INDUSTRIAL A N D ENGINEERING CHEMISTRY
1164
pieces show marked evidence of plastic deformation and a tough break which seems to go through the crystals rather than between them, as was the case in Figure 3. There is some, but much less marked, evidence of coarsening of grain in these steels. A microscopic examination of all these specimens in both Figures 3 and 4 shows that marked grain growth does occur in both types of steel, but it is very much more marked in the steels in Figure 3 than in Figure 4. Rezistal 7 and Stainless 24, as was observed, were above 25 per cent in chromium and, so far as impact resistance is concerned, the former shows some of the characteristics of the latter in the as-rolled condition and after 870" C. (1600" F.). In these two conditions Rezistal 7 shows, respectively, 4.10 and 6.0 foot-pounds (0.566 and 0.828 kilogram-meter) in the Charpy tests. Even this, however, is considerably higher than Stainless 24, which shows but 1.8 foot-pounds (0.248 kilogram-meter) after 870" C. (1600" F.), and at the higher heats is slightly diminished, while the austenitic steel, No. 7, a t the higher temperatures, showed a very great improvement and the Charpy test gave, respectively, 40.7 foot-pounds (5.62 kilogram-meters) after 980" C. (1800' F.), 52.6 foot-pounds (7.26 kilogram-meters) after 1090" C. (2000" F.), and 54.3 foot-pounds (7.49 kilogram-meter) after 1200' C. (2200" F.). Rezistals 2 and 4 were still higher in impact resistance after being subjected to high temperatures, while for the chromium and chromium-silicon steels a t 870" C. (1600" F.) or higher the maximum Charpy impact test showed 5.8 foot-pounds (0.80 kilogram-meter) and the minimum showed 1.4 foot-pounds (0.19 kilogrammeter). This very marked difference, particularly in the effect of temperature on the chromium and chromium-silicon steels versus the austenitic steels, is very remarkable and demonstrates why the latter are superior for applications involving high temperatures, particularly if there are any shock or impact conditions to be met, either intentionally or accidentally. Where only resistance t o scaling is concerned and high physical requirements are not necessary, Stainless 24 is very good. Under other conditions, however, the tougher steels, such as the austenitic types, are to be preferred as a matter of precaution. In case of exposure to products of combustion from high-sulfur fuels, such as lowgrade coal, those steels which are very high in nickel are liable to fail rapidly, but if the chromium is much higher than the nickel content there is still the possibility of pre-
Vol. 21, No. 12
serving the austenitic characteristics by the addition of a small amount of nickel to a high-chromium alloy and a t the same time producing a satisfactory steel for sulfurous conditions. Future of Alloy Steels
As t o the future of alloy steels for both ordinary corrosion and high temperature conditions, little imagination is required to foresee endless applications for products possessing such unusual properties. The austenitic heat-resistant steels have been used very successfully in furnace and retort work involving high temperatures for long periods of time. Welded carburizing boxes made from sheets or plates are gaining rapidly in popularity as compared with the much heavier cast boxes, which are not only more difficult to handle in the shop but also require a definite increase in the amount of fuel and time in order to carry out these carburizing operations. Boxes made of welded plates offer an economy in weight, time, and fuel. The same has been found true of annealing boxes for bright-annealing steel sheets and strips. This material has also been used in the form of endless belts, particularly in furnaces for continuous heat-treating operations. Endless belts built up of sheets and traveling over large pulleys sarry the work from the low-temperature end of the furnace up to the higher-temperature end. In other heat-treating furnaces the work is carried directly through the furnace on the surface of parallel driven rollers which constitute the hearth of the furnace. For this application strength, under load, and resistance to scaling are of great importance; while in the oil-cracking industry and other operations in chemical manufacturing, the same materials have been very successfully employed to withstand, not only chemical corrosion, but resistance to heat and scaling. While such applications have been very numerous, the writer is convinced that industry has only commenced to use these materials and in the very near future their employment will become much more general. Literature Cited (1) Evans, Trans. A m . Znst. Mtning M e t . Ens., Institute of Metals Div., 1929, p. 7. (2) Mathews, Zbid., 71, 568 (1935). (3) Pfeil, J . Zron Steel Znst. (London), 119,501 (1929). (4) Vernon, "Bibliography of Metallic Corrosion," E. Arnold & Co.,London, 192s.
