REFRACTORIES MANUFACTURERS ASSOCIATION’S INDUSTRIAL FELLOWSHIP No. 1 SPECIAL REPORT A STUDY OF THE TUNNEL KILN AND ITS APPLICATION TO THE BURNING OF REFRACTORIES BY RAYMOND M. HOWE INDUSTRIAL FELLOW MELLON INSTITUTE OF INDUSTRIAL RESEARCH UNIVERSITY OF PITTSBURGH PITTSBURGH, PENNSYLVANIA SEPTEMBER 19, 19 17 PRESS OF REED a WITTING CO. PITTSBURGH Digitized by the Internet Archive in 2017 with funding from University of Illinois Urbana-Champaign Alternates https://archive.org/details/studyoftunnelkilOOhowe REFRACTORIES MANUFACTURERS & ASSOCIATION’S • INDUSTRIAL FELLOWSHIP No. 1 SPECIAL REPORT A STUDY OF THE TUNNEL KILN AND ITS APPLICATION TO THE BURNING OF REFRACTORIES T HERE is in all probability no more vital a problem in the manufacture of refractories than that of burning. The ware must be burned uniformly, thoroughly, and as quickly as possible, in order to secure the maximum capacity of the kilns. At the present time many plants are either forced to use a poor quality of coal, pay exhorbitant prices, or, in some cases, to work their own mines. Under such conditions, which probably will not be overcome for some time, the manufacturer naturally wishes more than ever before to obtain the maximum output from the coal used. Exact figures are not available which give the percentage of heat used in maturing the ware, that which passes out of the stack, the radiation loss, etc. It is known, however, that only five to twenty per cent, of the heat liberated in intermittent kilns used for other purposes actually enters the ware. From thirty-three to seventy-one per cent, is lost by radiation and in the kiln, while another twenty to fifty-five per cent, is carried out by the hot stack gases. It is the latter loss that is most directly overcome by con¬ tinuous kilns. Continuous kilns of the tunnel type also have a Very low loss due to radiation as well as a small loss to the kiln. The hot waste gases are used to “water-smoke” the incoming ware, heat it to the temperature at which the combined water is driven off, and even to start the various chemical reactions incident to the development of a finished refractory. 799751 A kiln capable of doing this has necessarily attracted much attention. The General Electric Company has investigated the kiln abroad and in this country. The Bureau of Standards has made a heat balance of the tunnel kiln at Keasbey. Many other companies have applied the kiln to the production of various kinds of ceramic and metallurgical ware. Mr. L. E. Barringer has ably described the kiln* as follows (Trans. Am. Cer. Soc., 18, 111-116): The kiln is simply a straight tunnel, 197 feet in length, with a fixed firing zone somewhat nearer the discharge end than the charging end. A track extends through the kiln and upon this the cars are moved, being gradually brought up to the firing zone, given the necessary period of exposure to the maximum temperature and then passed to the cooling end of the kiln. By referring to Figure A the general operation of the kiln may be explained as follows: At the charging end of the kiln “A”, loaded cars are periodic¬ ally “fed” into the kiln, being first placed in a vestibule. This vestibule is necessary to prevent disturbance of the draft while the inner door of the kiln is opened to receive the loaded cars. After the entering car is in the vestibule and the outer door closed, the inner door of the kiln is opened and the car moved into the kiln proper, pushing all preceding cars before it. Movement of the cars is accomplished by a hand-windlass operating an endless chain just below the cars at the charging end of the kiln. This chain is provided with hook lugs which engage the axles of the car and move it forward as the windlass is operated. To facilitate the pushing of the entire train of cars, the track through the kiln floor has a slight drop from the charg¬ ing end to the discharging end. As each additional car of ware is placed in the kiln, the entire “train” of cars in the kiln progresses car-length by car- length, gradually passing through zone B, which for some wares might be called the “water-smoking” zone, into zone C and thence into the firing zone D, where the highest heat is imparted to the ware. After the car of ware has remained in the firing zone a sufficient length of time to bring about the desired shrink- *Tunnel kiln of the Didier-March type. 4 age in the ware, as indicated by pyrometric cones or other means, it passes into the cooling zone E, where the material is cooled slowly and uniformly. The car leaves the kiln at F in such con¬ dition that the material can at once be taken to storage and un¬ loaded, or, if somewhat too hot, can at least be sidetracked without interference with the operation of the kiln. This is an advantage in favor of the tunnel kiln, since in periodic kilns too hot to unload, the kiln space