not require any advanced mathematical knowledge. It is shown that when a slab is heated with a constant rate of surface temperature rise, the temperature lag at the centre of the slab ultimately becomes constant. Starting with this assumption, methods have been obtained for calculating: (1) The time during which a uniform rate of rise of surface temperature must be maintained in order to reach the above steady temperature lag at the centre, and the actual temperature lag achieved in any shorter time. (2) The temperature lag achieved in cases where the rate of rise of surface temperature is not uniform. (3) The 'soaking time' required at the end of a heating cycle, i.e. the time for which the maximum surface temperature must be held in order to allow the temperature lag to be reduced to an acceptable level. Results are given for typical cases, and the implications of these results discussed.

B. E. Moody, A. J. Madgwick & D. G. Monk

JSGT 1929 V13 T167-T182

The object of this paper is to give a brief review of an important metallurgical advance of recent years in the domain of high temperature engineering, and to indicate those properties of the new materials, generally known as heat-resisting steels, which are of special interest to the glass industry.

R. J. Sarjant

JSGT 1942 V26 T005-T021

In designing a recuperator for use at high temperatures metal parts may be employed, but the shape must be simple both from structural considerations and to keep the pressure drop in the gas passages low. The limits of temperature possible are usually set by the materials of construction. In calculating the design, attention is directed to the films at the surfaces of the metal. In the air side the coefficients are reasonably well established, but on the gas side the calculation of the heat transfer is complex. The radiation from the· waste gases is considerable and the rate of heat transfer is increased by the inclusion of brickwork amongst the tubes. This increase amounts to some 40% of the heat transferred, depending on the temperatures involved.

J. B. Wagstaff, S.M., B.Sc.

JSGT 1956 V40 T413-T428

This paper is the first report on a series of researches directed by the author for the production of high-melting glass fibres. Novel methods of operation were developed, and have been described in various Patents (1953 to 1956).

Alexander Silverman

JSGT 1917 V01 T003-T027

THE importance of the study of our resources of raw materials used in the manufacture· of glass needs no emphasis in a paper delivered to a Society largely composed of those directly concerned in the
industry. Despite this realisation, very little work has been done upon the raw materials of the glass-maker's "batch:'" That the investigation into the properties, behaviour, and impurities of our raw materials is receiving due, if tardy, recognition, is indicated by the appointment recently of a strong Standing Committee,
under the chairmanship of Prof. Herbert Jackson, F.R.S., by the Department of Scientific and Industrial Research for the purpose of considering problems dealing with glass and optical instruments.
The Committee has appointed a number of sub-committees, of which the first on the list is formed to deal with the question of raw materials for glass and glass-making.

By P. G. H. BoswELL. D.Se., F.G.S.
Read at the Leeds meeting .January 18th, 1917.)

JSGT 1917 V01 T027-T061

IN the manufacture of the best types of glass it is essential that each of the batch materials should be as free from iron compounds as possible, in order to ensure the production of a colourless glass.
Present-day manufacturing methods have reached so high a pitch of perfection that it can safely be said that batch materials which have to be manufactured can be obtained practically free from
iron. For example, such substances as sodium carbonate, potassium carbonate, potassium nitrate. red lead oxide, borax, precipitated calcium carbonate, all of which are manufactured in bulk for the glass maker, show IittIe or no trace of iron impurity. Unfortunately, this cannot be said of any batch material which is
used in its native mineral form, e.g. sand, limestone. felspar, kaolin, ground quartz.

JSGT 1917 V01 T061-T074

 THE subject of the annealing of glass has always been a matter of great importance for my firm. But my interest in it was until recently solely that of tlle optical glass worker, and I was chiefly
concerned with the conditions for annealing fairly massive pieces. The degree of annealing required for such pieces when they are to be used for the best class of optical work is much higher than is necessary for glassware.

By F TWYMAN

JSGT 1946 V30 T239-T242

The grey mark caused by glass-to-glass impact is shown to consist of a large number of microscopic flaws, the pattern of which is the same on both members subject to the impact. The origin of the flaws is not understood fully, but the result of the impact is a permanent rupture of the surface, due to flaws which may be invisible to the unaided eye.