Effect of Atmospheres on the Heat Treatment of Metals' E. G. de Coriolis and R. J. Cowan SURFACE COMBUSTION COMPANY, 2375 DORRST., TOLEDO, OHIO
M
ETALS are heat-treated for the purpose of effecting B change in their structure, their shape, or their composition. Some heat-treating operations may involve two of these changes simultaneously. It is not the purpose of this paper to deal with these changes per se, for they lie in the realm of the metallurgist. Incidental to them, however, are questions which deeply concern the chemist. As the name implies, heat treatment involves subjecting the metal to the effect of a thermal gradient, which in turn implies the derivation of heat in some form. As usually en1 Received August 10, 1929. Presented before the Division of Gas and Fuel Chemistry at the 78th Meeting of the American Chemical Society, Minneapolis, Minn., September 9 to 13, 1929, on behalf of the American Gas Association.
countered in the arts, the source may be the combustion of a fuel, in which case heat is imparted to the metal by convection and radiation, or radiation alone may be applied, as when a muffle is intercepted between the metal and the burning fuel or an electric current is utilized through a resistor, resulting in the conversion of electric to radiant heat energy. I n every instance some of the heat is imparted to the work by conduction, but this is not important in this analysis. Again the metal to be treated may itself be used as the resistor, resulting in the most direct possible application of heat. Finally, the metal may be placed within the field of a high-frequency current wherein by inductance energy is imparted t o the metal with a resultant thermal rise. The last two cases-via., resistance and inductance-make
December, 1929
INDUSTRIAL A N D ENGINEERING CHEMISTRY
1165
possible an application of heat wherein the metal under treat- atmosphere will contain, besides, free absorbable oxygen. A ment may be so confined as to reduce to a negligible factor the reducing atmosphere on partial analysis will contain no free effects of a surrounding atmosphere. In all other cases, how- oxygen but some free carbon monoxide as determined by the ever, there is present within the vessel or furnace an atmos- usual Orsat. However, as it results from the incomplete phere of some kind, the effects of which play an important part combustion of the fuel, some traces of sulfide or organic sulfur upon the particular process involved. This is specifically will be present together with some free hydrogen and partially decomposed hydrocarbons. In the case of a sulfur-free fuel, the subject matter of this paper. Lest there should be some thought as to the possible solu- such as natural gas, sulfur in any form is of course eliminated. The term “neutral,” however, originally implied that a tion of such atmosphere problems through the simple expedient of resorting to resistance and inductance methods, let it be piece of metal, particularly steel, heat-treated in such atsaid that such applications, for reasons that need not be dis- mosphere would remain unaffected as to its surface, and the cussed here, are limited by many practical considerations and term “reducing” implied that this particular atmosphere play but a very negligible part in the whole field of heat treat- was capable of reducing iron oxide or scale from the surface of the metal. Actually a neument. tral atmosphere is h i g h l y Notwithstanding the conoxidizing to steel at elevated siderable strides made This paper deals with the effects of various atmostemperatures and the comwithin the past ten years by pheres on metals during typical heating operations monly accepted reducing atthe application of the elecranging from 3.50” to 2500” F. These effects are shown mospheres are but comparatroradiant furnace to this to be essentially of a chemical nature. tively less oxidizing. The field, its usefulness is sharply Methods for the application of heat are duly conexplanation is that carbon limited by adverse economic sidered. The limitations of each are determined by d i o x i d e and water vapor, factors which no known dethe character of the atmosphere surrounding the work. both of which are present v e l o p m e n t s are likely to This has emphasized the necessity for a controlled in neutral and reducing atalter within the immediate atmosphere. The old designations of neutral, oxidizing, mospheres, are reactive to future. Were the enormous and reducing atmospheres have ceased to have a definite steel a t elevated temperatonnage of steel now daily meaning in the light of more advanced knowledge. tures, and the higher the treated d e p e n d e n t u p o n Some account is given of the research work of the temperature the greater is electrothermal e n e r g y , i t American Gas Association in connection with forging. this reactivity (1,4). Thus would r e q u i r e more than The matter of burning of steel is considered and some it is that a piece of steel