must be occupied until the jvare can be taken out. At a point “G” between the charging and discharging ends of the kiln the four grates for direct firing are located, two on either side, and form the base of the combustion chamber. This combustion chamber is separated from the main tunnel by an inner wall, and the products of combustion pass through the tuyeres H and enter the main tunnel where they become diffused and circulate around the ware, successively passing through and into the zones C, D, and B. At the point J the waste products of combustion and steam pass out of the kiln into the stack. Air for combustion enters the kiln through ports in the walls at the discharge end. The air passes through “serpentine’’ flues and becomes preheated by contact with the cooling ware. As the ware cools, the heat units are given up to the in-coming cold air, and this preheated air passes on until when nearly to the firing zone it is divided, part of the air passing under the grate bars to be used for combustion (primary air) and part of the air being deflected into the kiln chamber to complete combustion of the fuel gases (secondary air). The heated combustion gases pass through the tunnel towards the charging end and heat up the in-coming ware, leaving the kiln through the stack flue at a point about ten feet from the charging end of the kiln, as in¬ dicated by J. The kiln has a capacity of 36 cars, and one car is entered and one removed from the kiln every two hours. A car thus requires seventy-two hours to pass through the tunnel. An important feature of the cars is the “sand seal’’ to pre¬ vent the heat of the kiln from finding its way beneath the cars where it would soon damage the wheels and axles, to say nothing of disturbing the draft. This consists of sheet-iron aprons or 6 shields which move through troughs of sand on either side of the tunnel wall. The tunnel is provided with a “subway” which extends the entire length of the kiln beneath the cars and from which the heat is excluded by the “sand seal.” Youghiogheny coal is used at Keasbey for firing, and the four fire boxes are fired one at a time every fifteen minutes, so as to maintain as closely as possible the same conditions of oxida¬ tion and reduction, as well as temperature. About three tons of coal are used daily. The kiln at Schenectady is used in the firing of porcelain insulators to a temperature of 1300-1400°C., this class of ware of course all being fired in saggers. The kilns at Keasbey are used in the firing of fire-brick and refractories at practically the same temperature, and in both cases the uniformity of distribu¬ tion of temperature is well within one cone. As an indication of the efficiency of the kiln it may be noted that the temperature of the escaping gases is very low, averaging 200-220°C. at Schenectady and 1 50-200°C. at the damper in the Keasbey kilns. In the kilns at Keasbey 650 pounds of coal are used to burn 1000 standard fire-brick. The continuous kiln turns out about as much ware as three and one-half fourteen-foot periodic kilns. The cost of coal is about one-half that for the periodic kilns of the same capacity. Labor cost has been found to be 10 per cent, less for loading and unloading, and the labor cost of firing is the same as in periodic kilns. Such are the results of Barringer’s investigation on this type of kiln. Since the publication of his article, however, several questions have arisen. What is the temperature attained? How uniform is this temperature in various parts of the kiln? Are the refractories thoroughly burned to final shrinkage? Do any strains result from the rapid application of heat? What is the breakage, the saving in fuel, the cost of setting, the cost of un¬ loading and the cost of firing? Are there many ‘Turnovers?” Finally, is the kiln as efficient as possible? 7 In order to answer these questions a study of the tunnel kiln at Keasbey was undertaken*. This kiln was operating at the proper temperature, and was turning out refractory ware. The results obtained should apply to any well controlled tunnel kiln. Fire-brick, both burned and green, were secured which were typical for the various parts of the countryf. The green brick were measured with a micrometer, labeled, and placed in eleven cars. These cars were introduced at various intervals so that the first test was extended over a period of six days. The cones on top of the car were often vesicular in structure, being made from natural clay, and so these were compared with Orton cones. Cones were placed with all of the brick, and forty- one sets were scattered so as to measure all possible temperature conditions in the more sheltered interior of the car. After the samples were burned they were again measured and then tested for porosity, apparent specific gravity, crushing strength, and traverse strength. This