Februaqy, 1923
INDUSTRIAL A N D ENGINEERING CHEMISTRY
141
Pyrex Glass as a Material for Chemical Plant Construction' By A. E.Marshall 3034 ST. P A U L ST., BALTIMORE, MD.
course of a few hours and At the 1921 meeting of the Institute at Baltimore, the author read thereafter final collapse was dertaken construca short paper on uses of glass in chemical plant construction, and merely dependent on the tion work on chemipointed out certain desirable features of glass equipment. Refertime required for the acid cal plants, whether for the ence was made to the properties of Pyrex glass, and to a prospective fume to attack the wire manufacture of mineral investigation of the possibility of manufacturing Pyrex shapes reinforcement. acids or less corrosive prddsuited to the needs of the chemical industry. Other engineers have enucts, will admit that no i n the intervening period, development of the necessary manufacturdeavored to utilize glass in available material gives ing operations has proceeded satisfactorily, and it has also been various ways, and where satisfpctory service under possible to follow the behavior of the finished articles under actual temperature resistance and all conditions. working conditions in a large number of plants. mechanical strength have Materials in general use A s the Richmond meeting was mainly devoted to a symposium on not been essential, the mahave specific advantages plant materials, it seemed desirable to elaborate last year's note into terial has proved satisfacfor specific work, and the a more formal paper giving a summary of the development work, of tory. I have in mind, a s chemical engineer utilizes the present status of Pyrex as an industrial material, and of itsfuan instance, the glassa number of different mature possibilities. packed Gay-Lussac and terials in one piece of conThe following paper describes in detail the present status of inabsorption towers introstruction in order to develop dustrial Pyrex, and carries a suggestion of future possibilities. duced in England about maximum durability unIt is believed that Pyrex will in a short time occupy a definite place 1909 by Carmichael. These der varying conditions. Inalong with the other materials used for plant construction. i t is towers, usually of square cidentally, durability is not a universal panacea for construction ills, but, like everything section, are packed with in many cases synonymous else in our practice, it demands thought in its application and care annealed plate-glass sheets, with resistance to corrosion. in its use. set on edge and spaced Glass, because of its infairly closely. Each successolubility in acids, has always foind uses in the chemical industry, but its limitations sive layer of packing is placed a t right angles io the row below, in the way of temperature resistance have restricted the thus giving excellent surface contact and good gas distribution. field of application. Mention has been made of attempts to use and the actual Ordinary glass, if desired to withstand moderate temperature changes, has to be made into shapes with thin walls, and use of ordinary glass in plant construction, because such atsuch shapes when thin enough to survive slight heat shocks tempts afford evidence of a desire on the part of the chemical are much too thin to stand up under ordinary plant usage, or industry to utilize the valuable noncorrosive properties of i n many cases to survive rough handling during erection. glass. With the introduction of Pyrex as a laboratory material The desire to make use of the noncorrosive properties of glass has led to its being tried out under a variety of plant and as a domestic utility in the form of baking ware, the posoperating conditions, but the inherent disadvantages of ordi- sibilities of using glass in plant construction assumed a more nary glasses so far outweigh the useful features that it is promising outlook. Chemical manufacturers tried out Pyrex difficult to point to successful applications in plant-scale baking ware for small-scale chemical operations, drawn Pyrex tubes were used in Hart nitric acid condensers, and work except under special circumstances. Some years ago I made a survey of materials used in the con- many other minor uses were discovered for the standard struction of gas conveying lines from hydrochloric acid pots and shapes which were being produced for the laboratory or the muffles. One plant a t the time of my visit was using 15-in. home. I n December of last year the idea of the desirability of slip-joint glass pipes on the pot line. Very good results were being obtained because of the thorough cooling of the gas in gathering together these sporadic developments and investiits passage through the thin-walled pipes to the absorption gating the possibilities of producing a line of Pyrex products tower. Later inquiries indicated rather heavy breakage designed for the chemical industry was suggested to the Cornduring the winter months, the cause being ascribed to leakage ing Glass Works. The field appeared promising and a of melted snow through the roof and onto the pipes, or to the decision was reached to establish an Industrial Pyrex considerable difference in temperature between the atmos- Department. phere and the gas inside the pipes. CHARACTERISTICS OF PYREX There was a purely local reason for the use of glass on this line. The chemical plant adjoined a glass factory, and the Before entering into a description of the present state of glass pipes were thin cylinders from the plate glass depart- development, it is necessary to set forth the essential charment. acteristics of Pyrex and the difference in its properties and When wire glass was introduced it seemed to offer favor- those of ordinary glass. able possibilities t o the chemical industry, and it was tried out Pyrex is a low-expansion borosilicate glass of simple chemfor various uses without any great measure of success. I ical composition, containing no metals of the magnesia-limewas interested in sulfuric acid concentrators at that time, and zinc group and no heavy metals. A comparison of the substituted some wire-glass plates for the acid-proof cover linear expansion coefficients of Pyrex and a number of maslabs on a cascade concentrator. The glass cracked in the terials is given in the attached table, and, as will be seen, Pyrex has a smaller coefficient than porcelain, ordinary glass, 1 Presented before the 15th Annual Meeting of the American Institute of Chemical Engineers at Richmond, Va., December 8 to 9, 1922. or any of the usual metals. The low coefficient of expansion