J. G. Holmes

JSGT 1946 V30 T364-T380

A description is given of modern batch mixing and handling equipment used in the container section of the glass industry in the United States, followed by an account of machines used to deliver mixed batch to the glass-melting furnaces. In American furnaces high rates of melting are maintained, and most of them are fired with natural gas or fuel oil and operate at temperatures near 1550°. In general, the regenerators possess large volume and height, the width of the furnace ports is graded and combustion spaces are large. Improved furnace control equipment is mentioned. The forehearths in use are generally 14 feet or more in length, one type in common use being the Hartford-Empire K-type forehearth. The production speeds for containers are very high, and have been achieved by (l) light-weighting of bottles, and by alterations to bottle design, (2) a generous supply of mould-cooling air at high pressure, (3) internal cooling of containers in the blow mould, (4) double gob feeding, (5) improved forming machines and (6) improved operating technique and attention to maintenance.

T. Wardley

JSGT 1952 V36 T131-T136

The structure and composition of tank-block materials, especially of mullite and zircon–mullite composition, are discussed, and a method for measuring the corrosion of refractories by immersing 1 cm cubes of the material in molten glass is described.

T. S. Busby & J. H. Partridge

JSGT 1922 V06-T220

In gas reversing valves of the ordinary type the valve seating is liable to get warped or burned away, thus allowing a leakage of gas to the stack flues. This burning and warping is due to the current of hot waste gases turning downward over the valve seating on the way to the stack, alternating with the hot producer gas impinging on the sill on its way to the furnace. Any leakage of gas between the valve and its seating further aggravates the distortion or deformation of the sill, thus permitting an increasingly rapid escape of gas.

F. W. Knowles

JSGT_1923 V07 T003-T019

The present paper will deal with such improvements as have taken place since 1917, more particularly with the furnaces which have recently been built in this country to the author's design and patents. Before going into the design of the furnace itself, I should just like to mention briefly a few points not referred to in my previous paper relating to recuperates of the "Hermansen" type. In the author's view, these recuperators possess certain advantages not shared by other systems on the market, for instance, the Radot-Lencauchez type with vertical air channels, and apparently overlooked in the claims made for the same.

Th. Teisen

JSGT 1953 V37 T061-T070

A detailed exposition is given of some common faults in glass-making machinery.

N. A. Nichols

JSGT 1934 V18 T013-T018

The influence on the melting rate, thermal expansion, and resistance to water of the progressive addition of small amounts of boric oxide up to a maximum of 1·5% has been determined in reference to soda–lime–(magnesia)–silica glasses of the type employed for sheet glass. Melting determinations on some fifty glasses show that the rate of melting is definitely accelerated. Based on the examination of twenty glasses it is found that the thermal expansion is, in general, slightly reduced and therefore improved. The improvement in resistance to water is either slight or inappreciable.

Violet Dimbleby, Michael Parkin, Eric Seddon, W. E. S. Turner & Francis Winks

JSGT 1931 V15 T365-T372

In a series of earlier communications, we have recorded the results of measurements on the ultra-violet light transmission of some colourless bottle glasses and on types of glasses specially manufactured for the purpose of light transmission in the ultra-violet region. The second investigation led to a further one, in which a special series of pure synthetic glasses containing ferric oxide in amounts ranging from 0 to 1% was prepared and investigated. It was then announced that a similar series of glasses containing ferrous oxide was under preparation and investigation. The work was, however, interrupted by the removal of one of us from Sheffield; but that part of the study already then finished was referred to in a communication by one of us early in 1929, in which a comparison was made of the limit of transmission in the ultra-violet of glasses containing the same amounts of ferric and ferrous oxide, respectively. Towards the end of 1930 the opportunity was found of completing the work, and a full account is now recorded.

Links to 1928 JSGT papers T027, T307 and T325 and 1931 Glastech. Ber.

