Document 7Gb1xg5NvnvXe3kN77OngOMa

(Reprinted from The Journal of the American Ceramic Society, Vol. 23, No. 1. Jaouary, 1940.] DEFORMATION OF INSULATING FIREBRICK UNDER LOAD* By C. L. Norton, Jr., and V. J. Duplin, jR.t ABSTRACT An electrically heated furnace and a method to apply loads to full-size insulating firebrick heated throughout to thermal equilibrium are described. Sizteen'commercial brands of insulating firebrick of various use limits were tested under several loads over a range of temperatures. No correlation was found to exist between the weight and the resistance to hot loading or between the cold crushing strength and the hot loadbearing capacity. I. Object The object of this work was to devise a suitable test procedure to measure the relative load-bearing resist ance of insulating firebrick at various loads and tem peratures. For convenience as a routine laboratory test, similar to the corresponding test for dense fireclay refractories, the test was made of such duration that it could be run in an eight-hour day. A test of this short duration, however, would not give sufficient information for complete engineering or design purposes because the rate of deformation decreases as the time of test is continued and a uniform rate is not established until the sample has been under load for several days. Obviously such long-time tests are not justified unless some special design features are being investigated in connection with hot -load deformation. The foregoing ! Presented at the Forty-First Annual Meeting, Ameri can Ceramic Society, Chicago, 111., April 19, 1939 (Re fractories Division). Received June 10, 1939. t Mr. Norton is Technical Director, Refractories Division, and Mr. Duplin is in the Refractories Research Labora tory of the Babcock & Wilcox Co. remarks apply equally to insulating firebrick and dense firebrick. The question of load-bearing resistance of insulating firebrick for ordinary wall constructions usually is unimportant. The temperature gradient through a lightweight insulating brick is so steep that the cool portion is able to carry the load even though the hot face has exceeded its load-temperature limit. There are many applications, however, in which insulating firebrick must withstand load conditions, such as in sulated sprung arches of wide span, floors and car bottoms, metal holding furnaces, and parts of furnaces wherein insulating firebrick have been substituted for dense firebrick to reduce the weight of the furnace structure. The insulating brick in the latter case are sometimes heated on both sides so that the loadtemperature conditions have been severe. After selecting a suitable test procedure, the further object was to study the relationship of cold crushing strength anti of bulk density to the load-carrying ca pacity when the brick are hot. To determine this re lationship. sixteen brands of commercial insulating Vol. 23. No. 1 Bulletin 3-523 Deformation of Insulating Firebrick Under Load firebrick with widely differing properties were tested under the selected test procedure. II. Method Before establishing a test procedure, previous work by Phelps* was reviewed. He concluded that test brick should not bei loaded when heated on all sides because the deformation is greater under these'condi tions than is encountered in normal wall construction wherein the high temperature is confined to one face of the brick only. For this reason, Phelps has recom mended loading a wall which is heated on one face only. Because the cold part of the wall in this case supports the load, it'is obvious that a test of this type cannot supply any information concerning the loadtemperature conditions under which plastic flow results. It is also obvious that no relationships can be deter mined from such a test. For this reason, a test to be of value for the work outlined must be conducted on a sample which is heated on all sides. . In previous work on "soaking" heat load tests, the load was applied during a one and one-half hour soak ing period. It was j apparent that the center of the brick would not reach equilibrium until the tempera ture had been held for some time, thus reducing the 1 S. M. Phelps, "Properties of Insulating Refractories: I, Behavior Under Load at' High Temperatures," Amer. Refrac. Inst. Tech. Bull., No. 61, 6 pp. (December, 1936). effective loading time of the te&t. A variation in heating-up time would also cause a change in the tem perature lag between the inside and outside of the test brick. To eliminate these two factors, the furnace temperature was balanced long enough to bring the inside of the test brick to equilibrium before the load was applied and then the load was applied for one and one-half hours. To determine the temperature relation between the inside and outside of the test brick, thermocouples were placed in the center of the brick to determine how long a temperature-holding period would be required to bring the center to equilibrium conditions. Figure 1 shows the time-temperature relation between the inside and outside on a normal heating schedule. From these results, the furnace was brought up to the test tem perature, the temperature was balanced for one hour, and the load was applied for one and one-half hours in all test runs. III. Load Test Fumac* . The furnace used in these tests had been used previ ously. for testing dense firebrick (Figs. 2 and 3). It consists essentially of a rectangular furnace chamber heated by six Winch Globar elements. Loading is accomplished by means of load beams with suitable knife edges and, supports to reduce friction and ar ranged to give accurate vertical loading on the test brick. The lining of the furnace chamber is of insulat ing firebrick to reduce the heating-up time and to maintain a uniform temperature. In Fig. 2, four sighting windows are shown in the furnace casing, and opposite them are two reading telescopes, one of which has a filar micrometer eyepiece. This arrangement permits accurate measurement of subsidence of the test brick while at test temperature. When long time tests. are run, this method of measurement is necessary, but in the. short-time test described here, the deformation was determined by measuring the test Fig. 1.--Temperaturevs. heating time during load test ing of insulating firebrick. (1940) Fig. 2.--Electrically heated load test furnace. t 24 Journal of The American Ceramic Society--Norton and Duplin brick before the run and again after the test was com pleted. Furnace temperatures were measured by a platinumplatinum-rhodium thermocouple placed midway be tween the two test brick and protected by the usual porcelain tube. IV. Procedure The sample brick were measured carefully for length and were set in a vertical position on the 4*/- by 2`/rinch face. A fireclay brick was set upon this sample as a spacer and the loading assembly was completed (see Figs. 2 and 3). A slotted section of I-beam was used for the middle knife-edge support. The furnace was heated up, as shown in Fig. 1, held at test temperature. for one hour, and then loaded by slowly lowering the unloaded beam onto its support and by applying the weights gradually. The temperature was maintained for 6ne and one-half hours, after which the furnace was shut down and the samples were allowed to cool to room temperature under load. Table I Test Data on Commercial Insulating Firebrick Brand Use limit (*F) Weight (lb./9 in.strsight) Modulus of rupture 0b./ sq. in.) Cold crushing strength Ob./ sq. in.) A 2000 B C D O ** E 2200 F ** G ** N '* I 2500 J ** K 2600 L S M 2800 T 2900 1.72-1.82 2.21-2.75 1.64 2.27-2.32 1.88 1.87-1.96 2.71-3.07 2.27 2.44 2.40-2.63 2.76 2.30-2.50 2.68-2.80 2.40 2.42-2.48 2.90 95 120 197 172 71 104 107 559 106 176 165 193 224 377 129 208 66 235 196 248 251 408 133 128 203 372 127 196 140 150 235 220 Water-cooled holders-. , f/'Y ft 'S'//'/; r\ ^ $ -Q cai 9 a few it Specimen Sec.A-A Fire brick brand A B C D 0 E F G N 1 J K L S M T No. 1 No. 2 Table II 1#00*F. 10 25 lb. lb. 0.1% 1.6% 0.1 Failed 10 lb. 0.1% 0.6 1.0 0.9 Failed 0.1 0.1 0.1 Failed 0.1 0.0 0.1 2000 F. 25 lb. 0.1% 8.3 Failed 1.3 40 lb. 0.1% 2.5 0.0 0.1 0.4 1.8 1.8 7.6 0.8 1.2 0.2 0.2 0.2 0.4 0.1 0.2 0.1 0.2 Load Test Results 2100F. 10 25 40 lb. lb. lb. 0.0% 0.2% 0.3% 2.5 2.9 4.9 0.0 0.1 0.2 0.2 2.5 7.6 1.5 Failed 0.2 2.0 Failed 0.0 0.2 0.7 0.3 0.4 0.6 0.1 0.9 1.6 0.2 0.3 0.8 0.0 0.4 0.9 0.2 0.1 0.4 10 lb. 0.4% 3.8 0.1 2.5 1.9 0.1 0.3 0.9 0.4 0.1 0.2 2200*F. 25 40 ' lb. lb. 0.7% 1.1% Failed 0.1 0.4 Failed Failed 1.7 1.0 1.2 2.9 1.0 0.4 Failed 2.9 5.5 8.2 3.9 0.6 0.4 0.2 10 lb. 1-1% 1.6 1.4 0.4 0.3 2300* P. 25 40 lb. lb. 2.7% Failed 2.6 Failed Failed 1.3 Failed 0.8 0.9 Failed Vol. 23. No Determination of Insulating Firebrick Under Load 25 \ V. Samples Tested Sixteen brands of commercial insulating firebrick were chosen to be representative of the various temperature groups The use limit, cold strength, and bulk density of these brick are listed in Table I. Two brands of dense second-quality fireclay brick were ih'duded