doubling the total output of new facts are given. A graph is presented showing the heated to forging temperaall central stations a t a cost time-temperature relationship between carbon monture is invariably covered by so stupendous as to be welloxide and scale formation at forging temperatures of a layer of scale. A scalenigh inconceivable; and in steel. free forging is as yet unturn this would reflect itself The problem of bright-annealing is considered. This known in industry. on the cost of the finished is affected by gases evolved when a metal is heated. These atmosphere effects product in a measure not The use of steam for the bright-annealing of copper is permissible under p r e s e n t are of tremendous consediscussed. The effects of lubricants found on metal economic conditions. It quence to the whole field of surfaces present problems in relation to the specific will be pointed out later that metal treatment. For years atmospheres in bright-annealing. Ways have been they have been accepted as the electroradiant f u r n a c e indicated for overcoming these problems. offers of itself no solution to the natural concomitants of the steps involved in prothe atmosphere problems of duction, with but crude atheat treatment. In order more clearly to define the subject matter in hand, tempts a t mitigating the economic losses thus incurred. As let it be said that it covers all operations from the low-tem- steel scaled in process it was subsequently pickled. Along perature drawing of steel a t 350” F. through the range of an- with the scale was dissolved an appreciable percentage of the nealing, hardening, carburizing, normalizing, and high-tem- metal. Pickling losses of 5 per cent of metal treated are perature forging a t 2500” F. Lying between these extremes still common, and the acidulous pickle waste is still polluting are the annealing and forging of non-ferrous metals. All these streams in most industrial centers. In addition there enters operations are now being conducted to a preponderant degree the human element. The pickling department is to the steel industry what the fertilizer department is to the packing inin fuel-fired equipment. The increasingly exacting requirements of industry have dustry. None but the lower-most strata of labor will accept resulted in largely eliminating coal as a fuel for such opera- employment therein. The forward march of the automobile tions. Where economic conditions permit, gas is the pre- and bathtub are driving such industrial operations to ecoferred fuel, with oil as the next choice where a compromise nomic extinction. The pickling room must eventually be dismust be made between quality and cost. In either case there placed. What is the solution? is presented the factor of atmosphere resulting from the comThe problem is here presented as a challenge to the chemical bustion of the gas or gasified fuel and its effect upon the metal fraternity. Engrossed in the production of dyes, lacquers, and a host of synthetic substitutes, in the reclamation of under treatment. wastes on the farm and in the factory, it is surprising how little Atmospheres of Combustion attention chemists as a whole have devoted to this, one of It is pertinent a t this stage to consider briefly the com- the oldest waste problems of industry. Possibly this lack position of gaseous products of combustion. The old designa- of interest results from a misconception that the question was tions of neutral, oxidizing, and reducing atmospheres have purely a metallurgical one. Although knowledge of metalceased to have a definite meaning in the light of more ad- lurgy is necessarily implied, it must seem evident that a vanced knowledge. A neutral atmosphere merely implies proper solution can be arrived a t only through the application that upon analysis a t room temperature there are present of chemical reasoning and procedure. nitrogen, carbon dioxide, aqueous vapor, and sulfur dioxide Familiar as chemists are with the term “research” and its with no free oxygen or free carbon monoxide. An oxidizing implication of organized effort directed to the solution of a
INDUSTRIAL A N D ENGINEERING CHEMISTRY
1166
problem, i t may be surprising that this particular and seemingly orphaned problem should eventually find an apparently unrelated industry fostering the necessary organized effort toward its solution. The gas industry through its organic body, the American Gas Association, is now engaged in actively pursuing research on this subject. The realization that gas is the one fuel eventually capable of providing the means for overcoming these operating difficulties has provided the impetus for this research. I t is not possible a t this time to bring forth results of definite achievement. Some facts, however, may serve to elucidate the problem and indicate possible lines of further attack. Forging
Vol. 21, No. 12