David Starkie& W. E. S. Turner

JSGT 1929 V13 T047-T056

 After a survey of the results recorded, we are of the opinion that the most effective salt in producing easy melting and refining, combined with good working properties and least extensive corrosion, is sodium carbonate. Next in order comes borax. All the salts, however, led to very surprising results as compared with glasses prepared either entirely from cullet or from a mixture of cullet with 50% or less of batch. From our observations, indeed, it appeared that all the glasses prepared from cullet with small additions of sodium carbonate, borax, or saltcake mixed with carbon became plain before any glass prepared from a mixture of batch with cullet. This result, in itself, is not only striking, but is also important, both from the practical and the theoretical point of view.

F. W. Hodkin, W. E. S. Turner & F. Winks

JSGT 1955 V39 T173-T192

The use of infra-red reflection measurements for studying the structure of glasses is discussed. The positions of the peaks in the spectral reflection curves indicate the vibrational frequencies of the structural linkages present. Measurements have been made on a number of commercial glasses, fused silica, fused boric oxide and several series of experimental glasses comprising binary silicates, binary borates and sodium borosilicates. Spectral reflection curves for the glasses are presented, and assignments of the observed reflection peaks to structural linkages are made and are listed in tabular form. There is support for the existing hypothesis concerning the co-ordination of boron in sodium borate glasses. In the borosilicate glasses investigated it appears that all the boron can be tetrahedrally co-ordinated, provided that there is sufficient Na₂O and SiO₂ present; B–O–Si linkages are formed in preference to B–O–B linkages; additional oxygen atoms from the Na₂O associate with the boron atoms in preference to the silicon atoms. Attention is drawn to the effects of changes in composition involving the replacement of one metallic oxide by another. Thus, for example, BaO and Na₂O appear to play similar roles as constituents of binary borate glasses, but different roles in barium crown types of glass. PbO and Na₂O play different roles in binary silicate glasses.

P. E. Jellyman & J. P. Procter

JSGT 1954 V38 T421-T424

The infra-red transmission properties of black tellurite glasses containing CuO, MnO₂ or V₂O₅ are shown in the range 0·8–5·0 μm. These glasses, which absorb visible radiation very strongly, can transmit efficiently in the infra-red, in some cases at least to 5 μm.

J. A. James & J. E. Stanworth

JSGT 1954 V38 T383-T420

The basic methods which have been used to study internal friction or specific damping capacity are reviewed. Apparatus for studying the internal friction of thin rods of fused silica, a soda–lime–silica and a borosilicate glass undergoing transverse vibration (0·3–50 c/s) and torsional vibration (1–10 c/s) is described. The peculiar difficulties in obtaining an absolute measure of the internal friction in the case of transverse vibration at low frequencies are discussed. It is shown that there are two physical processes which contribute to the internal friction at these frequencies, a relaxation of alkali ions in the glass and the viscoelastic behaviour of the remainder of the vitreous network. This hypothesis is supported by a clear dependence of the internal friction on the thermal history of the specimen. At constant temperature the internal friction decreases with increasing frequency, but there is no evidence of a maximum value at low frequencies. The variation of internal friction with temperature shows a frequency-dependent peak value, indicating a process with an activation energy similar to that governing the movement of alkali ions in the glass. At higher temperatures the dependence of the internal friction on frequency and thermal history is illustrated, and note is made of the use of the results in an estimation of the distribution of relaxation periods among viscoelastic elements in the network. Incidental results are appended giving the variation of elastic moduli with temperature, and .the possibility of thermo-elastic coupling in glass is discussed.

P. L. Kirby

JSGT 1951 V35 T041-T042

This is a brief note to indicate our concurrence in the view that “Wallner's lines” are an effect of the passage of transverse (“shake”) wave across the head of an advancing fissure. This interpretation was first called to my attention, I believe, by E. F. Poncelet about 1945, when he was working at this laboratory on these phenomena. The details have been elegantly worked out by Prof. Adolf Smekal, who showed us much of his work on his visit here in 1949, and some of it has since been published in the Glastechnische Berichte.