for purposes of comparison. VI. (Results ' The results obtained are shown in Table II. The loads were taken as 10 lb., 25 lb., and 40 lb. per square inch and the temperatures were 1900, 2000, 2100, 2200, and 2300F. The 'deformation values are also shown in Fig. 4. VII. Conclusions There is no apparent relation from these results be tween cold strength and plastic flow. It might be con cluded that a dense strong brick usually would carry more load when heated than the lighter and more fragile brands. This is not the.case, however, inasmuch as some of the brands with thebestload-bearingcapacity are also among the liglitest;|>ruids tested* The &ulk density and'.cold strength^in general, do not whether or hot the brick(structure is capable of ,with-'(^pfiw standing hot-load conditions'^.2 ' iOne point of considerable .interest riasthatsome brands of insulating-firebrick'chn actually carry more i load at the same-test tdl5erature than the two samples r of dense firebrick t&ted. It-is surprising that a'.refractory structure which consisted 80% solids will not stand as much actual loading as another refractory " structure containing only 20% solids. Rbpractoubs Division Babcock & Wilcox Company Nbw York, N. Y. (1940) Babcock & Wilcox Refractories Division Product Information 1712 APPLICATION-B&W FIREBRICK Steam Generators Selection of Refractories for 776 Oil Fired Furnaces A common phenomena observed in furnaces fired with 776 fuel oil is slag formation on the surface of the brick. The average operator refers to "melting" of the brick. Actually a reaction is taking place between the brick and impurities in the fuel oil which result in lowering the melting point of the exposed surface. Although the impurities may be well under 1%, the total weight of oil introduced into a furnace may be hundreds or thousands of pounds per hour and the weight of impurities can build up rapidly. x Usually there will be a period of operation with no slag followed s by a gradual build-up as the concentration becomes sufficiently high to permit lowering of the melting point of the exposed face. The attached article "Petroleum Ash Components and Their Effect on Refractories" by Jones and Hardy covers the basic factors involved in slagging of furnace walls, the impurities which are responsible for the effects, and proposed remedies by improved selection of refractories. We disagree with their conclusion on the use of basic brick because such brick will spall severely in the average industrial or boiler furnace service. We do agree with the use of higher alumina content bricks. The Navy has been using B&W Allmul in their severe duty furnaces of older design for some years because slagging can be eliminated by their use. However, Allmuls should not be used if Na02 is present in substantial quantities in the ash because a nephelite formation may result with a marked expansion increase. B&W Kaomul has shown a definite improvement over brick such as superduty of regular or high burned variety. B&W Kao-70 would be good for such service although 70% alumina brick of 25-30% jJorosity would not do as well. Wash coatings of B&W Mulset or B&W Chrome Mortar applied to new brick of high duty or superduty grade will postpone slag formation for a period of a few weeks. To be of permanent value they must be reapplied every month or so before slag formation has started. 1 7 12 oo ooz 3oFirebrick Insulating Firebrick Refractory Castables Plastics Ramming Mixes Mortars Ceramic Fiber Special Oxide Refractories CO CL DC LlI a </) SEQUENCE Petroleum Ash Components and Their Effect on Refractories MINOR C. K. JONES AND ROBERT L. HARDY Esso Laboratories, Standard Oil Development Co., Linden, N. J. While the ash content of petroleum is extremely small, varying from less than 0.001 to about 0.05% of the crude, this material is nonvolatile and concentrates in residual fractions marketed as heavy fuel oil. A study was made of the reactivity of these residual ash components on the refractories used in industrial furnaces. In addition, the ashes from various crudes were analyzed. Petroleum ashes containing vanadium are very destructive to fireclay brick upon direct contact at high temperatures. The alkaline earth and alkali metals (particularly sodium) and certain other metals will also lower the fusion point of refractories but are not as destructive as ash which includes vanadium. Fuel oils derived from most crudes contain ash which is potentially destructive to refractories in furnaces of improper design, as shown by the ash analysis of 25 crudes. The importance of proper furnace design is emphasized, particularly as to the selection of refractory brick resistant to attack by ash components. NE of the chief advantages usually associated with petro O leum fuels is the absence of the ash disposal problem. The ash in a petroleum crude is indeed small, varying from less and disturbing the balance of turbine rotors, ash deposits can cause severe damage by corrosion. A more common example is the corrosion found in industrial boilers, where, under certain than 0.001 to about 0.05% of the crude. This material, however, conditions, ash residues can cause severe damage to boiler tubes is nonvolatile and concentrates in residual fractions marketed and to the refractory brickwork. as heavy fuel oil. When large quantities of such fuels are burned, The present paper deals with the effects of these residual ash troublesome deposits may build up despite the low ash content components on the refractories used in industrial furnaces. of petroleum. Some of these ash residues are very reactive and Individual components of a typical ash have been evaluated to under certain conditions hav'e been known to cause severe damage isolate the chief offenders and the resistance to damage of several to equipment. For example, this is of particular concern in the typical refractory bricks has been studied. In addition, the operation of gas turbines, where accumulations on the turbine analyses of the ash found in a number of crudes are presented to blades cannot be tolerated. In addition to reducing clearances show thadistribution of the harmful components. November 1952 INDUSTRIAL AND ENGINEERING CHEMISTRY 2615 EFFECT OF ASH COMPONENTS ON REFRACTORIES In many oil-fired furnaces where there is a direct impingement of flame on the refractory, the firebrick becomes coated with a thin glassy substance. This has been attributed to actual fusion of the refractory. However, one investigation has shown (2) that the surface temperature of the brick, even with direct flame impingement, is not high enough to cause fusion; the coating on the wall was, rather, the ash originating in the oil or a slag formed by the reaction of the ash and the refractory. In the particular furnace investigated, the major cause of failure of the refractories was spalling. The investigation showed that this spalling was caused by the formation of slag which was absorbed into the brick. When the temperature of the furnace changed, because of a change in the firing rate, cracks and spalling of the brick resulted from the differential expansion of the brick and the slag. Another type of refractory failure led to the present investi gation. This was a case of certain ships in trans-Atlantic service each firing four marine boilers with approximately 10,000 barrels of fuel oil per round trip. The brick work at and below the area of flame impingement had a glazed surface marked by numerous vertical channels where rivulets of molten slag had run down to the furnace floor. The damage was so severe that during each round trip about 1'/, tons of slag accumulated on the floor of each boiler. This slag was in a molten state under furnace conditions at sea but solidified when the boiler load was reduced in port. Every four trips the accumulation of about 10 inches of this hard, glassy slag had to be removed laboriously by chip ping. Analyses of this slag and of furnace dust collected on the radiant section tubes of the boiler are shown in Table I. more reactive ash components i- liurncd. For this reason it was felt that a determination of the chemical component- of petroleum ash actually responsible for the destruction of refrac tories would be useful. The effects of ash components on fusion point of firebrick and chemical reactivity of the slags produced were chosen for investigation. Experimental. Petroleum ash for fusion experiments was prepared by distilling crude to coke, burning the coke in a stain less steel furnace, grinding to 60 mesh, finally igniting in platinum or silica dishes in an electric furnace, and again grinding to 200 mesh in an agate mortar. Iron introduced by crushing and grinding the original coke was removed by a magnet. The fusion point determinations were carried out in a gas-fired furnace as in the ASTM procedure D-271-48 for the fusibility of coal ash. Ground refractory brick and completely oxidized ash or ash component were used in the determinations. The desired mixture of the