same data were secured from the stock brick on hand, it being necessary, however, to calculate the fire shrinkage of these from the average dimensions of the green brick. The crushing strength was calculated in pounds per square inch, the tests being made on the sides of the brick. These tests were made in order to determine whether the bond had been as thoroughly developed in the tunnel kiln as in the periodic kilns. The traverse strength tests were made between two parallel fulcrums, six inches apart. It is given in pounds per brick. This test was designed as a means of discovering any cracks which might be developed. *The writer wishes to acknowledge his indebtedness to the Didier-March Company, and especially to Mr. G. A. Balz, for hearty co-operation in carry¬ ing out this investigation. fSome of these have been delayed in transit and have not been tested at this time. These will be available in a short time. 8 Results of the Tests The temperature variations found are shown diagramatically. Some of the cars were burned to cone 13 and some were burned to cone 14. In order to simplify the tabulation of results each car was considered as having been burned to cone 14. The neces¬ sary corrections were made. The results are given below. At no time did the cars receive a burn of less than cone 13. In order, then, to calculate the minimum temperatures, simply subtract one cone from the temperatures given in the diagrams. The physical tests have likewise been tabulated in as simple a form as possible. From 12 to 24 measurements of the dimen¬ sions of the brick were made as the foundation of the fire shrink¬ age calculations. The porosity tests were made on four samples of each brand. Crushing strength tests were made on from 4 to 12 samples of each brand, more being made on the bricks which showed a difference in quality due to the two methods of burning. The average results are given in the following table: Cone Equivalents in Temperature Degrees Cone Numbers Centigrade Degrees Fahrenheit Degrees 10 1330 2426 11 1350 2462 12 1370 2498 13 1390 2534 14 1410 2570 Average Physical Tests Sample Percentage Porosity Per Cent. Fire Shrinkage Crushing Strength in Lbs. per Sq. In. T raverse Strength in Lbs. per Brick Stock Brick Tunnel Kiln Brick Stock Brick Tunnel Kiln Brick Stock Brick Tunnel Kiln Brick Stock Brick Tunnel Kiln Brick I 28.2 27.7 .70 1.53 1890 1833 1100 1140 II 24.1 22.1 3.23 3.68 2029 2096 1185 1670 III 21 1 20.6 3.05 2.24 1750 1919 1550 1450 IV 26.3 25.6 3.89 4.75 1772 1621 1450 1440 V 17.7 17.8 2.18 2.30 2684 1117 1570 1145 VI 23.1 20.8 2.86 4.57 2044 1818 1275 1350 VII 21.5 23.1 1 .61 1 .47 1337 819 1220 1250 VIII 17.4 20.3 4.67 3.78 5846 5204 3250 3050 9 LAYER I—(BOTTOM OF CAR) 10 LAYER IV 11 LAYER VII 13- 13- 13 13 13 LAYER IX 12- 12 + 12 + 12 LAYER X LAYER XI 13 LAYER XII LAYER XIII—(TOP OF CAR) 14 Discussion of Cone Tests The variation of cones which was found is much greater than that given by Barringer. This may be easily explained. At the time of his investigation a ware was being burned which required a uniform temperature, which was obtained by careful firing and the use of new cars. The present investigation was conducted under more adverse conditions. The cars were in need of repair and allowed cold air to seep up from the “subway” below. The kiln was not fired as carefully. Furthermore it had been running at maximum capacity for four years. Since the ware was mature at cone 10 no attempt was made to control the burning more carefully, this variation being considered per¬ missible. Discussion of Physical Tests I. These brick were made by the stiff mud process, using an auger machine, and were steam repressed. New Jersey fire¬ clay, grog and gannister were used in the batch. The ware burned in the tunnel kiln showed the same physical properties as the ware from the intermittent kilns. The slight difference observed was more likely due to experimental error rather than conditions of burning. II. These brick were made by the soft mud process from 13 per cent, bond clay and 83 per cent, flint clay, part of which had been calcined. The clays were from the Cambria district. The ware coming from the tunnel kiln showed greater shrinkage, greater strength, and lower porosity, all of which indicated that the tunnel kiln burn was more thorough. III. These brick were much the same as those of type B— being higher in flint clay. The ware from the two types of kilns was essentially the same. IV. These brick were from the Southern Ohio district. The two burns showed practically the same quality of ware. V. These brick contained twenty per cent, bond clay and eighty per cent, flint clay, part of which was one-half an inch in diameter. These large particles of flint clay often protruded from the sides of the firebrick as they came from the tunnel kiln. 15 A great