F. W. Preston

JSGT 1953 V37 T027-T031

An introductory review of the early development of the blow-blow bottle-making machine and of the flow feeder.

F. W. Hodkin

JSGT 1959 V43 T030-T036

The analytical uses of ethylenediamine tetra-acetic acid (EDTA) are reviewed. These uses arise from the ability of EDTA to form highly stable complex compounds with almost all metals. The main analytical use of EDTA is 'complex metric titration' in which a metal is determined by titration with EDTA solution; more than forty metals can be determined in this manner. Choice of pH of solution, or use of selective precipitating or complexing agents, often permits titration of one metal in a mixture of metals. Such methods are rapid where they avoid need for separations and where they can replace gravimetric methods. In glass analysis complexometric titration of Ca, Mg, Ba, Al, Zn, and Pb are of proved or potential value. EDTA has also been used extensively as a masking agent to improve the selectivity of various analytical reactions by inhibiting interfering metals. Examples quoted include determination of Mo, W, U, Be, Ca, Ba, K, Cu, sulphate, and borate. Several such applications are relevant to problems in glass analysis. The third main analytical use of EDTA relates to its effect on oxidation–reduction equilibria, principally in polarography.

H. J. Cluley

JSGT 1934 V18 T201-T208

Employing a rotating cylinder viscometer the viscosities of Na₂B₄O₇–NaH₂PO₄ mixtures were determined from 527° to 920°. The viscosity of fused borax ranged from 3·30 poises at 880°to 83,750 at 590°, and for a mixture containing 25% Na₂B₄O₇, 75%. NaH₂PO₄ the values were 15·7 poises at 732° and 38,370 at 532°. There was no indication of the formation of any intermediate complex compounds.

M. Volarovich

JSGT 1958 V42 T088-T100

The glass flow in a number of tank furnaces has been investigated by adding a very small amount of radioactive tracer to the batch, and measuring the amount of radioactivity in the glass containers produced. Details are given of the methods used for preparing and adding the tracer, consisting of a glass containing active phosphorus, and for measuring the activity of the glass samples. Some typical results are discussed.

B. E. Moody & D. G. Monk

JSGT 1931 V15 T128-T140

Much useful and interesting information concerning the flow of glass in tanks has been obtained previously by investigations carried out with the aid of deep blue glass or grains of a nearly insoluble oxide, such as chromium oxide, but such methods are subject to many disadvantages. For instance, the introduction of blue glass or of insoluble grains into colourless glass ruins the glass for commercial purposes, and therefore, for economic reasons,  investigations employing those indicating media can only be carried out when a tank furnace (usually in a very worn condition) is about to be put out, and the experiment has to be stopped at, or shortly after, the initial appearance of the blue glass or black specks at the fabricating machines; thus the opportunities for such investigations are strictly limited, and usually the information gained can only be claimed to apply strictly to worn tanks and to a comparatively short time interval. Further, investigations on dark glass tank furnaces cannot be made easily and precisely with the aid of blue glass or insoluble grains.

E. J. C. Bowmaker & J. D. Cauwood

JSGT 1927 V11 T039-T052

 The method of forced crystallisation makes it possible to crystallise every glass and to determine the temperature of the beginning of crystallisation as well as the melting point of the newly-formed crystals. In the glass under investigation a constant lowering of temperature is maintained so that the temperature is reached at which the number of spontaneously forming nuclei reaches its maximum value. This value can be very small, as, for instance, in my study of acid borates of sodium in mixtures approaching pure B₂O₃, which I was obliged to maintain at such temperature conditions for three weeks uninterruptedly until the crystals began to form.