two was moistened with a 10% dextrin solution and worked into a plastic mass. The mass was then molded into small triangular pyramids 3/t inch high and >/4 inch wide at the base. These "cones1' were then dried in an oven and mounted on a refractory base composed of equal parts of kaolin and calcined alumina moistened and worked into a small plate as specified in the ASTM procedure. Usually five or six cones of the same material were mounted to ensure accurate observa tions. The mounted cones were placed in the furnace and the tem perature was gradually increased at a uniform rate. The fusion or softening point was taken as the temperature at which the cone fused into a hemisphere of homogeneous slag. In addi tion to the fusion temperature, the destructive action of the slag on the kaolin-alumina base was also noted. In the following dis cussions this destructive action (or reactivity) of the slag formed by fusion of the mixture is reported as follows: Table 1. Analysis of Slag and Furnace Dust Analysis, Wfc. % Vanadium (as V*0) Refractory materials* Other metals* Alkali metals, sulfur, and halogens Ignition loss Fusion point, 0 F. Slag from Boiler Furnace Floor 12.5 80.9 2.3 3.6 Nil 2199 Dust from Radiant Section Boiler Tubes 35.3 7.8 6.9 47.0 2.8 1400 Alumina, silica, etc. * Including iron, molybdenum, nickel, cobalt, tin, and zinc. The dust from the radiant section boiler tubes should be fairly representative of the ash content of the fuel oil. This dust contained high perentages of vanadium and alkali metals, partic ularly sodium. Slag from the furnace floor also contained relatively high percentages of vanadium, but the amount of refractory materials (alumina, silica, etc.) was about tenfold higher than in the dust from the tubes. It is quite evident from these analyses that the fuel oil ash reacted with the refractory material and lowered its fusion point. Fusion point measurement on the slag showed that it melted at about 2200 F., which is far below the softening point of the refractory material used. There are many factors which probably contribute to failures of the type observed in these marine boilers. Among these are flame impingement on the wall, which increases wall tempera tures to around 2800 F.; furnace atmosphere, which is known to be severely detrimental to furnace refractories when a high concentration of carbon monoxide is present; the composition of the refractory used; the furnace design; and the composition and quantity of the fuel oil ash. Of these, the last is the only factor over which fuel oil suppliers have any degree of control. Proper furnace design and choice of refractory materials will reduce chemical attack from any type of petroleum ash, but particular care should be taken when fuel oil containing the 1. Complete decomposition. The slag reacted with the entire kaolin-alumina baseplate causing it to fuse completely. 2. Extensive, considerable, slight, and evident decomposi tion. The slag reacted with the kaolin-alumina base in varying degrees ranging from fusion of a large area to a minor attack around the base of the spherical lump of original slag. 3. No decomposition. The slag formed a spherical lump with no evidence of attack on the base. These observations are necessarily qualitative in nature. However, they are thought to be quite significant in the evaluation of the chemical reactivity of the slag since the kaolin-alumina mixture is typical of refractory materials. These qualitative observations of the chemical reactivity were made at the fusion point of the sample under test. The severity of chemical attack probably increases with temperature and, thus, a slag having a low fusion point but showing a slight evidence of chemical attack may actually be very reactive under higher temperature furnace conditions. Table II. Fusion Point and Chemical Reactivity of Undiluted Petroleum Ashes Ash Source Lagunillas crude Quiriquire crude West Texa&iCrude East Texas, West Texas, Talco crude mixture Fusion Point, 0 F. 2760 2445 2400 Reactivity with KaolinAlumina Base at Fusioo Point Complete decomposition Some decomposition Considerable decomposition 2155 Considerable decomposition Discussion of Results. Table II shows the fusion point and the chemical reactivity of undiluted petroleum ashes from a number of crude sources. The fusion point of these pure ashes varied from about 2100 to 2800 F., and all resulted in considerable de composition of the kaolin-alumina base at the fusion temperature of the ash. 2616 INDUSTRIAL AND ENGINEERING CHEMISTRY Vol. 44, No. II ____ PETROLEUM--COMPOSITION_______ Table 111. Fusion Point and Chemical Reactivity of Vanadium- and Sodium-Containing Mixtures Mixture, % Fusion Point, Reactivity with Kaolin- F. Alumina Base at Fusion Point Firebrick, 100 V0, 100 VjO, 10; firebrick, 90 VtO*. 50; firebrick, 50 VtO, 80; firebrick, 20 NaVOi, 10; firebrick, 90 NaVOi, 50; firebrick, 50 NaVOi, 80; firebrick, 20 Above 2800 1274* Abo\e 2855 2085 1290 2fi40 2G20 Approx. 1800 No decomposition No decomposition Evident decomposition Evident decomposition Slight decomposition Kvident decomposition Extensi\e decomposition * Employing Pennsylvania fireclay brick containing 39% Al?Oi. * From {/). These data show that chemical reactivity is not limited to the ash from one crude source but is typical of the ash from a wide variety of crudes. The ash from the Lagunillas crude completely destroyed the kaolin-alumina base but the ash from other crudes showed considerable reactivity even though their fusion points were, lower. Table IV. Fusion Point of Brick with Alkali and Alkaline Earth Metals (Mixture, 50% compound and 50% Pennsylvania fireclay brick) Compound MgSO< MgO CaSO CaCOi NaCl NajSO* K*SO NajCO* Fusion Point, 0 F. 2530 Above 2900 2320 2300 2670 1950 2740 1840 The presence of high concentrations of vanadium and sodium in the ash causing damage to the marine boiler refractories described above, suggested that these metals are particularly destructive. Consequently, 10, 50, and 80% mixtures of vana dium pentoxide and of sodium metavanadate with fireclay brick were tested for fusion point and chemical reactivity. The data are shown in Table III. All compositions except the one contain ing 10% vanadium pentoxide showed a marked reduction in the fusion point of the brick. This effect increased, as might be ex pected, with the vanadium content. Sodium metavanadate in low concentrations (10%) is evidently more active than the vanadium pentoxide in reducing the fusion point, but in high concentrations, vanadium pentoxide is the more destructive. In all cases where fusion occurred there was evidence of chemical attack on the kaolin-alumina base. This reactivity increased with the concentra tion of the vanadium compound even though the fusion tempera ture at which the attack was observed decreased. If anything, the sodium vanadate was more destructive to the kaolin-alumina base than was the pure vanadium pentoxide. It may be con cluded from these data that vanadium compounds in general are reactive with refractory materials and will cause significant reduction of the fusion point of such materials. The slags formed are very reactive with certain other refractories. These effects are accelerated by the presence of sodium compounds. Another large class of compounds present in petroleum ashes is the alkali and alkaline earth metals. The effect of their salts on firebrick is shown in Table IV. Salts of magnesium, calcium, sodium, and potassium all lower the fusion point of the brick, with sodium being by far the worst offender. The potassium compound tested was not nearly as harmful as the sodium compounds. This observation was confirmed by tests using brick of other com positions. In no case was there any evidence of chemical attack on the kaolin-alumina base by the slags formed by the alkaline earth metals, indicating that these slags are not chemically active. The anions evidently play some role in determining the extent to which metals damage refractories. The carbonates and the sulfates appear to be particularly potent in lowering fusion point while chlorides (at least in the case of sodium) are considerably less harmful. Substitution of magnesium oxide for magnesium sulfate eliminated the effect entirely. The effect of a number of other metals known to occur in petro leum was determined in 50-50 mixtures of the oxide of the metal and firebrick. The data for the effect of various metal oxideon a fireclay brick are shown in Table V. Iron, tin, and lead are shown here to be detrimental to the firebrick. The remainder of. the metals tested are either not destructive at the temperatures measured or only very slightly destructive. In no case was there any evidence of chemical attack on the kaolin-alumina base. So far, the data have shown the effect of the components of petroleum ash on a single refractory brick. Table VI shows the effect of a typical ash on a number of different bricks. This table gives the chemical composition of the bricks