NYONE who has un-
A
IhlDUSTRIAL AND ENGI
142
introduces a marked distinction from ordinary glass whether of the lead or lime-soda type. The melting point of glass is not particularly valuable as an engineering consideration, as there is usually a fairly wide range between the initial softening point and final melting point. In the case of Pyrex the softening point is ahout 800" C., hut the material will soften slightly, especially under pressure if maintained for a long time above 600" C . In connection with the softening point it is usefn1 to remember that devitrification, which is a serious factor in the use of some materials, does not affect Pyrex in its working range. Lrmnn EXPANSION COSPIIECIENTS
Fm. I
Acid resistance is not usually given much thought in the case of glass, the general assumption being that all glasses are equally resistant. Bullelin 107 of tho Bureau of Standards gives considerable data on the acid-resisting qualities of Pyrex and other glasses, and is well worth study. Researches have also been conducted in the Corning laboratories on the resistance of Pyrex to perchloric, phosphoric, constant-boiling hydrochloric, and concentrated sulfuric acids under a variety of conditions. The action of hydrochloric and sulfuric acids is imperceptible, constanhhoiling hydrochloric acid attacking Pyrex a t a rate of 0.oowO6 g. per sq. om. per hr. Concentrated sulfuric acid in 4 hrs. a t the fuming temperature shows an attack of 0.oooO02 g. per sq. cm. per hr. Both figures relate to initial surface attack, as a f k r the lapse of a few hours a state of practicalstahility is rached. Work on the coefficient of heat transference is being carried out a t Corning but has not been completed. Preliminary results on the relative efficiencies of Pyrex, porcelain, and stonewareindicate Pyrex and porcelain as equal, whereas stoneware shows slightly less than one-half the Pyrex value. It can be said that while Pyrex is superficially a glass, its physical characteristics justify consideration from the engineering standpoint BS a special and distinct material adapted to a variety of industrial uses to which ordinary glass cannot be applied.
DEVELOPMENT OF SHAPES Proceeding from a consideration of useful properties to the development of definite industrial shapes, i t is obvious that the upper limit of size is a factor of great importance. If
equipment was constructed recently for the du Pout Company. This was in the form of an experimental denitrating tower 6 in. in diameter by 8 ft.high (Fig. 3).
*
h
~
February, 1923
INDUSTRIAL A N D ENGINEERING CHEMISTRY
One plant, which is now completely equipped with Pyrex dishes, reports handling breakages as nil, and states that the men treat the dishes as glass, setting them down with care and so obviating breakages through dropping. It has been the experience of every user of Pyrex, irrespective of the shape of the article, that transparency creates care in handling. It is a natural assumption that an opaque object will withstand shocks and that a transparent one will not, so a transparent material which is a t least equal to stoneware and porcelain in mechanical strength has a much better chance to survive a t the hands of a workman. Transparency has other advantages which are not psychological. Th e first and most important is that a transparent article cannot have blow holes or other FIG.3 hidden flaws. Even strains can be detected by the use of a special polariscope, this being one of the routine tests to which every piece of Pyrex equipment is subjected in the factory. Next comes the question of cleanliness. Opaque equipment may or may not be clean, but transparent equipment always supplies its own positive answer on this point. Finally, there is the feature of controlling reactions through direct observation. Another interesting feature which has been brought forward by users is in relation to the process of manufacture. Pyrex is to a certain extent competitive with chemical stoneware. The process of making stoneware is quite lengthy, the time required being about two months. Breakage of a special piece of stoneware, provided there is no duplicate in the plant storeroom or a t the stoneware manufacturer’s works, means a long delay in starting up after the shutdown. If a mold exists for the piece in Pyrex, manufacture can be completed in three days a t a minimum, although factory conditions might necessitate a delay of a few additional days on account of prior routing of work. Touching on the phase of excess costs of construction materials, it seems very desirable to emphasize the lack of standardization in the chemical‘ industry. Molds are costly, whether they are intended for stoneware, silica, Pyrex, or other materials. The manufacturer of the equipment has to charge up his mold cost to the user, and there seems to be too great a demand for “specials” which may vary only l / g in. from a stock mold. Some concerted effort to standardize shapes would cheapen the products, and, of equal importance, would enable producers to carry representative stocks. A striking example of the lack of standardization is shown by the Corning Glass Works list of sight glass molds. Continuous efforts have been made to keep down the number of sizes, but the success can be judged when it is shown that in. thickness to between the range of 21/2in. diameter by B1/2 in. by 3/4 in. it has been necessary to provide 41 molds. I n the case of plant equipment, a further effort is being made towork out shapes that will suit a variety of uses, and it is hoped that a full measure of cooperation will be extended by plant managers and engineers, through the use of stock rather than special molds.