many surface cracks were found in the brick coming from the tunnel kiln. The brick were also weaker than those from the regular burn, showing that the surface cracks evidently extended throughout the brick. VI. These brick contained ninety per cent, flint clay and ten per cent, bond clay. They showed that they had received a harder burn in the tunnel kiln than in the intermittent kiln. VII. These brick were from the Pennsylvania-Ohio dis¬ trict and were made by the soft mud process, twenty-five per cent, bond clay and seventy-five per cent, flint clay. The brick were discolored and cracked as they came from the tunnel kiln, showing a large amount of iron both on the surface of and throughout the brick. The ware from the intermittent kiln showed some traces of iron but it was not discolored as was the ware from the tunnel kiln. VIII. These brick were made with an auger machine from the same batch as given in VII. The stock brick were nearly white, dense and very strong. The ware coming from the tunnel kiln was spotted and slightly weaker. General Conclusions I. Fire-brick made from calcined clay grog and bond clay may be successfully fired in the tunnel kiln. The quality of the ware coming from the tunnel kiln is of a quality equal to that coming from the periodic kiln. II. Fire-brick made from flint clay and bond clay may be successfully fired in the tunnel kiln. The flint clay, however, should not be introduced in too large particles if the ware is to be burned and cooled in seventy-two hours. If some of the flint clay has been previously calcined very good results are secured. III. When the bond clay becomes dense at a low tempera¬ ture, difficulties are liable to arise when such a clay is burned in the tunnel kiln. Bricks of types seven and eight show this. It is believed that such brick, those which become very dense, should be fired to a lower temperature, or else be fired in a longer tunnel kiln. Either one of these porcedures, if followed, would result in the lessening of the severity of the heat treatment. 16 IV. Iron, if present, results in the formation of dark spots on the surface of the brick coming from the tunnel kiln. If an excess of air is present, such as is found in the periodic kiln, it oxidizes the iron and its color is less pronounced. This excess of air lowers the efficiency of the kiln. As a result, if the high efficiency of the tunnel kiln is to be retained, such iron spots are unavoidable. V. Brick were also sent through the tunnel kiln which were made by the dry press method. They were packed in such a way that the original moisture content was present when the bricks were introduced into the kilns. The results were not satisfactory for two reasons. The brick are ordinarily burned to cone 10; here they were burned to cone 14 and in a very short time. As a result cracks were developed. Were the kiln longer the brick would not have been subjected to such a strain during the water¬ smoking period. It is believed that by using a longer kiln that dry pressed bricks can be set directly on the tunnel kiln car and be burned successfully. VI. By using a kiln of suitable length, properly proportion¬ ing the raw materials, properly sizing the raw materials, and burning to a proper temperature there appears to be no reason why all refractory fire-clay brick can not be successfully burned in the tunnel kiln. Miscellaneous Details 1. Setting The brick are set on a car such as is shown in Figure B. Fig¬ ures C, D, E, F, G, H, and I indicate the manner in which the brick are set. Layers six, ten, and two are similar; layers three, seven, and eleven are alike; layer eight is like layer four and layer nine is the same as layer five. The upper layers, twelve and thirteen, conform to the shape of the roof of the kiln. The lower layers are set openly in order to secure a better draft in the portion of the car which is naturally coolest. Each layer is so placed that an interlocking effect is secured which overcomes any tendency for the load to be dislodged while pass¬ ing through the kiln. Four men were able to set seven cars, each of which held from 920 to 950 brick, in 105 minutes, or one car every fifteen minutes. 