IL L INO I S UNIVERSITY OF ILLINOIS AT URBANA-CHAMPAIGN PRODUCTION NOTE University of Illinois at Urbana-Champaign Library Large-scale Digitization Project, 2007. UNIVERSITY OF ILLINOIS ENGINEERING EXPERIMENT STATION BULLETIN NO. 11 APRIL 1907 EFFECT OF SCALE ON THE TRANSMISSION OF HEAT THROUGH LOCOMOTIVE BOILER TUB)ES BY EDWARD C. SCHMIDT, M. E., ASSOCIATE PROFESSOR OF RAILWAY ENGINEERING, AND JOHN M. SNODGRASS, B. S., INSTRUCTOR IN RAILWAY ENGINEERING During the past twenty or thirty years there has been consid- erable discussion in railroad circles as to the effect of scale upon the heat-transmitting properties of tube surfaces, and the con- sequent effect upon the consumption of fuel. Statements as to the extent to which deposits of scale affect the conductivity of a tube or sheet have been made from time to time and have differed widely. In a committee report on boiler incrustation in the Proceed- ings of the American Railway Master Mechanics Association of 1872, we find the following quotation from a paper by Dr. Joseph G. Rodgers before the American Association for the Advance- men't of Science, given as the best information which the commit- tee had been able to obtain: "The evil effects of scale are due to the fact that it is relatively a non-conductor of heat. Its conduct- ing power compared with that of iron is as 1 to 37.5. This known, it is readily appreciated that more fuel is required to heat water through scale and iron than through iron alone. It has been demonstrated that a scale -y in. thick requires the extra expendi- ture of 15% more fuel. As the scale thickens the ratio increases. Thus when it is I in. thick, 60% more is required; ............." ILLINOTS ENGINEERING EXPERIMENT STATION The report continues as follows: "On most western roads incrusta- tions will form to a thickness of from J in. to A in. in the course of one year, and will increase at a still greater ratio as long as the engine is kept in service. Thus after four months' time, there will have accumulated in our engines nearly ii in. of scale. If Dr. Rodgers' theory be correct, after one month's service our engines will consume 31 % more fuel than at first; after two months' service 7%/ and so on, making an average for the year of over 20% more fuel than they would have consumed if using pure water." In a report before the same Society in the year 1877 upon "Feed Water" a committee under the sub-heading, "The Effect of Incrustations on the Consumption of Fuel," reports in part as fol- lows: "The increase in the consumption of fuel, on account of incrustations on the heating surfaces of boilers, varies with the thickness and density of the deposit. When porous the water will penetrate it, but when hard and compact it presents a com- plete barrier to the contact of the water with the heating surfaces. As incrustations are poor conductors of heat, an increased con- sumption of fuel is inevitable where they exist." The committee then cites a number of cases for which sufficient data were collect- ed to estimate the per cent loss that was occasioned due to scale deposits. A table showing the average miles run to one ton of coal by engines upon the Illinois Central Railroad for three months prior to and for three months after the removal of incrustations, including 120 such cases and extending over a period of three years, showed as a general average an increase of 11% in.the con- sumption of coal for three months prior to the cleaning of the boilers, as compared with the three months immediately succeed- ing. The result of 11% loss due to scale is, of course, entirely a general result, as individual cases often showed less miles run per ton of coal after cleaning than before. This difference from the general result could in most cases be accounted for by weather differences. A second case is cited for two passenger engines, which were of the same size and pattern, and which were run with the same trains on alternate days. Records were kept for the six months preceding and six months following the cleaning of the boilers. Both engines had previously been cleaned at the same time and had made an average of 34,047 miles before the test be- gan. The tests as run showed a difference in favor of clean EFFECT OF SCALE ON BOILER TUBES heating