J. F. Ponomareff

JSGT 1936 V20 T144-T151

Comments are made on the results of recent determinations made by Heidtkamp & Endell of the density and viscosity at high temperatures of the soda–silica glasses. The likelihood of changes of curvature in the temperature–viscosity–composition curves being indicative of the presence of the compounds Na₂O.2SiO₂ and Na₂O.SiO₂ is discussed. The form of the density–composition curves at different temperatures is also abnormal in the same sense as those of viscosity and probably for the same reasons. Comparison with other work only emphasises these differences. The evidence afforded by the study of glass-making reactions, and the volatilisation of alkali from a similar series of glasses in connection with this problem is pointed out, and also the obvious chemical changes taking place in coloured glasses. The compounds Na₂O.SiO₂ and Na₂O.2SiO₂ known to be stable at their melting points, should not be denied existence in the liquid state without experimental proof, and any theory of the constitution of glasses should not ignore the elementary facts of chemical combination.

Eric Preston & W. E. S. Turner

JSGT 1932 V16 T145-T170

 The description is given of an apparatus used to determine the tensile strength of glass specimens of approximately 1 mm diameter, the specimens having constrictions at their centres. These constrictions were formed in an electrically heated tube at a determined temperature. It was found that the type of fracture obtained, the thickness of the specimen, acid polishing the specimens and matt etching the specimens considerably affected the tensile strength values, but that the influence of different loading methods, the temperature of constriction formation, the presence of strain, and the influence of sulphurous fumes during annealing were very small, if existent. Experiments were made to attempt to correlate tensile strength and thermal endurance results. The conclusion was drawn that the presence of surface flaws in glassware, as proposed by A. A. Griffith, would explain the results obtained.

E. J. Gooding

JSGT 1917 V01 T074-T087

GLASS furnaces can be divided into two main classes, pot furnaces and tank furnaces. I should here like to make a few remarks regarding the first class, namely, pot furnaces, as these furnaces are mostly used for high-class glass.

JSGT 1936 V20 T016-T018

The sites of Jean Carré’s four glass furnaces in "Fernefol" have been identified, and the foundations of one of them, recently excavated, are here described. The furnace, 20 ft 10 in long by 2 ft 5 in wide, had walls 1 ft 11 in thick. A few glass objects of skilled technique were found.

S. E. Winbolt

JSGT 1923 V07 T263-T276

J'ai suivi avec le plus grand interet le journal de votre societe depuis sa creation. J'ai pu voir l'importance que vous attachez aux recherches scientifiques et la confiance que vous leur accordez pour le perfectionnement de votre industrie. Vous ne serez donc pas surpris que j'aie choisi pour cette conference l'etude d'un des points de contact de la Science et de l'industrie.

Henry Le Chatelier

JSGT 1947 V31 T153-T169

Barium and lanthanum oxides decrease the reciprocal relative dispersion (ν-value) comparatively less than other constituents which impart equal increases of refractive index to high index, low dispersion crowns. Quantitative data were obtained when 7·5 mol% of various oxides were introduced into the simple base glass 27·5 BaO, 15 B₂O₃, 7·5 Al₂O₃, 50 SiO₂. The chemical resistance is impaired by increase of the BaO as well as by La₂O₃, Small amounts of Al₂O₃, ZrO₂, ThO₂, and particularly of TiO₂, Nb₂O₅ and Ta₂O₅, increase the chemical resistance of barium and lanthanum glasses more than SiO₂. In a large area of the field BaO–La₂O₃–B₂O₃, formation of glasses having favourable index-dispersion relations was observed, particularly near the narrow zone of lanthanum borates containing little or no barium. Lanthanum borate, modified by barium, may be considered the basis of modern complex non-silica and low silica glasses of high index and low dispersion. Lanthanum oxide was found to form glasses with phosphoric acid and barium oxide. A pure lanthanum metaphosphate glass exists and has a refractive index of 1·60 and a reciprocal relative dispersion of 60. Data obtained in the borosilicate, borate and phosphate system permit the discussion of the application of barium and lanthanum in conjunction with other more or less frequently-used elements in the construction of high index, low dispersion crowns.