investigated and the fusion point of 50-50 mixtures of these bricks with a petroleum ash derived from Lagunillas crude. Also included is the reactivity of the slags formed with the alumina-kaolin base used to support the test cones. In Table VI, the types of brick are arranged in order of increas ing cost (with the possible exception of some of the fireclay bricks, which are approximately the same grade). As a class, the fire clay bricks are subject to damage by petroleum ash containing vanadium. The reduction in their fusion points does not appear to bear any relation to the alumina content of the fireclay brick and is probably a function of the other constituents of the clay used. The method of manufacture which affects the porosity of the brick may also play an important part. The slags formed by the fusion of these firebricks were all chemically reactive to the kaolin-alumina base. Generally, fireclay brick as a class should not be used in furnaces where vanadium-bearing fuel oils are burned and there is direct impingement of Same on the refractory. The high alumina bricks, however, appear to be suitable for use in furnaces where such conditions occur. The fusion point Table V. Fusion Point of Brick with Various Metal Oxides (Mixture, 50% oxide and 50% Pennsylvania fireclay brick) Oxide Fusion Point, NiO CojOi FeiOj SnOt ZnO TiO* SbiO, PbO Cr*0* 2835 2555 2125 2100 Above 2910 2855 Above 2900 2230 2900 Table VI. . Effect of Petroleum Ash on Various Refractory Bricks (Mixture, 50% Lagunillas ash and 50% commercial brick) Fusion Point, Composition and Type of Brick 0 F. Reactivity with Kaolin- Alumina Base at Fusion Point Fireclay brick New Jersey clay, 15% AljOi New Jersey clay, 25% AljOi Pennsylvania clay, 39% AUOt Missouri clay, 44% AljO* Georgia JmoUn, 45% AlzOi High alumina brick 60% A1,Oj 70% AljOi 80% AIzO* 2710 2670 2695 2670 2555 Considerable decomposition Some decomposition Complete decomposition Complete decomposition Some decomposition 2800 2940 Above 3000 In these tests the ash melted and ran into the base causing considerable damage before the fusion point of the mixture was reached Magnesia brick Firebonded, 82-85% MgO Chemically bonded^ Firebonded, 90%+ MgOc Above 2855 Above 2910 Above 2900 Some damage to the base by melting of the ash; no slag formed " Classes and individual brick other than fireclay as listed in order of increasing cost. t Contained some chrome ore. e Contained 2.5% FetO,. November 1952 INDUSTRIAL AND ENGINEERING CHEMISTRY 2617 J.ixiiniuiKi ami Mississippi crudes North Louisiana Central Louisiana South Louisiana Southwest Louisiana Louisiana-Mississippi Heidelberg-Eucutta (Miss.) 01 la (La.) Texas crudes Conroe Yates-Pecos Imogene-West Texas East Texas Talco Salt Flat Coastal heavy Refugio light Midcontinent and Western crudes lllinois-Indiana Louden (III.) Oklahoma-Kansas Big Horn mixture (Wyo.) Venezuelan crudes Quiriquire San Joaquin Jusepin I.agunillas La Rosa medium Tia Juana Table VII. Analyses of Oil-Soluble Ash in Crude Oils Chemical, Wt. % on Ash U\i, U'l.% XiO NasO SjO* V,Oi I'c/L AljOj CaO M >(h SOj Spe< trogruphi*- Tru< e, 0.1-1.0 ht.',;, <>n ash I'rtx iu, <0.1 ut.' , on a-h 0 00 A 1 0 0008 0 0010 0 000.5 0.0011 0 0310 0 0100 17 09 17 2 12 01 24 72 40 .5 43. 1 9.2 13.4 01.7 16 3 35.6 43 10.6 33 7.0 3.4 5.0 15 8 8.8 0.3 29 0 0.1 2.7 8.6 0 8 1.0 28 J 9 11 0 1 1 0.1 12 9 23 0 30 1 12.8 42 1 .50 7 .59 6 \1. Mg Si. Al, On, Mg Al, C.t Ni, Ag, Pb On, >i 1 e, C:i, Mo Y, A), Ca Sr. Pb, Mn Cr, Cu, Ag. Pb, Mn Cr. Cu, Ag. Pb, Mn. Mir ( 'o Cr. Cu, Mr,. Mg, T, Ag. 1'b. Mg. Mn. >, Cu, Mg On, Mg. sr, Co 0 0007 42 0 7.1 0 2 16 5 2.7 13 2 28.9 Xi, Ph 0.0022 15.1 4 1 11.7 03.3 0.7 12.1 1.1 1 A 10.4 Ca 0.0000 3.3 33. 1 7.0 23.0 1.9 2.4 30.3 Ca, Mo 0.0030 3. 1 46.1 4 9 6.0 2.0 7.0 43.4 Ca, Co 0 0109 3.0 22.8 14.5 2.7 42 0 Al, Ca 0.002G 7.0 5 9 3.7 56.9 9.4 i i' 4 13 1 Ca 0 0073 4.0 33 2 2.3 1.5 .55 2 Ca 0.0003 17.9 7.7 3^9 24.1 9.8 2.2 10.0 Ca, B Cr, Cu, Mn, Mg. Ti, P, Cr, Pb, Mg, K Cu, Cr Cu, Pb, Mg Cr, Cu, Mg Cr, Cu, PI), Mn, Mg Al, Cu, Pb. Mg, Co Cr, Cu, Pb, Mn, Mg, M< 0.0075 2 3 28 6 9.1 2.2 0.0014 5.6 16 7 0.9 15.6 5 8 i! i L6 0.0049 2 0 33.3 7.0 10.9 1.0 3,3 2.8 0.0044 10.4 3.0 4.3 64.9 2 6 3.2 47 7 41 8 38 7 14.6 Al, Ca Ca ' Cu, Pb, Mg, Mn Cr, Cu, Pb, Mg, Mn, Sr Cr, Cu, Pb, Mg. Mn Cr, Cu, Pb, Mg, Mo 0.0236 0 0005 0.0160 0.0348 0.0394 0.0388 3.2 10.0 3.6 29.5 1.6 23.1 3.0 8.7 3.2 12.6 2.5 14.3 16.9 4.8 11.2 10.7 19.3 23.7 11.7 16.5 59.9 60.2 61.8 1.2 12.8 4.2 0.4 0.3 0.9 2 1 33.2 3.5 0.5 1.3 i.' 3 1.5 31.5 15.4 41.0 0.2 0.2 0.7 Si, Cu, Mo Cr, Mg, Pb Cu.Ca Mg of mixtures of these bricks and petroleum ash increases with the alumina content, but even the bricks of the lowest alumina con tent fuse only at or above 2800 F., the temperature level of furnace walls with flame impingement. When these bricks were tested, little or no slag was formed but the ash tended to melt away from the mix, leaving the cone structure intact. It appears, therefore, that no reactive slag would be formed. The magnesia bricks are at least the equivalent of the high alu mina brick and are probably suitable for even more severe service. None of these bricks fused with the petroleum ash. They are, however, more expensive than high alumina brick and would probably be used in only the most severe service. Summation. In brief, it may be concluded that petroleum ashes containing vanadium are very destructive to fireclay brick upon direct contact at high temperatures. The alkaline earth and alkali metals (particularly sodium) and certain other metals will lower the fusion point of refractories, but are not as destruc tive as ashes which include vanadium. In these cases the damage is probably limited to the surface and will usually cause a hard, glassy film on the brick which may result in failure by spalling if the brick is porous. The attack of vanadium com pounds is much more extensive. In such cases, extensive damage may occur from the fluxing action of the slag. There are several means of preventing or controlling such damage. When flame impingement cannot be eliminated, damage could be minimized by cooling the refractory. In cases where this is impossible, the proper selection of refractory brick will go far in eliminating destruction from petroleum ash. ANALYSES OF ASH OF CRUDES In view of the conclusions reached in the first part of this paper, information is required as to whether the metals reactive writh refractories occur in most crudes or whether the problem is peculiar to only a few crude sources. In order to answer this question, the analyses of 25 different crudes for their ash com ponents are presented. These crudes were procured, sampled, and analyzed under controlled conditions. The importance of vanadium compounds makes a knowledge of vanadium content of crude samples particularly pertinent. Vanadium analyses are available on a considerable number of crudes, and these are included as a separate listing. The data on the systematic study and the vanadium content of various crudes (all independent analyses) are given in Tables VII and VIII, respectively. The remainder of the discussion in this paper will be confined to the analyses obtained in the systematic study of the 25 crude samples. Table VUI. Vanadium Content of Various Crudes Texas crudes Light Refugio Coastal mixture Talco West Texas California crudes Ventura Ave. Signal Hill Lompoc San Joaquin Las Flores Bradley Sands Orcutt Santa Maria Purisma Nicolai Oxnard Illinois crude Louden Canadian crudes Leduc Redwater Mexican crudes Poza Rica El Plan El Burro Tonala Naranjos Filisola --Pjinuco Venezuela -rudes Jusepin Quiriquire Oficina heavy Lagunillas light Tia Juana La Rosa light Pedernalles Amacuro Lagunillas heavy Colombian crude Colombian Arabian crude Abqaiq Wt. % as V*0 0.0001 0.0001 0.0011 0.0016 0.0045 0.005 0.0067 0.008 0.019 0.024 0.029 0.036 0.039 0.044 0.072 0.0001 <0.0001 <0.0001 0.0019 0.0024 0.0037 0.0038 0.0072 0.0111 0.0138 0.003 0.007 0.023 0.027 0.030 0.037 0.040 0.050 0.057 0.018 <0.0001 2618 INDUSTRIAL AND ENGINEERING CHEMISTRY Vol. 44, No. 11 The agh content of petroleum fuels may be derived from a Dumber of different sources. These include the following: 1. The oil-soluble compounds which are present in the crude itself 2. The water-soluble (ompoumls which are usually salts and occur in solution in water originating with the crude (on distillation these water-soluble compounds are left, behind in the residual fractions as suspended salts) 3. Ash compounds from corrosion caused by the reaction of naphthenic and other acids present in the crude with the metal tanks and pipelines (like the oil-soluble ash, these are usually present as soluble matter in the residual fractions) 4. From contamination with scale, salt water, etc. 5. Caustic, carbonate*, etc., injected into distillation equip ment for control of corrosion or to neutralize acids Sources 2 to 5 are variable and individual determinations are meaningless because of this variation. For this reason the systematic study was limited to