surfaces of 17.5%. Mr. Wells, the master mechanic mak- ing this test, however, concludes, on account of the scaled tubes being run more often during the winter months than the cleaned tubes, that of the 17.5% difference in consumption of fuel between clean and incrusted heating surfaces about 2% was due to temper- ature and 15i% to the effects of incrustation. Similar tests with two freight engines gave a difference of 26% in favor of clean heat- ing surfaces. A correction of 4% was applied to this on account of different atmospheric temperatures under which the tests were made, "thus giving a net saving of 22% in favor of clean boilers on two freight engines." Other cases might be mentioned giving results more or less similar, also cases in which little or no loss was found to be occa- sioned by the presence of scale. Likewise, the opinion has been not uncommonly expressed that there is either no fuel loss due to the presence of scale in the usual amounts or that loss is so small as to be of little practical importance. During the last few years there have been made by the Rail- way Engineering department of the University of Illinois four series of experiments to determine the relative conductivities for heat of clean and scale-covered locomotive boiler heating surfaces. A fifth series of tests is now being carried on along the same gen- eral lines as the others. It is the purpose of this bulletin to report upon the results of the first four series of these tests. The tests were planned with the purpose of determining, not only the actual transmission loss due to scale in individual cases, but also the relation of this loss to the scale thickness. The last three series were arranged especially to try to determine whether there is any regularity of variation of heat transmission loss with scale thickness and to study at the same time the effects of chem- ical composition on this loss. It was recognized from the outset that in any series of com- parative tests for the purpose of determining the loss in heat transmission due to the presence of scale, practically exact similarity of conditions was essential for trustworthy results. A study of previods work done along this line, such as the cases and results already referred to, also served to emphasize the ne- cessity of such care. In all of the work hereinafter reported the greatest stress has been laid upon this point, i. e. the elimination of variations in conditions except the scale itself. The difficulties ILLINOIS ENGINEERING EXPERIMENT STATION -which have been encountered while prosecuting this work have still further emphasized this necessity. The difficulties attend- ing road-testing are well understood both as to exact measure- ments and similarity of conditions. These considerations were of weight in determining that the tests herein reported should be largely of laboratory character rather than road tests. The first series* of tests was made during May and June 1898. The method employed in making the tests was as follows:- A Mogul freight locomotive, which had been in service 21 months and which was about to be sent to the shops for repairs and new tubes, was set in the roundhouse and the boiler tested by the standard method. The locomotive was then sent to the shops and the boiler carefully cleaned and retubed. All the scale was re- moved and samples analyzed from nine different parts of the boiler. It was then sent back and again tested for evaporation under the same conditions as before cleaning. Before making the trials with the clean tubes the locomotive was allowed to make one or two trips on the road so as to insure its being thoroughly clean. The tests were made in the round-house at Champaign, Illinois. The locomotive was set in the roundhouse over a pit and the tender removed. A car of coal was then run in back of the en- gine and on this were arranged the scales for weighing the coal. All of the feed water was weighed and then delivered into a tank placed on a platform by the side of the car, and connected with the suction pipe of the injector. The -slide valve on one side of the locomotive was moved back far enough, by disconnecting the valve rod, so thatlthe steam generated could pass directly into the exhaust, and thus out through the nozzle