It will be noted that the temperature of operation is 2100” F. I n practical forging operation it would be difficult and costly to achieve higher temperatures by the use of a muffle. On the other hand, it is not feasible to burn city gas while maintaining such a high percentage of carbon monoxide and secure the desired temperatures. These conditions impose requirements of furnace construction not easy to achieve. However, it is believed that for the first time these conditions have been defined and the objective is now clearly in view. A practical solution would be of considerable import to the whole forging industry. Hardening and Tempering
Whereas the scaling of steel at forging heats involves mechanical difficulties of operation which it would be desirable to eliminate, it is nevertheless possible to produce good forgMichigan has disclosed that, in addition to the usual scaling ings under present conditions because the affected outer surat these high temperatures, the composition of the atmosphere face is usually removed by grinding or machining in the prochas a definite effect upon the burning of steel. The term ess of converting the rough forging into a finished article. “burning” is used to define such steel which, after being strucA machined forging, however, is now seldom incorporated turally damaged a t these elevated temperatures, cannot be re- into mechanical equipment without subsequent heat treatstored to normal structure by any subsequent heat treatment. ment. This may involve carburizing or case-hardening followed by subsequent reheating,.quenching, and drawing. Atmosphere effects in the carburizing process are not of consequence in so far as the metal surface is concerned, because the process itself necessitates the presence of highly reducing atmospheres containing an excess of free carbon. A discussion of these atmospheres and their effects upon the steel structure would of themselves constitute the subject of a lengthy paper. Suffice to say that unburned city gas as such is frequently used as the carbon-carrying agent surrounding the steel within the retort in which the process is carried out. The reheating of a machined and carburized steel forging does present some problems in that scaling of the piece a t this stage seriously affects the internal structure as well as the ultimate dimensions of the finished part. Many such 1 1 I I I parts must be made to fit into a mechanical assembly within IO 30 4 0 2o MIN. close limits of tolerance and must maintain these tolerances As reported by Jominy, a low-carbon steel, when heated in as determined by the previous machining operation. I n a an atmosphere containing 6 per cent carbon monoxide, can gas-fired reheating furnace it is the universal practice to mainbe heated to a temperature 100” F. higher without burning tain the atmosphere on the reducing side, and in the majority than a similar steel heated in an atmosphere containing free of cases the very slight oxidation which takes place on the oxygen. It is the practice in most forge shops to heat steel surface is not sufficient to affect the tolerances of the finished to the highest possible temperature to facilitate the subse- piece. Where the carburizing operation has been so conducted that quent hammering operation. The atmosphere in an electrically heated forge is highly oxidizing. In a gas-fired forge the carbon content of the case is close to the eutectoid point, the atmosphere can be maintained as desired to a 6 per cent the usual reducing atmosphere of the reheating furnace may carbon monoxide content. The possibility of such applica- not prove sufficient to prevent slight decarburization with tion from the standpoint of insuring structural strength of resultant softening of the case. Greater precautions must then be taken. It has been found possible to harden carthe forging marks an advance heretofore unrecognized. As previously set forth, however, even as high a content burized forgings perfectly by reheating in a gas-fired muffle as 6 per cent carbon monoxide at forging temperatures does not furnace in which a slow current of unburned city gas is caused eliminate oxidation of the metal surface. Such oxidation to flow through the muffle. The atmosphere is then similar to that in which the carburizing process itself was conducted, or scaling is highly undesirable. Conjointly with the work undertaken a t the University of and therefore no decarburizing can occur, nor can the piece be Michigan, the Surface Combustion Company in their labora- scaled. The tempering or drawing operation which completes the tories in Toledo are cooperating, on behalf of the American Gas Association, in working out practical applications to these heat treatment of the forging is conducted a t temperatures forging problems. I n these laboratories it has been deter- sufficiently low that the atmosphere effects are not a factor mined that the non-scaling of steel is possible if the atmos- either as to the surface or structure of the finished piece. phere can be maintained a t a still higher carbon monoxide Taken as a whole, therefore, the heat-treating steps in precontent. A muffle furnace was used to heat the steel and paring a machined forging for finished use do not present a partially burned atmosphere of city gas was passed through serious atmosphere problems which cannot be met by present the muffle with the results shown in the accompanying graph. equipment and methods. The composition of Toledo city gas is as follows: Annealing