the determination of the oilsoluble constituents of petroleum ash. This had a further advan tage of concentrating the vanadium content of the crude since vanadium usually occurs as an oil-soluble metallo-organic com pound. The vanadium determinations of these crudes should therefore be particularly accurate. In order to avoid contamination, the crude samples were obtained from points as close to the producing wells as possible. In the case of individual crudes, specially cleaned glass containers were sent to the collection points and samples were taken under controlled conditions to avoid any possible contamination. In the case of crude mixtures from several different sources, the samples were taken from the pipeline carrying these mixtures. Since the object of the survey was to determine the oil-soluble metallic compounds present in the crudes, each sample was paper filtered to remove suspended matter and water. Tests showed that this procedure removed all traces of water and any compounds not soluble in oil. Each filtered crude was reduced to coke by burning at as low a temperature as possible. The coke thus produced was ignited at 1000 F. in platinum dishes to produce at least 100 mg. of ash. This procedure was repeated until consistent results were obtained. Ash contents and analyses are given in Table VII. Each of the ash samples was analyzed semiquantitatively in an emission spectrograph. This analysis reports the metals present in the ash as major (10 to 100%), minor (1 to 10%), trace (0.1 to 1%), and present (below 0.1%). Quantitative chemical analyses for all metals present in concentrations above 1% were made. These are reported in Table VII as the oxides. Sulfur determinations were also made and are reported as SO* content, but no analyses were made for other possible anions. Table VII also contains a list of the trace and present metals according to the semiquantitative spectrographic analysis. PETROLEUM--COMPOSITION. The results arc best summarized in Table IX which is based on the semiquantitative spectrographic analysis of the crudes. This table shows the percentage of this group of crudes containing each metal at various concentration levels. Table IX. Summary of Semiquantitative Spectrographic Analyses of Crudes* 1 2 Element0 Sodium Vanadium Iron Nickel Aluminum Silicon Calcium Molybdeuuin Copper Magnesium Lead Silver Chromium Manganese Percentage of Crudes Showing Concentrations of Major (10%) Minor Trace Present (1.0-10%) <0.1-1.0%) (0.1%) Not presei <J0 10 80 20 5 30 70 5 45 45 15 5 00 30 5 5 15 50 15 25 40 60 5 15 io 70 15 75 10 10 85 5 10 65 25 5 15 80 70 30 50 50 a Cobalt, strontium, boron, titanium, potassium, and platinum appeared as present in 20% or less of the samples; no other elements were reported. In general the quantitative results shown in Table VII check the semiquantitative results in Table IX quite well. Sodium and vanadium are shown to occur in the ash of most crudes in concentrations above 10% and in the ash of the remaining crudes in concentrations above 1 %. There are only a few cases where vanadium is below 1%. The next most prevalent elements are iron, nickel, aluminum, silicon, and calcium. Over half the crudes investigated contain traces of copper, magnesium, lead, chromium, and manganese, while a number of other elements such as cobalt, strontium, boron, titanium, potassium, platinum, silver, and molybdenum are present in isolated samples. These data show that fuel oils derived from most crudes con tain ash which is potentially destructive to refractories. The extent of actual attack in any given furnace will, of course, depend on the concentration of ash in the fuel oil but in time the ash from most crude sources will damage refractories in furnaces of im proper design. The importance of proper furnace design is emphasized, particularly as to the selection of refractory brick resistant to attack by ash components. LITERATURE CITED (1) "Handbook of Chemisty and Physics," 30th ed., p. 517, Cleve land, Chemical Rubber Co., 1948. (2) U. S. Bur. Mines, Bull. 334 (1931). Received for review May 14, 1952. Accepted September 11, 1952. Presented as part of the Symposium on Nonhydrocarbon Constituents of Petroleum before the Division of Petroleum Chemistry at the 121st Meeting of the American Chemical Society, Milwaukee, Wis., 1952. November 1952 Refractories: What They Are, How They Are Made, and What They Are Used For Refractories are important to an engineer. They not only meet his needs for heat resisting materials, but also provide him materials that will withstand such effects as abrasion, thermal shock, pressure, corrosion, and erosion. Through the years, a wide variety of refractory materials have been developed. Each has definite physical properties. This requires that an engineer be careful in his selection of a refractory for a specific appli cation. The environment in which a refractory will be placed first must be taken into account, then the refractory selected that has the physical properties necessary to with stand that environment. Along with an increase in the number of refractories have come improved methods for their manufacture. Manufacturing methods have evolved from hand molding to the currently used thixotropic casting process developed by AC Spark Plug Division. mmmmmm mmmmmm mmmmmm _tliltt____ f)Of)>f)f> mmmmmm mmmmmm mmmmmm mmmmmmj Corrosion resistance Thermal shock resistance ire is the basis of modern industrial alumina, mullite, fused magnesia, stabi Flife. The industrial use of fire is joined lized zirconia, and fusion cast refractories. to important economic considerations which, besides fuel costs, include the cost Value of Refractory of constructing and maintaining installa Determined by its tions in which fires are contained. Fur Physical Properties naces, kilns, and other heat treating The ability of a refractory material to equipment require construction materials that will withstand the effects of heat, flame erosion, and other conditions created by the industrial use of fire and withstand various service conditions is determined to a large extent by its chemi cal composition and physical properties. The most commonly determined physical other heat producing means. properties for refractory materials are as The need 'for heat and fire resistant follows: materials has led to the development of a class of products known as refractories. Pyrometric Cone Equivalent Broadly defined, refractories are materials capable of resisting heat over a wide Average bulk density temperature range. In addition to resist ing heat, however, a refractory often is Weight of a 9 in. by 4)4 in- by 2)4 required to resist corrosion, thermal in. brick gradients, thermal shock, load, and abra sion; attack by molten glass, metals, and Porosity i slags; and attack by atmospheres, either reducing or oxidizing, or both. Permeability The first refractory materials used in this country were naturally occurring Mean specific heat materials such as mica-schist or siliceous rock. Today, the majority of refractory Mean coefficient of expansion bricks and other shapes are manufac tured from fireclays (either bauxitic, Thermal conductivity kaolin, or flint types), silica minerals, magnesite, alumina, chromite, olivine, Modulus of rupture and graphite. Another class of present day refractories, called super refractories, Load test (heated) is manufactured from electric furnace materials or materials that are heated to Compressive strength high temperatures by other means. These materials include silicon carbide, fused Abrasion resistance Maximum recommended tempera ture. Refractories, such as refractory brick (Table I1), are tailored to have the maximum desired physical properties required for a specific application. For example, a refractory to be used in con tact with molten metal will generally be dense, impermeable, and resistant to abrasion and corrosion (Table II). It also will have low porosity, a relatively high termal conductivity, a high modu lus of rupture, and high thermal shock resistance. On the other hand, an insu lating refractory to be used as back-up in a location where it does not come into contact with metal or flame impingement will be porous, permeable, rather soft and non-resistant to abrasion, have a low thermal conductivity, a relatively low modulus of rupture, and moderate to good thermal shock resistance (Table III). The desired physical properties in re fractories are obtained by the selection and preparation of suitable materials; by the manufacturing process used; and by providing variations in composition, grain size, molding pressures, water con tent, and firing temperature. Physical Properties Explained An explanation of the meaning and importance of the physical properties of refractories is necessary to understand their significance to the industry and to the user. Pyrometric Com Equivalent When a solid material is transformed 30 GENERAL MOTORS ENGINEERING JOURNAL By ARTHUR P. WATTS and ARCHIE A. SHUKLE AC