and produce the necessary draft as usual. A 2-in. pipe was also run from the dome to the atmosphere, a valve in the pipe furnishing additional means of disposing of the steam generated. The tests were started by the standard method, i. e., raising steam to the running pressure, drawing the fire and start- ing with weighed wood. At the end of the tests the ashes were all weighed. One of the regular road firemen fired for all the tests and'the boiler and furnace were operated under the usual road conditions. A series of observations was made during these tests, to determine the re- *This test constituted the thesis for graduation of Messrs. F. H. Armstrong and J. N. 'Herwig. Fie. No. 1 LOCOMOTIVE No. 420 EFFECT OF SCALE ON BOILER TUBES 5 lation between the blast-pipe pressures and vacuum in smoke box and furnace, as well as the velocity of the gases in the stack at various points along two diameters at right angles to each other. The locomotive upon which the tests were made was a Mogul freight engine made by the Rogers Locomotive Works,, and was one of nineteen in use at that time on the Chicago division of the Illinois Central Railroad between Champaign and Centralia, Illi- nois. Fig. 1 shows the arrangements just described and gives a general view of the locomotive. Leading dimensions: No. of Locomotive............. ....................... 420 Diameter of cylinder................................... 19 in. Stroke......................................... ........ 26 in. Diameter of drivers................................... 564 in. Weight on drivers..................................... 106,400 ibs. W eight on trucks..................................... 19,600 lbs. Total weight of engine ....................... ....... 126,000 lbs. Diameter of boiler .................................... 62 in. Number of tubes...................................... 236 Diameter of tubes................... .... ............ 2 in. Length of tubes............. ....................... 11 ft. 1 in. over tube sheets Length of firebox.............................. ...... 1144 in. W idth of firebox........................ ............ 33g in. Depth of firebox, front end........... ................ 674 in. Depth of firebox, back end...................... .... 59 in. Length of grate ..................................... 114k in. Width of grate ............................. ......... 33H in. Diameter of dry pipe................................. 8 in. outside Diameter of steam dome.... ......................... 29N in. inside Height of steam dome................................ 28 in. Kind of lagging........................................ Magnesia sec- Governing proportions: tional Grate area ......................... ..... ......... 26.45 sq. ft. Total heating surface................................. 1531.6 sq. ft. Area of draft through tubes.......................... 573.5 sq. in. Ratio of grate to heating surface.. ............... 57.9 Fuel used: Commercial name..................................... Odin Commercial size... .................................... M ine run Lum ps per cent ........ .............. ................ 75 Sm all coal per cent.... ..................... ......... 20 Slack per cent....................................... 5 Heat units per lb. of dry coal (by calorimeter)......... 12,240 The results of these tests are exhibited in the accompanying tables. ILLINOIS ENGINEERING EXPERIMENT STATION TABLE 1 LOG OF OBSERVATIONS GIVING AVERAGE VALUES LOCOMOTIVE No. 420 ILLINOIS CENTRAL RAILROAD First Series Secoi Scale in C Boiler I Date of Trial (1898) ....... ..................... May 2 May 3 May 31 Duration of trial, hours............................ 8.33 May 3 8.0 31 Steam pressure by gage ............................ 143 140 116 40 Vacuum in smoke box (in. of water) ...............2 2 2.9 Temperature of roundhouse (degrees F.)... ... 72 62 79 Temperature of feed water in tank (degrees F.) 57 54 58.5 Temperature of escaping gases (degrees F.)..... 623 670 621 Temperature of steam (degrees F.)......... ....362 360 348. Moisture in coal, per cent ............. .. .. .. 4.0 4.0 Percentage of ash (from ash pan)........... .... 15.6 15.6 16.6 Percentage of moisture in steam................ 