Gordon F. Brewster, Norbert J. Kreidl & Tyler G. Pett

JSGT 1939 V23 T076-T081

 Evidence is given, from X-ray analysis, for concluding that the opacifying substance in a lead glass containing arsenic is lead arsenate, Pb₃(AsO₄)₂, although no direct comparison with a compound of this composition was made. The replacement of arsenic with phosphorus produces an opal glass in which the opacifying substance is lead phosphate, Pb₃(PO₄)₂ In this case, the lead phosphate was identified directly. The crystal structures of the lead arsenate in the opal glass and lead phosphate, and their relations to the structures of some members of the apatite group of minerals are briefly  discussed, dimensions of the unit hexagonal cells being determined. For lead arsenate a₀ = 10.02 A.U., c₀ = 7.37 A.U. The formation of solid solutions between lead phosphate and lead chloro-phosphate is described.

H. P. Rooksby, B.Sc., F.Inst.P.

JSGT 1950 V34 T005-T009

In an earlier communication describing sealing glasses, brief reference was made to a glass (C76) suitable for sealing to iron, and to its use in the manufacture of glass-metal inspection windows. J. H. Partridge in his monograph on glass-to-metal seals assumes, not incorrectly, that C76 is of the same type as the glass No. 542 described by A. W. Hull, E. E. Burger & L. Navias. The latter glass, however, contains 30% lead oxide, whereas C76 is entirely free from lead, a difference of some interest, and one which has led to some useful applications.

J. E. Stanworth

JSGT 1951 V35 T043-T057

This investigation was undertaken to obtain evidence relating to the ways in which various “network-modifiers” are held in silicate glasses. Continuity of structure is essential in any glass, but in an alkali–silica glass the presence of the alkali causes discontinuities in the structure wherever systems such as ≡Si–O–Na or, as is more probable, ≡Si–O– and a free Na⁺ ion occur. The limiting compositions of alkali-silica glasses should thus correspond with Na₂O.SiO₂, K₂O.SiO₂, etc. The introduction of the oxides of divalent and trivalent elements could, on the other hand, provide “bridge” linkages of the form ≡Si–O–Ca–O–Si≡ or ≡Si–O–Al–(Si≡)₂, which would preserve continuity even if the glasses contained the limiting proportions of the alkalis. In the experimental work a glass was regarded as stable if a sample weighing about 10 g remained clear when cast and allowed to cool in air sufficiently slowly to prevent shattering. As judged by this test the limiting compositions of soda–silica and potash–silica glasses were found to be 0·9Na₂O.SiO₂ and K₂O.SiO₂, respectively. The addition of the oxides of divalent and trivalent metals to glasses approximating to the above limiting compositions enabled stable glasses to be made containing RO and R₂O₃ in considerable amounts. It is concluded, therefore, that oxides of the types RO and R₂O₃ act as strong “bridging” links in the glass structure, by the formation of strong ionic or covalent bonds between the metallic atom and the oxygen atoms at the corners of adjoining silica tetrahedra. The introduction of lead oxide enables stable glasses of compositions K₂O.2PbO.SiO₂ to be obtained. This requires either the formation of “double-span” bridges ≡Si–O–Pb–O–Pb–O–Si≡ in the structure, or the formation of PbO₄ tetrahedra. In either case the lead is present in such high proportions that it must play a large part in building up the glass structure.

H. Moore & M. Carey

JSGT 1956 V40 T252-T259

The region of glass formation was determined in the ternary system ZnO–B₂O₃–V₂O₅ by using a micro-scale glass-melting technique. Thermal expansion measurements were made on a number of glasses in the system, when it was found that the effect of adding up to 20 wt% of V₂O₅ to zinc borate glasses results in a considerable lowering of the deformation temperature, Mg, without appreciably affecting the expansion coefficient, which lies in the range 45–50×10⁻⁷ • One glass composition was selected and it was shown that, with small modifications in its composition to improve its durability, this glass could be used as a solder glass for Fernico sealing glasses. The temperature at which the solder glass formed vacuum-tight joints was 625°C.