A comprehensive study conducted on behalf of the American Gas Association by W. E. Jominy at the University of
Methane Ethylene Hydrogen Carbon monoxide
0.331 0.022 0.478 0.076
Carbon dioxide Oxygen Nitrogen Propylene
0.025 0.004 0.051 0.013
When mechanical force is applied to metals in the process of altering their shape, strains are set up in the structure of
December, 1929
INDUSTRIAL AND ENGINEERING CHEMISTRY
the metal which it is desirable or necessary to remove before the metal can be further worked. The usual procedure is to heat it up to a certain temperature varying with its nature and composition, followed by slow, gradual, or rapid cooling. “Annealing” is the generic term applied to all operations of this type. The temperature range of annealing operations varies from 700” F. for non-ferrous metals to 2100’ F. for certain hightemperature alloys. Almost invariably the metals so treated, if exposed t o air, will suffer some oxidation of their surfaces. Subsequent mechanical operations require that the annealed metal be free from surface oxide. The well-nigh universal corrective is pickling. Industry is becoming increasingly insistent that the corrective be incorporated as a part of the annealing process and that metals be “bright-annealed.” This imposes requirements on a heating process which, it can be readily appreciated, are extremely difficult to meet. Considerable work has been done by industry itself with a view to achieving this end, but apparently no concerted effort had been made to view these operations as a whole and to study the fundamentals involved until the inception of the research activity fostered by the American Gas Association. Perhaps the most common misconception in relation to bright-annealing is that, oxygen being the offending gas, any means of excluding its presence would of itself insure freedom from oxidation. Of these the simplest and the one most frequently employed is to pack the metal in a container and fill the voids therein with an inert substance. This is known as “pack-annealing.” A very large tonnage of steel sheet and strip is so treated. The results to date have been fairly satisfactory. The process is costly, however, and improved methods are eagerly being sought. Similar treatment applied to non-ferrous metals, particularly yellow brass, has never proved successful. There is a t present no “bright-annealed” brass being produced in industry. A better understanding of the chemistry of metals will reveal why this is so. Effects of Atmospheres
The most readily available inert substance with which to fill an annealing muffle or pack is pure nitrogen. With yellow brass in particular there is no known reaction between its constituent metals and pure nitrogen a t the usual annealing temperatures, 700” to 1300” F. Yet repeated attempts to anneal yellow brass in an atmosphere of nitrogen have fallen short of expectations. It must be stressed that these experiments were conducted with commercially pure nitrogen further purified by passing through oxygen-absorbing chemicals and over heated metallic copper so that the resultant gas was oxygen-free as determined by ordinary chemical analysis. A film of observable zinc oxide invariably was formed when such purified nitrogen was passed over the heated brass. If the nitrogen, after passing over the heated metal, was caused to bubble through caustic potash solution, the weight of the potash bulb was found to have increased. A search through the literature reveals that considerable work has been done on the occlusion of gases by metals. A general discussion conducted by the Faraday Society (3) lists over seventy-three papers dealing with this matter. A great many subsequent studies have been made. I n 1927 the French chemists Guillet and Roux (3) determined that yellow brass when heated would liberate carbon dioxide, carbon inonoxide, and oxygen. Whether these gases were occluded within the brass or were in actual chemical combination with the metal was not revealed. More recent work by the Bureau of Standards has shown that yellow brass is apparently not the only metal capable of containing or retaining gases within the body of the metal. It was found that steel subjected to prolonged heating under
1167