Spark Plug Division Physical properties an important factor when selecting a refractory to a liquid by means of heat, the transi tion may be referred to as the melting point, fusion temperature, or softening temperature. Most commercial refracto ries do not have true melting points, because they soften progressively over a temperature range within which the re fractory contains both crystalline mate rial and a liquid melt. The melting points of the various crystalline compo nents of a refractory, plus the amount and viscosity of any liquid which is formed, will determine the ability of a refractory to remain rigid at high tem peratures. The melting behavior of refractories varies greatly at high temperatures. Those refractories, consisting almost entirely of pure oxides in crystalline form, such as silica (SiOj), corundum (AliOj), and magnesite (MgO), have a relatively nar row melting range. Fireclay refractories and other refractories containing several oxides with various impurities fuse and soften gradually over a relatively wide range of temperatures. The high temperature softening beha vior of fireclay and high-alumina refrac tories is determined by the Pyrometric Cone Equivalent (P.C.E.) evaluation. This method consists of comparing test cones, made from a ground sample of the mate rial to be tested, with standard cones having known time-temperature soften ing \ alues. Both the test cones and stand ard cones are numbered, then mounted in a ceramic plaque at a slight angle to the vertical (Fig- 0- The plaque is heated at a definite rate until the test cones soften and bend. The number of the standard cone whose tip touches the plaque at the same time as the tip of the test cone is reported as the P.C.E . clue of the test cone. FIRST QUART! R 1 '64 Table I--Listed here are characteristic properties of various types and classes of refractory brick. In general, basic refractories are less resistant to thermal shodt than fireclay and high-alumina refractories. The spall resistance of some basic brick of special compositions, however, is equal to that of superduty fireclay brick. Basic brick of any class may be either metal encased or metal encased and internally plated. Such brick have better resistance to spalling than brick without the casings. Most metal-encased brick are of magnesite-chrome compositions. 31 1 PHYSICAL PROPERTIES PYROMETRIC CONE EQUIVALENT, Oft NOMINAL FUSION POINT (F) AVERAGE BULK DENSITY (G PER CU CM) WEIGHT OF x 4-1/2" x 2-1/2" BRICK (LB) POROSITY (PER CENT) PERMEABILITY TO COLD AIR MEAN SPECIFIC HEAT MEAN COEFFICIENT OF EXPANSION (IN. PER IN. PER C) THERMAL CONDUCTIVITY (BTU PW HR PER SQ FT PER IN. THICKNESS PER F) MODULUS Of RUFTURE (LB FES SQ IN.) LOW DUTY FIRE BRICK 19-29 2,770-3,018 1.9-2.15 MEDIUM DUTY FIRE BRICK 29-31 3,018-3,061 1.9-2.15 FIRE HIGH DUTY FIRE BRICK 32-33 3,123-3,169 1.9-2.2 CLAY PHYSICAL REFRAC TOR1E S SUPER V DUTY * FIRE BRICK 33-34 3,169-3,205 V50% Al203 34-35 3,205-3,245 60% *i2o3 36-37 3,279-3,308 70% AIj03 37-38 3,308-3,335 2.25-2.45 2.05-2.15 2.2-2.3 2.3-2.4 PROPERTIES V. ai203 38-40 3,335-3,425 2.6-2.7 ALUMINA 90-997. Al23 37-40 3,300-3,425 2.9 7 - 8-1/2 7 - 8-1/2 7 - 8-1/2 7-1/2 - 9 16-25 16-25 10-24 7-20 LOW TO HIGH LOW TO HIGH LOW TO HIGH moderate TO HIGH 0.25-0.26 20-1,000 C 0.25-0.26 20-1,000 C 0.25-0.26 20-1,000 C 0.25-0.26 20-1,000 C 7-1/2 - 9 18-24 MODERATE 0.26-0.27 20-1,000 C 8.2-8.S 20-27 MODERATE 0.26-0.27 20-1,000 C 8.4-8.8 20-35 VERY LOW TO LOW 0.26-0.27 20-1,000 C 9.5-10 20-30 VERY LOW TO LOW 0.26-0.27 20-1,000 C 10.5 14-24 3-51 0.28 0-1,700 C 5.1 x I0`6 20-1,200 C 8-9 AT 2,000 F 400-1,700 AT 70 F 5.1 x I0-6 20-1,200 C 9-10 AT 2,400 F 1,000-1,500 AT 70 F 5.1 x I0~6 20-1,200 C 10-11 AT 2,400 F 1,000-1,600 AT 70 F 5.1 x 10"6 20-1,200 C 5.5 x 10'6 20-1,200 C 10-11 AT 2,400 F 1,000-3,000 AT 70 F 10 - 12 AT 2,400 F 1,000-1,600 AT 70 F 6.5 x 10"6 20-1,200 C 10 AT 2,000 F 900-2,000 AT 70 F 6.5 x 10'6 20-1,200 C 7.0 x 10~4 20-1,200 C 10.5 AT 2,000 F 1,000-2,000 AT 70 F 12 AT 2.000 F 800-2,000 AT 70 F 7.3 TO 10.2 x lO-4 25-1,500 C 15-24 AT 2,200 F 1,200-4,600 AT 70 F HOT LOAD STRENGTH SUBSIDENCE IN 1-1/2 HR TEST UNDER 25 PSI LOAD (PER CENT) 3 - 10% AT 2,460 F 0.5 - 7.07. AT 2,460 F 0.5 - 7.0% AT 2,640 F * - 77. AT 2,640 F '.5 - 77. AT 2,640 F 0.5 - 3.07. AT 2,640 F 0.5 - 2.5% AT 2,640 F 0.3 - 2.07, AT 2,900 F COLD CRUSHING STRENGTH (PSI) ABRASION RESISTANCE 3,000 - 5,000 AT 70 F VARIES 2,000-3,000 AT 70 F VARIES 2,000-5,000 AT 70 F VARIES 1,500-4,000 AT 70 F VARIES 3,000-6,000 AT 70 F GENERALLY GOOD 2,000-4,500 AT 70 F GENERALLY GOOD 3,500-7,000 AT 70 F GENERALLY GOOD 3,500-6,000 AT 70 F GENERALLY GOOO 4,000-25,000 AT 70 F GOOD EROSION RESISTANCE VARIES VARIES VARIES VARIES GOOD GOOO GOOD GOOD GOOO THERMAL SHOCK RESISTANCE VARIES VARIES VARIES VARIES GOOD GOOD GOOD GOOD GOOO MAXIMUM RECOMMENDED HOT FACE TEMPBIATURE (F) 1,400-1,700 2,200-2,300 2,500-2,600 2,600-2,700 VARIES WITH VARIES WITH VARIES WITH VARIES WITH CONDITIONS CONDITIONS CONDITIONS CONDITIONS 2,000-3,400 AIRFLOW IN Table II--Values for the physical properties of dense, non-insulating refractories are listed here. These values, which are a composite of data obtained from information published by various manufacturers of refractories, represent a range of values and should not be considered as specific for any one refractory. The data for a commercial refractory may differ from the values given here because of differences in raw materials, bonding materials, and manufacturing process. The values for specific heat, thermal expansion, and thermal conductivity may be used in a general way for engineering purposes. The P.C.E. does not indicate a definite melting point or fusion point for a refrac tory, because the test is not a measure ment of temperature alone. P.C.E., how ever, does furnish a means of comparing the refractoriness of materials. Average Bulk Density The density of refractories is a factor that affects their capacity to store heat. The average bulk density of a refractory may be expressed as weight per cu ft or as grams per cu cm. The bulk density of a refractory is almost invariably less than the true specific gravity of the material composing^ it because of the effect of porosity. Porosity Porosity has an effect on a refractory's thermal conductivity, permeability to gases, and resistance to penetration by slags, molten metal, glasses, aaf| fluxes. Permeability Permeability to cold air is a relative measure of the number of open pores and their character. When a refractory has both high porosity and high permeability, a large number of connected pores is indicated. A high porosity and relatively low permeability indicates the structure consists largely of closed pores. Permea bility, therefore, is an indication of the extent to which the refractory will be penetrated by gases, fluid slags, metals, and fluxes. Mean Specific Heat The mean specific heat of a refractory material is a number that expresses the ratio between the amount of heat re quired to raise one gram of the material one degree Centigrade, between 0 C and its mean temperature, and the amount required to raise one gram of water from 19.5 C to 20.5 C. The units for mean specific heat in the English system are Btu per lb per deg F. The numerical value for mean specific heat is the same in both the English and metric systems for any refractory material. The heat content of a refractory brick or other shape at any temperature is the 32 GENERAL MOTORS ENGINEERING JOURNAL Ot D EN SE MU LUTE FUSION CAST ALUMINA RE RACTORl ES SILICON CARBIDE CLAY BONDED SiN BONDED ZIRCON 38-39 3,335-3,389 39-40 3,389-3,425 +42 2.3-2.6 2.B-3.2 2.54 2.85 3.3 S.5-9.5 10-12 9.25 10.25 12 15-22 2.5-7.5 13-14 7-9 15-30 STABILIZED ZIRCONIA 4,600-4,700 4.3-4.4 16 15-30 3-40 *0.5-2.0 '2 - 6 *1 - 4 *10-78 *4-32 0.25 20-1,000 C 0.28-0.32 0.285 20-1,400 C 0-1,400 C 0.29 0-1,400 C 0.18 20-1,050 C 0.175 25-1,400 C 4.5 TO 6.0 * io"6 25-1,400 C 8, 4 x IQ-6 25-1,400 C 4.4 x IO'6 25-1,400 C 4.4 x 10'6 25-1,400 C 4.2 x I0'6 20-1,550 C 8 x 10-6 0-1,100 C 14-16 500-2,000 F 1,200-1,800 AT 70 F 24-31 109 AT 2,200 F AT 2,200 F 113 AT 2,200 F 700-1,500 2,000 AT 2,462 F AT 2,500 F 5,000-7,000 AT 2,500 F 13.5 390-1,832 F 5 AT 2,000 F 1,500-2,200 AT 70 F 1900 AT 70 F SILICA REFRACTORIES QUARTZITE 31-33 3,061-3,169 1.65-1.9 6-7 19-30 HIGH 0.265 20-1,000 C SEMI-SILICA 27-31 2,984-3,061 1.85-2.0 6-3/4 - 7-1/4 23-30 MOOERATE TO HIGH 0.26 20-1,000 C MAGNESITE (BURNED) 38-41 3,335-3,578 2.6-2.75 9-1/2 - 10 20-28 LOW TO MODERATE 0.28 20-1,000 C BASIC REFRACTORIES MAGNESITE CHROMITE CHROME - (CHEMICALLY (BURNED) BONDED) FORSTER ITE VARIES 38-41 3,335-3,578 40 3,425 2.8-3.2 10-1/4 11-1/2 3.0-3.3 11-12 2.45-2.62 9 - 9-1/2 20-28 20-28 22-28 HIGH VARIES WITH TEMPERATURE *10-25 0.23 20-1,000 C 0.22 AT 1,000 C 0.27 20-1,000 C 43 X 10`6 20-300 C 3 x IO-6 300-1,100 C 13 390-1,832 F 600-1,200 AT 70 F 6.5 x 10"6 20-1,200 C 8 400-2,000 F 300-700 AT 70 F 14.5 x I0'6 20-1,425 C n x io-6 20-1,500 c: 8 IO'6 20-1,000 C 11 x 10`6 20-1,500 C 14-18 AT 2,000 