E. P. Denton & H. Rawson

JSGT_1919_V03_T275_T277 (3 pages)

Magnesia glasses do not as yet find a place in commercial manufacture in the same sense as lime glasses; that is to say, three component glasses of magnesia, soda, and silica have no place corresponding with those of lime, soda, and silica. Since, however, magnesia is naturally found associated with lime, there are few lime glasses which do not contain some very small quantity of it at least. As a sequel to our investigation of the physical properties of the lime-soda series, we have undertaken the examination of corresponding magnesia-soda glasses. The general molecular type was, as before, that of the tri-silicate glasses, and the range from 6SiO₂.Na₂O to 6SiO₂.0·9Na₂O, 1·1MgO was successfully prepared and tested.

J. H. Davidson, F. W. Hodkin & W. E. S. Turner

JSGT 1951 V35 T058-T092

A study has been made of the magnetic properties of two series of glasses containing iron, with the object of identifying the state or states in which the iron was present. The glasses of Series I were soda–lime–silica glasses made under strongly reducing conditions so as to ensure, as far as possible, that the whole of the iron was present as ferrous iron. Series II consisted of potash–lead oxide–silica glasses melted under conditions designed to maintain the whole of the iron in the fully ferric state, sub-series II(a) being made on a base-composition: PbO 70%,

K₂O 5%, SiO₂ 25%; and sub-series II(b) on a base-composition: PbO 70%, K₂O 10%, SiO₂ 20%. From the magnetic susceptibilities of the glasses of Series I the mass susceptibility of the iron at 25° was determined as approximately 220×10^-6, which corresponds very closely with the mass susceptibility of iron in the ferrous condition. The mass susceptibility of the iron in the glasses of Series II ranged from 210×10^-6 to 350×10^-6, the variation being associated with a marked increase in the intensity of the ferric colour. In one glass, the colour of which showed a large “grey” component, the mass susceptibility of the iron was determined as 369×10^-6. These values indicate that the mass susceptibility of the colouring ferric iron is greater than 300 and probably greater than 350×10^-6, and that the iron contributing the grey colour has a mass susceptibility at least as large as that of the colouring ferric iron, probably larger. They also suggest that the "colourless" iron has a mass susceptibility less than 200×10^-6. The conclusions drawn are that iron can exist in glasses as: (1) ferrous iron, producing a blue colour; (2) ferric iron in the form of colloidal ferric oxide, giving a yellow or amber colour, (3) ferroso-ferric iron, in the form of colloidal ferroso-ferric oxide, which contributes a grey colour; and (4) iron which is linked into the glass structure in a form which contributes no colour. The low mass susceptibility of the colourless iron suggests that the iron in this form is linked into the structure by four bonds, probably as FeO₄ tetrahedra. This cannot, however, be regarded as definitely established without additional experimental evidence.

H. Moore & S. Kumar

JSGT_1936_V20_T200_T217

The manufacture is described of crucibles, rods and tubing from the pure refractory oxides Al₂O₃, MgO, BeO, ThO₂, and ZrO₂, and a general account given of the production of highly refractory articles by fusion casting, slip casting, and by the use of temporary bonding agents. Spacers of high electrical resistance for wireless valves are produced by moulding Al₂O₃ and MgO under pressure, to tolerances of ±0·5%. Wet paste and slip casting processes are employed for the production of crucibles. For the extrusion of rod or tubing a plasticising agent of flour containing NH₄OH is used for alumina, and a cellulose composition for magnesia. The particle size must be graded to suit the diameter of the extruded shape. The extent of the firing shrinkage is controlled by particle size grading, the ratio of plasticiser to material, and by the use of material pre-calcined at different temperatures. Data are given on the chemical, thermal, and electrical properties of the materials produced.

J.H. Partridge & J.R. Lait

JSGT 1919 V03 T014-T026 (13 pages)

To be perfectly frank, when I attempted to introduce my experience of costing in order to obtain results in a glass factory, I was confronted with problems which at :6.rst sight appeared to me to be absolutely without solution, and it is only by the unstinted assistance that I have received from the managing and technical staffs of the various factories that I gradually obtained a clear view of most of these solutions and I am convinced that it is up to the accountant and the technical expert to put their heads together before any progress can be made in obtaining a suitable costing system for a glass factory.

J. Dearden Mills

JSGT 1956 V40 T032-T038