17 SCALE— 1" =8" Front View of Tunnel Kiln Car Fig. B. 18 Fire Clay Top b—Sand Shields c—Iron Framework LAYER ONE Front of Car SCALE— 1'»= 10 19 LAYER TWO 20 LAYER THREE Fig. E. 21 LAYER FOUR 22 LAYER FIVE 23 LAYER TWELVE 24 TOP LAYER THIRTEEN Fig. I. 25 In ordinary practice they work from six to seven hours a day in setting the cars for one tunnel kiln, including two extra cars per day for the Sunday supply. If it were not for the large amount of special shapes sent through with the fire-brick still better time could be made. 2. Fuel and Firing Barringer reported the use of 650 pounds of coal per 1000 fire-brick. At the time of this investigation 748 pounds were used for 1000 fire-brick. Two firemen were necessary for one kiln, one in the day time, who also oiled the cars, and one night fireman. One man hauled the ashes from two kilns. These same firemen, with the aid of other men from the plant, introduced the new cars into the kiln and drew the finished cars. Four or five men were required for this operation and spent fifteen minutes every two hours in doing it. 3. Unloading Two men unloaded twelve cars of brick in seven hours. The time required for unloading standard brick was 35, 30, 16, 18, 16, and 35 minutes per car of 920-960 brick. They experienced the same difficulty as the setters in working with a variety of ware. Were the kiln used for standard brick alone the two men could undoubtedly accomplish more. Two other men took the brick from the unloaders and placed them in the yard, on the boat, or in the car. 4. Breakage The kiln breakage for standard fire-brick was extremely small. 2, 2, 5, 4, 1, and 3 broken brick were found on six cars of standard brick, or about 0.3 per cent. When mixed cars were burned, as was generally the case, the breakage was naturally higher. It was however less than 1.0 per cent., there being about fifty broken fire-brick from every 12,000 output. 5. Upkeep The kiln investigated had been in continuous operation for four years. Each part of the kiln had been maintained at ap- 26 proximately the same temperature for the entire period. Be¬ cause of this the deterioration in a kiln due to repeated heating and cooling is avoided. The cars last about two years, by which time the tops be¬ come cracked and uneven. One man can repair two of these cars in a day. 6. Adaptability Large pieces of ware were sent through the tunnel kiln without experiencing breakage. Since such special shapes could be secured four days after being set, the kiln was used largely for the burning of such ware. In one instance a gas retort was burned successfully in this manner, although it is not the usual practice. Possible Improvements The kiln is very easily changed to suit conditions, and offers a marked contrast in this respect to the periodic kiln. Producer gas, natural gas, and oil have been successfully used as fuel. Because of the location of the fire-boxes, the fuel may be easily interchanged; i. e., natural gas may be used in the summer and producer gas in the winter. If the value of one fuel increased and another does not, advantage may be taken (in fact, it has already been done), of the changing market con¬ ditions. Gravity conveyors for handling the burned bricks can also be applied to this type of kiln. Since the brick are always found at the same place, the position of the conveyor, once being fixed, need not be changed. The tunnel kiln car itself might be moved directly to the railroad car, or into the yard of a well designed plant. Professor Harrop has found that a large amount of heat is lost by radiation. Although one may touch it at any one point the kiln is so large that about forty per cent, of the heat is lost in this way. This type of a kiln may be insulated rather easily. The crown is not so large as in the periodic kiln and hence under less pressure. Being under less pressure the crown would not be so liable to fall if the kiln were more thoroughly insulated. Fur¬ thermore, the insulation may be graded to suit conditions. In 27 the type of continuous kiln where the fire moves, each part must be insulated to withstand the maximum temperature, but in this type of a kiln such a procedure is not necessary. The tem¬ perature at each point is fixed and the insulation may be adapted to the condition found at each part of the kiln. It is expected, therefore, that a study of heat insulation will result in increased thermal efficiency of the tunnel kiln. 28 Other Types of Tunnel Kilns 1. The Owens Kiln The tunnel kiln designed by J. B. Owens is in operation at Metuchen, New Jersey. It is used for the burning of floor and wall tile, fire-brick, etc. In many respects this kiln is similar to the Didier-March kiln. It is not a muffle kiln, saggers being used for the burning of floor and wall tile. The in-coming air is drawn over the hot cooling ware and in that way is preheated, which accounts for the small amount of fuel necessary for its operation. • z: —r £: T I §i Fig. M. Outlet of Kiln, Showing Cars Loaded with Tile in Saggers Just Drawn from Kiln Ready for Unloading (Patent Pending on this Style Sagger) 29 mmmm The cars are introduced into the kiln intermittently, the intervals being gauged by the character of the ware being burned. The method employed has already been described, being essenti¬ ally the same as that used in the Didier-March kilns. The mechanism used is shown in Figure N. Fig. N. Entrance to Kiln, Showing Transmission, End Door and Loaded Car Ready to Enter (Patent Pending) The cars are