F 1,200-4,000 AT 70 F 13.8 400-2,200 F 550-1,400 AT 70 F 12.8 AT 2,400 F 800-1,500 AT 70 F 10.3 AT 2,400 F 700-1,300 AT 70 F 0 - 2.0"/. AT 3,100 F 00 AT 2,730 F AT 2,730 F 0 AT 2,730 F 2 - 5% 2,910 F LESS THAN 1% AT 15 PSI AT 1,815 C PER 100 MIN FAILS AT 2,900-3,070 F 0.5 - 2.0% AT 2,640 F FAILS AT FAILS AT 2,900-3,200 F 3,000 F FAILS AT FAILS AT 2,300-2,700 F 2,950-3,050 F 5,000-7,000 AT 70 F FAIR FAIR TO GOOD 10,000- 15,000 AT 70 F 15,00020,000 AT 70 F EXCRLENT EXCELLENT 5,00012,000 AT 70 F FAIR FAIR 2,000-4,000 AT 70 F GOOD 1,400-3,000 AT 70 F GOOD 6,000-12.000 AT 70 F USUALLY POOR 1,800-6,000 AT 70 F 3,000-5,000 AT 70 F -- 1,500-4,000 AT 70 F FAIR GOOD GOOD RESISTANT TO ACIDS ACIDS: GOOD ALKALIES: POOR RESISTANT TO ACK> AND SILICEOUS SLAGS RESISTANT TO ALKALIES AND MOST METALS RESISTANT TO ACID SLAGS RESISTANT TO ALKALI FUMES AND DUST RESISTANT TO BASIC SLAGS RESISTANT TO BASIC SLAGS VERY GOOD POOR TO FAIR FAIR TO GOOD VERY GOOD GOOD FAIR POOR BELOW 1,200 F. GOOD ABOVE 1,200 F GOOD FAIR GOOD 3,000-3,200 3,300 2,500-3,000 2,900-3,000 3,000 4,250 F PROPBLY SUPPORTED 2,900-3,000 2,800-2,900 VARIES VARIES CU FT PER HR SQ FT PER IN. FOR ONE INCH H^O PRESSURE DROP AS REPORTED BY REFRACTORIES DIVISION, CARBORUNDUM CORPORATION RESISTANT TO BASIC AND MODERATELY ACID SLAGS RESISTANT TO ALKALIES AND BASIC SLAGS FAIR FAIR VARIES VARIES product of its specific heat and its weight. This product is used when calculating the heat stored in a furnace structure. Coefficient of Thermal Conductivity The thermal conductivity of refractory materials is affected mainly by the chem ical and mineral composition of the refractory, the structure of the material, its porosity, and its; temperature. In general, the thermal conductivity of any one particular type of refractory is more dependent upon the porosity than any other factor. ThermaJ conductivity de creases with increasing porosity. The formula for heat flow through a refrac tory wall in service under steady-state conditions is: n _ k (Ti - T,)At FIRST QUARTER 1964 where Q heat flow k coefficient of thermal conduc tivity for the refractory material Tx - Ti temperature drop from the hot face to the cold face A area of the wall t time d = thickness of the wall. In English units, k representsjjie quan tity of heat in British thermal units which will flow through the refractory per hour, per sq ft of surface, per degree F temperature difference between the hotter and the colder faces, and for a one-in. thickness. Modulus of Rupture The modulus of rupture is a measure of the strength of a refractory. The modulus usually is determined at room temperature, although it may be deter mined at some higher temperature if desired. When determined at room tem perature, the cold, strength of the refrac tory is the quantity measured. The strength of a refractory at room tem perature may have little or no relationship to the strength of the same refractory at higher temperatures in the normal oper ating range. Modulus of rupture deter minations at room temperature may serve to indicate whether a refractory of a particular type has been properly processed and fired. Load Test Subsidence under load at high tem perature is a standard laboratory test for determining the behavior of a refractory at or near its normal operating tempera- 33 PHYSICAL PROPERTIES OF PHYSIOL PROPERTIES AVERAGE BULK DENSITY (G PER CU CM) WEIGHT OF 9" x 4-1/T x 2-1/2" BRICK {IB) POROSITY (PER CENT) 1,600 F 0. 30-0.47 1.1-1.7 2,000 F 2,300 F 0.46-0.73 1.65-2.65 65-80 0.50-0.78 1.6-2.85 65-75 INSULATING REFRACTORIES CLASSIFICATION 2/00 F 2,800 F 3,000 F 3,200 F 0.67-0.82 0.69-1.00 0.91-1.13 1.27 3,300 F ALUMINA BUBBLE BRICK 1.29-1.45 2.45-3.00 2.50-3/5 3.3-4.1 4.6 4.7-5.3 63-73 50-70 50-65 50-64 50-67 FUSEO ZIRCONIA SILICA INSULATING INSULATING 2.45 0.91-0.98 9.00 3.3-3.6 PERMEABILITY TO COLD AIR FLOW (IN.3 PE* SEC PER IN.2 PER IN. THICKNESS PER LB PRESSURE) 5.9-30.2 6.7-39.2 12.2-42.7 -- -- MEAN SPECIFIC HEAT MEAN COEFFICIENT OF EXPANSION (IN. PER IN. PER F) THERMAL CONDUCTIVITY (8TU PER HR PE* SO FT PER IN. THICKNESS PER F) 0.230 AT 1,000 F 2.5 x 10"6 AT 1,000 F 0.95-1.63 AT 1,000 F 0.245 AT 1,500 F 0.245 AT 1,500 F 0.260 AT 2,000 F 2.65 x 10'6 2.65 x 10*6 2.5 10*6 AT 1,500 F AT 1,500 F AT 2,000 F 1.28-2.10 1.43-2.20 2.82-3.20 AT 1,500 F AT 1/00 F AT 2,000 F 0.260 AT 2,000 F 3.1 x 10"6 AT 2,000 F 2.60-3.60 AT 2,000 F 0.260 AT 2,000 F 3.3 x 10"6 AT 2,000 F 3.6-4.1 AT 2,000 F 0.270 AT 2,400 F 4.5-5.5 AT 2,400 F 0.320 70-2,550 F 5.0 x I0"6 AT 2,400 F 5.0-7.0 AT 2,400 F 0.175 70-2,550 F 0.270 2,000 F 5.1 x 10'6 SAME AS HIGH 85 - 2,730 F OENSITY SILICA 5.0 5.5 AT 2,000 F AT 2,000 F MODULUS OF RUPTURE (LB PER SO IN.) 40-100 75-140 80-155 100-350 150-450 175-300 450-750 150-450 300 160-200 HOT LOAD STRENGTH - SUBSIDENCE IN 1-1/2 HR TEST UNDER 10 PSI LOAD (PER CENT) 0.0% AT 2,000 F 0.0-0/% 0.3-O.5% 0.0-4.0% 0.0-4.0% 0.5-2.0V# 0.0-0.7% AT 2,000 F AT 2,200 F AT 2,400 F AT 2,600 F AT 2,600 F AT 2,800 F COLD CRUSHING STRENGTH (PSI) 50-110 90-160 100-210 150-450 350-900 400-750 1,000-1/00 350-1,300 150-200 ABRASION RESISTANCE POOR POOR POOR POOR POOR POOR POOR POOR POOR POOR CORROSION RESISTANCE REDUCED ABOVE 2,900 F IN CONTACT WITH CARBON RESISTANT TO ACIO SLAGS AND IRON OXIDE THERMAL SHOCK RESISTANCE GENERALLY GOOD GENERALLY GOOO GENERALLY GENERALLY GOOD GOOD GENERALLY GENERALLY GOOO GOOO GENERALLY GOOD GOOD GOOO ABOVE 1,200 F . MAXIMUM RECOMMENDED HOT FACE 1 TEMPERATURE (F) EXPOSEO 1/00 F BACKUP 2,000 F 2,000 F 2,300 F 2/00 F 2,800 F 3,000 F 3,200 F 2,8003/00 F 4,000 F PROPERLY SUPPORTED 3,000 F Table III--Insulating refractories, whose physical properties are listed here, are used to withstand temperature and act as insulation. These refrac tories are classified primarily by division in groups suitable for specified maximum recom mended hot face temperatures. The groups are listed as 16, 20, 23, 26, 28, and 30; When multi plied by 100, the group number gives the recom mended temperature. Insulating refractories are light weight and porous. The weight increases and porosity decreases as the group number increases. When used in furnace structures, insulating refrac tories give reduced heat storage due to their limiter weight and a reduced heat toss from the outer surfaces of the furnace due to low thermal conductivity. The values listed here are a com posite of values obtained from several manufac turers of refractories and do not apply specifically to the products of any one manufacturer. ture. When conducted according to the American Society for Testing and Mate rials method Cl 6, the temperature is raised at a prescribed rate until a specified holding temperature is reached. The specified holding temperature differs according to the classification of the refractory undergoing test, ranging from 2,460 F for high duty fireclay to 3,000 F for 99 per cent alumina brick. A vertical load of 25 psi is applied to a brick standing on its end. In most tests the maximum temperature is maintained for 1 }/> hours, with the furnace then being allowed to cool by radiation to 1,830 F or lower before the load is removed. The contraction or amount of subsidence is reported as a percentage of the original length. Variations of the load test are made at times by some manufacturers to suit specialized products. In normal service, refractories ,are not generally required to bear loads of 25 psi. The load test, therefore, subjects the brick to conditions that are more severe than normal and thus accelerates plastic deformation. Cold Crushing Strength Cold crushing strength is admeasure of the strength of a refractory under compression at room temperature. Here, as with modulus of rupture, the cold strength of a refractory cannot be relied upon to indicate its behavior at high temperature. The cold strength, how ever, is a valuable measure of resistance to impact and ability to withstand the abuse encountered in handling and ship ping. It also may be of value in determin ing whether processing and firing pro cedures have been satisfactory. Abrasion Resistance While the ability of refractories to resist abrasion is important in some appli cations, it may be of no importance in others. Refractories that must resist abra sion are intentionally made to be hard, well bonded, and strong. Abrasion re sistance is most often measured by im pingement of a stream of abrasive par ticles against the face of a test brick, or by rubbing against an abrasive surface. The modulus of rupture test also is an indication of the relative resistance to abrasion. Corrosion Resistance Refractories may be subjected to a corrosive environment due to chemical reactions that occur when they are in contact with metals, slags, fluxes, dust, furnace gases, combustion products of fuels, and other refractories. As a gen eral rule, fireclay, alumina, and silica refractories Eire considered to be resistant to acid fluxes and slags, while magnesia, 34 GENERAL MOTORS ENGINEERING JOURNAL chrome, and forsterite are considered to be resistant to basic fluxes