143
INSULATORS The industrial use of Pyrex is spreading into many fields, and, while not a strictly chemical application, decided interest attaches to the development by the research laboratories of the Corning Glass Works of Pyrex high-tension insulators. The generally accepted causes of failure of porcelain hightension insulators are: (1) change in structure of porcelain with time and absorption of moisture; (2) breakage due to thermal changes; (3) flaws in the porcelain body causing dielectric and other failures; (4)failure due to expansion of cements used in attaching hardware to the insulator; ( 5 ) mechanical weakness: These failures may be generally classified as failure due to the properties of the insulator and failure due to design. Pyrex glass seems to have ideal properties for an insulator for it apparently is not subject to any of the intrinsic weaknesses of the porcelain insulator. It does not change in structure, can be inspected for any defects thus assuring a uniform product, has a sufficiently high dielectric strength, a great resistance to thermal changes (its thermal expansion coefficient being lower than that of porcelain), and in addition is not heated by direct solar radiation as much as porcelain. Consequently, if a Pyrex glass insulator could be built free from the design defects of the porcelain, there is no question but that it would be a better insulator. The drawing shows the construction of the Pyrex insulator (Fig. 4). It is evident a t a glance that this construction and design gives a cement-free, all metal and glass insulator, and that the design is one of great resistance to tension as the material of the insulator is largely under compression. Considerable data have been collected showing the relative properties of Pyrex which show the increase in temperature due to solar radiation to be 3.65 times as much for the porcelain as it is for the glass. The glass is transparent and the porcelain absorbs the heat radiation. This proves that in service the glass is not subjected to anywhere near as severe heat changes as is porcelain. One of the large manufacturers of electrical equipment says: “The tests on Pyrex suspension insulators show quite conclusively that the material has good characteristics. The uniformity of electrical puncture is a distinct advantage over porcelain.” It is evident that Pyrex glass has very good dielecFIG.4 t r i c properties, more thall ample for the service, as well as being uniform. Tensile strength tests of the better grades of porcelain insulators show that they will not stand over 10,000 lbs. load as a maximum, and many break below that. It is therefore evident that the glass insulator has a strength far superior to anything yet developed. This property should allow for longer spans, and in many cases where two strings of porcelain are used in parallel one of Pyrex will be ample to carry the load. There are other points of probable superiority of the glass insulators which will have to be proved by service tests. Pyrex glass has a great resistance to water absorption and
I N D U S T R I A L A N D ENGINEERING CHEMISTRY
144
surface attack, and it is therefore durable under long exposures to severe atmospheric conditions, as, for instance, around chemical plants. It is probable that glass insulators will not be as attractive to birds and spiders for nesting sites since they will not have the dark shadows that exist in a string of porcelain insulators.
Vol. 15, No. 2
The chief advantages of the Pyrex insulator are the facts that it does not absorb heat as does porcelain; has no cement in its construction to gradually absorb water and expand, ultimately fracturing the insulator; and has a mechanical strength that will average twice as high as any porcelain insulator yet on the market.
Tools of t h e Chemical Engineer.