vacuum gave off gases which in this case were combustible. This led one facetious newspaper correspondent to express the hope that before long devices might be worked out t o propel an automobile from the very gases contained within its metallic frame. Be that as it may, the important facts are that metals do contain gases which are liberated under heat and that these gases are capable of affecting the metal surface, thus eliminating from consideration any process dependent solely upon annealing in a so-called “inert” atmosphere. It would appear obvious that, if oxidizing gases are given off by the metal, the way to compensate for this is to fill the reaction vessel with a reducing gas. Some progress has been made along these lines. For instance, nickel is now successfully annealed in an atmosphere of pure hydrogen. The preparation of this gas is expensive. Besides, its high explodability introduces an element of danger to the operation. A cheaper source of hydrogen would be blue gas or water gas. For some metals this might prove suitable, but not for nickel, owing to its reactivity with carbon monoxide contained in the gas. A highly reducing gas, such as hydrogen, is not, however, suitable for copper and its alloys. Although very effective in protecting the surface from oxidation, its absorption within the metallic structure under heat gives curious effects. T o understand these it is necessary t o consider briefly some steps in the process of manufacturing copper. The purest available commercial copper is that deposited by electrolysis. Electrolytic copper, however, is not sold as such for fabrication into articles of copper. The deposited cathode metal must first be melted down for casting into bars of suitable shape and size. During the melting process some oxygen is absorbed, and this is reduced as far as possible by “poling” the molten bath with wooden poles. Nevertheless, the cast copper contains a slight amount of oxygen which appears as a copper-copper oxide eutectic. When annealed in an atmosphere of hydrogen, this eutectic is reduced with the formation of water vapor, which causes the disruption of the metal structure, resulting in such brittleness that the annealed metal is termed “cold-short” and is unsuitable for drawing into wire or sheet. Curiously enough, the most common method of annealing copper wire is to heat it in a steam-sealed muffle wherein the wire is completely surrounded by an atmosphere of superheated steam a t slightly above atmospheric pressure. To the chemist this process may appear eminently absurd. The earliest lessons in chemistry have taught him that steam passed over heated copper is broken down with formation of copper oxide. The reconciliation of these apparently divergent facts is that the reaction, although actually taking place as it must, does so a t a very slow rate. The atmosphere within the muffle is also quiescent. The weight of copper within the muffle is measured in tons; the weight of the surrounding steam in ounces. This weight of copper in the form of copper wire represents acres of reacting surface. Therefore, such oxidation as does take place is not readily observable and the product is acceptable as being commercially “brightannealed.” Most research chemists have had the unfortunate experience of developing processes which have worked successfully in the laboratory to find their hopes blasted when the process is extended to full-scale operation, owing to the amplification on the larger scale of adverse factors which in the laboratory were so far minimized as t o be unobservable. The steam process for annealing copper is here presented as a rare example of a reverse case where on a large scale a process is commercially workable which in the laboratory would be discarded as being theoretically unsound. The steam process has never proved successful in annealing
INDUSTRIAL AND ENGINEERING CHEMISTRY
1168
large copper sheets, nor has it ever worked with brass in any form, for reasons that will now appear obvious. The effect of a hydrogen atmosphere on copper in causing the metal to become brittle apparently does not produce the same results when applied to alloys of copper, more particularly yellow brass. But its behavior is, however, consistent in that its powerful deoxidizing effect produces a somewhat unexpected result. If a coil of yellow brass sheet is heated to 1150” F. in an atmosphere of flue gas or in the oxidizing atmosphere of a radiant electric furnace, the annealed metal will have the same degree of softness in either case. However, the same brass heated to the same temperature in an atmosphere of hydrogen will be considerably softer, as though it had been heated to a much higher temperature. It is deduced from this that in deoxidizing the heated metal the reaction, which is exothermic, had been sufficiently intense to raise the temperature of the metal itself above that indicated by the pyrometer in the furnace muffle, thus rendering the process uncontrollable. Effects of Lubricants Thus far we have considered the effect of these atmospheres on the surface of metals which it has been presumed were “bright” before being subjected to the annealing process. Although the term implies that the metal surface was free from tarnish or oxide, it does not follow that it was also clean. When metals are worked cold by drawing, stamping, or rolling, it is customary to use a lubricant to prolong the life of the die or roll. This may be a mineral oil, fat, soap, or a mixture. The worked metal, ready for the next annealing operation, is invariably coated with a film of the lubricant. If it is a very light mineral oil, there is a chance that the oil will vaporize under heat without marring the surface of the metal. In most cases, however, the lubricant will crack and carbonize on the metal surface in such a manner as to nullify any at-