and slags. No definite rule can be given, however, because conditions vary greatly. The applicadons best suited to certain re fractory types have been established dirough long experience and a process of trial and error. Manufacturers of refractories can generally recommend die type suitable for a particular purpose. Thermal Shock Resistance Thermal shock resulting from abrupt and rapid changes in temperature can cause rapid deterioration of a refractory due to the generation of cracks and the loss of pieces from the face of the refrac tory. The cracking and spalling are caused by mechanical stresses that result from non-uniform temperature changes that occur throughout the piece. Resistance to thermal shock can be increased by a combination of physical properties that tend to reduce the mag nitude of the stresses occurring during a temperature change. Properties that pro mote resistance to thermal shock, for example, are low thermal expansion, high thermal conductivity, a structure having some elasticity, and the absence of crystalline phase changes. In many applications thermal shock resistance is of minor importance. For example, the operating temperature of a furnace or kiln may be slowly raised, then held there for years. In other appli cations such as kiln cars, which are continually going from room temperature to the operating temperature of the kiln and then back to room temperature, thermal shock resistance is of vital impor tance. The present trend towards short ening the time cycle in kilns and other industrial furnaces also has made thermal shock resistance of major importance for many applications. Maximum Safe Temperature Manufacturers of refractories usually specify a maximum safe operating tem perature for their products. This is done, however, only when the manufacturer knows the exact environmental condi tions to which his product will be subjected. Environment and operating variables exert so much influence upon the behavior of refractories that a maxi mum recommended temperature cannot always be given. Long experience has made it possible, however, to correlate industrial use of refractories with lab Fig. I--Illustrated here are test cones and standard cones used in the Pyrometric Cone Equivalent evaluation method to determine the temperature softening behavior of fireclay and high-alumina refrac tories. Cones I. 2. 3, and 4 are test cones and cones 32 through 39 are standard cones. oratory tests to discover the temperature range where failure is apt to occur. Various Methods Used to Manufacture Refractories Several methods are used to manufac ture refractories. Some of the methods are well established practices, while others are of comparatively recent origin. Pressing Press forming of refractories is done by manual, power driven, or hydraulic presses. In most cases where pressing is used, the material is granular and con tains relatively small amounts of moisture. Various binders are added to famish a cold bond that aids in handling the shape after it is out of the press ajSS' in the drying and firing operations. Hydraulic presses are used to exert the high pres sures necessary when extremely high density and strength properties are neces sary. Extrusion Materials that are sufficiendy plastic can be extruded as a column from a die through which the material is forced. As the column issues from the die it is cut into appropriate lengths for the subsequent reforming operations usually needed to obtain the desired shapes. Slip Casting In slip casting, the refractory ingre dients are mixed with water to form a slurry which then is poured into a porous plaster mold. The mold absorbs water from the slurry, causing a layer of the refractory material to be formed on the inside surface of the mold cavity. Either solid or hollow shapes can be formed by the slip casting method. Air Ramming Air-operated rammers or vibrating hammers can be used to form refractory shapes in molds made of steel or wood. The process is used primarily for special shapes, or when conditions require only a limited number of pieces. The material usually is rammed in layers. The layers are bonded to one another by roughening the surface of the layer just rammed before the next layer of material is introduced into the mold. While good pieces can be made by this method if FIRST QUARTER 1964 33 REFRACTORY MATERIAL FLOWS INTO MOLD CAVITIES Fig. 2--The thixotropic casting process developed by AC Spark Plug Division is used to produce a variety of refractory products. Shown here are the principal steps involved in casting a sagger by the thixotropic method. care is used, it is impossible to have uniform density throughout the piece; therefore, other methods are considered more desirable. Hand Molding Hand molding formerly was a common method used to make special refractory shapes. This method is used today only when the cost of dies and molds for machine operations is not justified. Drying and Firing Refractories must be thoroughly dry before firing. Even a small amount of moisture is apt to cause bursting and cracks when the pieces are heated. The firing process is done either in periodic or tunnel kilns which produce a timetemperature cycle that has been found to be satisfactory for the particular con tent. In general, firing temperatures for refractories are sufficiently high? to initiate chemical and physical changes that result in the development of the desii;l struc ture and composition, the elimination of volume changes, and the development of strength required for handling, shipping, and enduring structural loads in furnaces. AC Develops Unique Process for Manufacturing Refractories AC Spark Plug Division has had long experience in the manufacture of refrac tories, both for its own use and for others. AC was originally a ceramic plant, and the manufacture of saggers (ceramic boxes used to contain spark plug insula tors during the firing process) and kiln refractories was continued after it became a Division of General Motors. The saggers were originally made of a plastic refrac tory composition containing a fairly high percentage of clay. The saggers were manufactured by forming them in steel dies on a steam-powered press. When AC became interested in manu facturing alumina spark plug insulators in the early 1930's, the high temperature refractories required for a kiln operating in the temperature range of 3,000 F to 3,200 F were not available in this coun try. It became necessary, therefore, for AC to produce its own high temperature refractories for the kiln and kiln cars. 36 GENERAL MOTORS ENGINEERING JOURNAL CORE EXTRACTED FROM MOLD TYPICAL REFRACTORIES MADE BY THIXOTROPIC CASTING PROCESS t LOADED TUNNEL KILN CAR WITH SAGGERS. THE GIRDERS, CAR TOP BLOCKS, AND CAR BASE BLOCKS ALSO ARE MADE BY THE THIXOTROPIC CASTING PROCESS. At first, the refractories were hand molded. Bricks and other shapes also were made by hydraulic pressing meth ods. Hand molding then was replaced by air hammer ramming. Later on, AC developed a variation of slip casting known as the thixotropic casting process. Both the process and composition, pat ented by AC, now are used to produce refractory brick, shapes, and saggers used in kilns operating at temperatures up to 3,200 F and to make refractories for use in the power house and die cast areas. The thixotropic casting process also is used to make specialized refractories for unusual applications. The thixotropic casting process makes use of electrolytes and surface active agents to obtain high density mixtures containing relatively little water. The mixtures, however, have sufficient fluidity to flow into and fill comparatively small spaces in a mold (Fig. 2), especially when the mold is vibrated. After the mold is filled, it is allowed to stand for a period of time until the refractory material has stiffened sufficiently to permit the piece to be removed from the mold. The removal of the piece from the mold must be done carefully because any appreciable vibration or movement of the material after the initial setting will make it fluid again due to its thixotropic quality. Summary Refractories represent a class of engi neering materials that possess character istics well suited to applications where resistance to heat, abrasion, corrosion, thermal shock, and attack by molten glass, metals, slags, and reducing or oxidizing atmospheres is required. To apply refractory materials properly, how ever, the engineer must be familiar with their varied physical properties and char acteristics. Knowing the environmental conditions in which a refractory will be placed is a primary requirement in the proper selection of a refractory material. This will indicate the type or class of refractory material required. Examina tion of physical properties then will nar row the selection to the proper refractory material to use. References 1...............Modern Refractory Practice (Pittsburgh, Pennsylvania: HarbisonWalker Refractories Company) 4th Edition, 1961. FIRST QUARTER 1964 37