I--Agitators’
By D.R.Killeffer 50
UNDAMENTALLY,
EAST4
1 s STRBET, ~ Nsw YORK, N. Y ,
stream of liquid caused by Agitation, as the term is applied to chemical engineering practice, agitation is accomthe rotating paddle. The may be defined as the operation causing mutual movement of particles plished by applying centrifugal component of of matter for the purpose of bringing about intimate contact between force toamore or less fluid the applied force builds u p them. Zn general practice the term is applied to the mixing of mass and then opposing the rapidly as the speed of rotadiyerent liquids, diyerent solids, and liquids and solids, although resulting movement in such tion increases, and is orin a strict sense its application to mixtures involving a gas phase a way as to alter its direcdinarily only opposed by is equally logical. The purpose of the present article is to rliscuss tion or otherwise emphasize the containing walls. The some of the fundamental principles of agitation and the means its selective action on varimovement in the mass available for putting them to work in the chemical plant. Experious parts of the mass. caused by this force is utilmental eoidence of the e$ectiaeness of any particular form of agitator Movement must be brought ized more or less incidenis sadly lacking, and nothing better than the old “cut and try” method about in parts of the mass tally in the ordinary paddle is available to the prospectioe user of such apparatus. There is no in such a way as to force mixer but becomes the prinmethod at present in use by which the effectiveness of any particular changes in the relative posicipal force in turbine mixers type of machine may be compared with any other, either as to speed tions of its various partiwhich will be considered of mixing or power consumed to dccomplish it. The result is that cles. The means by which later. Motion from the cenchemical engineers generally are inclined to instal tools which will this result is accomplished ter toward the walls of t h e give greater mixing than is actually necessary with consequent waste in practice may be roughcontaining vessel is changed of power. It is to be hoped that the generalizations put forth here ly classified into six groups: in directionat that obstrucmay lead to a closer study of the subject by those best equipped to tion into up or down motion (1) simple paddles and undertde it. modifications, (2) turbines following the wall, and is laraelv influenced by the and modifications, (3) propellers and modifications, (4)processes involving movement strong rotational movement which is-especially rapid at the of the containing vessel, (5) combinations of two or more periphery of the vessel. Few machines are now built which of the above, ( 6 ) miscellaneous unclassified types. are of the simple paddle type, as most builders incline the paddles in such a way as to add a force component in the PADDLES The simple paddle in the hands of a workman is seldom direction of the paddle’s axis. Combinations of paddles are frequently used. Two paddles found profitable in modern operations, but frequently its immediate successor, single or multiple vanes mounted on a mounted to revolve in the same direction a t the focuses of rotating shaft, is found quite efficient where movement need an elliptical vessel, three paddles in a clover-leaf vessel, etc., not be violent, as in those cases where liquids of very nearly generate confficting currents in the liquid and bring about The mounting of the same specific gravity are to be mixed and where mixing is very efficient mixing in many processes. two sets of paddles on the same axis to revolve in opposite used as a means of aiding osmotic forces. Dissolvers and directions has been found to consume a disproportionate leachers of this class with slow-motion paddles are found to be efficient, as in these tools the problem is really the mixing amount of power. The use of single or multiple sets of of the strong solutio? in direct contact with the solute with the paddles mounted off-center in cylindrical tanks gives economical mixing in many cases where a centrally mounted paddle more dilute solvent. Paddle agitators involve the application of force with two would result in a simple rotation of the mass without r e d directional components, one in the direction of rotation and a mixing. Inclining the axis of a cylindrical vessel with a second centrifugally and a t right angles to the f i s t . The simple paddle revolving in its center produces a marked inrotating force is comparatively greatest and the centrifugal crease in mixing by utilizing gravitational force. TURBINES force least when the movement of the paddle is slow and tends The turbine type agitator is really a special case of the to bring the mass as a whole into rotation at a speed approximating that of the paddle. Actual mixing in this event is paddle type, but since the principal force used is centrifugal comparatively slight. The inertia of the mass, as well as the rather than rotational, it must be considered separately. resistance to rotation offered by the walls of the containing In this type of agitator the rotating unit is so designed as t o vessel, tends to start cross currents of various kinds and thus minimize the rotational component of force and increase the promotes some mixing. Increases in the resistance to this centrifugal Component to a maximum. I t s operation is in all rotational motion are made by introducing stationary paddles respects similar to that of a centrifugal pump, with the exor simple vanes projecting from the sides of the containing ception that the motion of the liquid is permitted in every vessel in various ways which break up the otherwise solid radial direction instead of being forced in a single direction by a casing or pipe. The radial currents are directed toward 1 Received October 26, 1922.
F