lighter than the cars used in the other tunnel kilns. No special shapes are necessary, standard fire-brick being used in the top of the car. The capacity of the car is good, no space being lost as is the case with the car used in the Dressier kiln. 30 Oil is used as the source of fuel at Metuchen, New Jersey, having been installed as soon* as the price of coal advanced. The insulation of the kiln is not as thorough as that of the other kilns. Because of this there is a marked difference in the cost of construction, the Owens kiln being only one-half as expen¬ sive as some of the more complex and massive kilns. The Dressier Kiln The ware enters this type of kiln at point 0 (Figure J). It is placed on cars which are introduced continuously by means of a hydraulic ram located at position F. As this car progresses the preceding cars are pushed through at the same time. The rate at which this “train” progresses depends entirely upon the ware being fired. The ware first becomes water-smoked and then becomes heated according to its position in the kiln. As it does become heated the circulation of the hot air in the kiln takes place as follows (see Figure L) The hot air from the combustion chambers gradually rises through the passages, L, until the top of the kiln is reached. Here it strikes the cooler ware and falls, passing through the ware R, and the temporary air passage M. From here the cooler air passes through the passages N, which are in the car, over to the hot combustion chamber. In this way the air moves in a spiral path, tending to make the cooler portions warmer, and the warmer portions cooler, with the resulting small temperature variation found in this type of kiln (less than one cone). As the car passes the hot zone an entirely different condition occurs. The ware is now hotter than the cooling pipes which take the place of the combustion chambers. The air now rises up through the hot ware, strikes the cool roof of the kiln and settles again (Figure K). 31 Dressier Tunnel Kiln A—Gas Producer B—Gas Duct C—Car Track D—Combustion Chambers O—Mouth End of Kiln Fig. J. E—Exhaust Fan P—Tail End of Kiln F—Propelling Apparatus G—Gas Inlet G1—Air Inlet G2—Outlet of Hot Flue Gases H—Pipes in Cooling Zone I—Cooling Zone Pipe Out¬ lets K—Inspecting Chamber S—Insulating Material Cross Section—Hot Zone Fig. L. D—Combustion Chambers R—Ware to be Fired LMN—Air S—Insulating Material 32 Cross Section—Cooling Zone Fig. K. H—Cooling Pipes In this way the incoming air becomes heated to a high temp¬ erature and is available for use either in drying or for the com¬ bustion of the fuel. It is found that the best results are secured when the heated air is divided, part being used for each of the two possible purposes. Were the air used entirely for combustion an excess would be present and the efficiency of the kiln would be lowered. The ware finally reaches the point P (Figure J), where it is taken from the kiln at a temperature so low that glazed ware does not dunt. This kiln is generally fired with producer gas or natural gas, or with both according to the time of the year. There is no reason why oil would not be used as well as in the Owens kiln. The cars used are shown in Figures L and K. The ware, R, is placed upon these cars in two sections which are separated by the temporary air passage M. The air passages N are perm¬ anently located in the top of the car. 33 The car moves on small rolls rather than wheels, these rolls being cooled by air pipes which are located in such a way that the rolls do not rust or scale due to heat. This method of cooling the rolls gives satisfaction, as does the sand seal of the other cars. The combustion chambers D serve as muffles, the heat being taken from them by the circulating air. Because of the high temperature necessary to mature refractories, these muffles are made of carborundum, no other material being as capable of withstanding the intense heat without deterioration. From 10 to 12 per cent, of fuel is necessary for burning ware to 1250°C. (cone 6 or 2282°F.). Draft is maintained by fans and is controlled very accur¬ ately. The artificial draft makes it possible for very accurate control of the kiln atmosphere. The kiln is more expensive than the other tunnel kilns be¬ cause of its more complex structure. This feature coupled with the lower capacity of the cars are the only bad features of the kiln. They are to be met by the uniformity in temperature which is secured and the lack of flashing. The continuous motion of the ware eliminates any sudden increase in temperature of the ware, and the artificial draft, as has been said, results in good control. These latter features are applicable to any kiln. Special arrangements are made for the payment for such kilns. INDUSTRIAL FELLOW MELLON INSTITUTE OF INDUSTRIAL RESEARCH UNIVERSITY OF PITTSBURGH. 34