Vol. 21, No. 12
tempts at bright-annealing, If the atmosphere during annealing is oxidizing, the carbon will burn and dissipate. The metal, however, is then oxidized. If the atmosphere is reducing, the result may be even worse for, whereas the oxidized metal may be subsequently pickled bright, a carbonized surface may defy the action of the strongest pickling bath until the pickle solution has actually dissolved sufficient of the pure metal to loosen the grip of the adhering carbon particles. In either case pickling would have to be employed, thus defeating the very intent of the process. On these premises, a successful bright-annealing process necessarily involves two stages-an oxidizing phase which will effectively dispose of the lubricant, and a deoxidizing stage in which the reagent is powerful enough to free the metal surface from oxide but without deleterious effects on the internal structure of the metal. Conclusion This paper has been presented with a view to outlining the problems involved in atmosphere effects on the heat treatment of metals. An attempt has been made to define more clearly than heretofore the characteristics of these effects and to point a way to the possible solution of these problems. Work is now in progress along the lines indicated. It is hoped that within the near future commercial processes will be developed tending to eliminate or greatly minimize those adverse factors inherent to present methods of heattreatment and that these processes will place the gas industry in a still further preferred position in relation to this whole field. Literature Cited (1) Baur and Glaessner, 2. physik. Chem., 43, 354 (1903). (2) Guillet and Roux, Compl. rend., 184, 724 (1927). (3) Hadfield. Trans. Faraday Soc., 14, 173 (1919). (4) Matsubara, Trans. A m . Ins;. Mining Met. Eng., 61, 3 (1922).
Continuous Process for Plate Glass at the Ford River Rouge Plant’ Everett P. Partridge 1440 EASTPARKPLACE, ANN ARBOR,MICH.
I
N 1923 there was an acute shortage of plate glass. Prices
ran up to $1.25 a square foot, and even at this level the automobile industry experienced much trouble in procuring glass of the desired quality in the quantities required by its expanding production. The manufacture of plate glass at that time required an immense amount of specialized labor, and was far from being an efficient process. No continuous process had ever been put into successful production, but the evident advantages of continuous operation had influenced the Ford Motor Company to attempt the development of such a process. Coincident with the shortage in 1923, a new plate-glass plant was begun at River Rouge to use this process. The first furnace of the new plant went into operation on August 10, 1923. After a strenuous period of developmental difficulties, satisfactory production was finally established. Today the plant contains four complete units, each of which a t the present rate of operation can produce 16,000 square feet of glass per day of 24 hours. A second plant using the same continuous process has been put in operation by the Ford Motor Company a t St. Paul, Minn. 1
Received October 7, 1929.
These plants are not a t all similar to the traditional ones for the manufacture of plate glass. In place of the cumbersome handling of pots from furnace to casting table, and of the rolled plate from the casting table to the annealing lehr necessary in the old process, the new continuous process substitutes the direct flow of molten glass from a tank furnace to rolls which flatten it into a continuous sheet which passes immediately into the lehr. As the annealed glass leaves the lehr it is cut into convenient lengths and is transferred directly to a continuous grinding and polishing operation. The finished pieces of plate glass from the polishing tables are washed, inspected for defects, laid out for cutting, and cut to desired sizes, and the cut pieces pass on conveyors to the shipping room. From the feed of raw material to the furnaces through to the final transfer of the finished pieces, the whole process moves steadily along with a maximum of automatic mechanical control and a minimum of manual labor. Seven hours and a half after the melted glass flows from the furnace to the rolls, it is in the shipping room as cut pieces for Model A or miscellaneous sizes for sale outside of the company. This article is intended as a concise description of the