2.25 2.25 2.85 nd Series leaned Boiler June 1 8.16 114 2.8 89 59.4 687 346 4.0 18.7 2.85 TABLE 2 RESULTS OF EVAPORATION TEST OF LOCOMOTIVE BOILER ENGINE NO. 420, ILLINOIS CENTRAL RAILROAD First Series: After running 21 months and accumulating a scale deposit 3 to A inch thick. Second Series: After cleaning and putting in new tubes. First Series Second Series Scale in Clean Boiler Boiler Date of Trial (1898)............... May 2 May 3 May 31 June 1 Mean Mean lbs. lbs. lbs. lbs. lbs. lbs. . Water actually evaporated per lb. of dry coal ........ ........ .. 5.21 5.27 5.24 5.81 5.85 5.83 F. per lb. of dry coal........... 6.29 6.39 6.34 6.99 7.01 7.01 S Water actually evaporated per lb. of combustible......... 6......6.17 625 6.21 6.95 7.16 7.05 Equivalent water from and at 21S2 F. per lb. of combustible...... 7.46 7.59 7.53 8.36 8.61 8.48 0 6 Q Dry coal burned per hour' per sq. 010 p ft. of grate surface.............. 57.45 58.51 57.95 59.80 60.00 59.90 Per sq. ft. of tubeopening ......... 394.80 402.10 398.40 411.00 412.80 411.90 g Per sq. ft. of water heating surface .93 .95 .94 .97 .98 .97 ,M'. p Water evaporated per hour from and at 212" F. per sq. ft. of grate surface ...................... 361.80 374.40 368.10 418.00 416.00 417.00 Persq. ft. of tube opening .......... 486.00 2573.00 2529.00 2874.00 2857.00 2865.00 S Per sq. ft. of waterheating surface 5.89 6.09 5.99 6.81 6.76 6.79 EFFECT OF. SCALE ON BOILER TUBES 7 The loss due to scale in this boiler was (7.01 minus 6.34) divided by 7.01 or 9.55 %. The water used in the locomotive tested was taken from tanks at Centralia, Kinmundy, Little Effingham, Neoga, Dorans, Galton and Champaign. From the thickness of scale deposited during the 21 months it is evident that these waters are comparatively good for this section of the country. The average thickness of the scale on the principal heating surfaces was -r in. The total weight of scale removed on clean- ing was 485 lbs. The boiler'had been in regular service during the 21 months. The locomotive was cleaned and retubed at the Burnside shops of the Illinois Central Railroad. When the boiler was opened all the scale removed was carefully weighed, the scale on the tubes being determined by weighing the tubes before and after cleaning them. The scale from the shell and firebox sheets that could be removed was carefully collected. The total weikht of scale was as follows:- W eight of scale from flues........................... 360 lbs. W eight of scale from shell................. ......... 125 lbs. Total weight of scale ............... ................ 485 lbs. At nine different points in the boiler the thickness of the scale was determined by the average of many measurements, and samples were secured for analysis as follows:- Point 1. Near injector discharge, hard and soft scale J in. thick. 2. On upper tubes, hard smooth scale uniform thickness A in. 3. On lower tubes, hard scale near middle, a in. thick. 4. Mud covering hard scale at No. 3, A in. thick. 5. Scale from side sheet, flue sheet, and tubes rough and scaly. 6. From bottom of barrel, 4 ft. from flue sheet. 7. On crown stays, 3 in. to 6 in. from crown sheet. 8. On crown sheet, rivet heads and base of stays. 9. From stay bolts at water line. The results of the analyses of these scales calculated to com- pounds are shown in Table 3. ILLINOIS ENGINEERING EXPERIMENT STATION TABLh 3 RESULTS OF THE ANALYSES OF BOILER SCALE FROM ENGINE NO. 420 SCALE CONSTITUENTS CALCULATED TO COMPOUNDS AND EXPRESSED IN PER CENT 3 8.00 4 7.84 5 15.89 6 11.25 7 18.25 8 13.05 9 *2.70 100 B a, '3 .5) 0. 00 ~CJ2 0 '5 0 a '5 0 0 .0 aO 00 '5 0 '5 0 5) 00 C 0 5) 5) a .3 F' 0 a 0 5) n bS a The loss, as found by these trials, due to the presence of scale, was 9.55 % of the fuel. EXPERIMENTS WITH SINGLE TUBES The last three series of experim4nts to determine the loss due to scale have been laboratory experiments entirely. They were made during the years 1901, 1904 and 1905, and are referred to as the series of 1901, 1904'and 1905 respectively.* The locomotive boiler tubes upon which the experiments were made in 1901 were furnished by the Peoria and Eastern division of the Cleveland, Cincinnati, Chicago and St. Louis, the Illinois Central, the Chicago, Burlington and Quincy, and the Chicago, Milwaukee and St. Paul Railways. The tubes used in 1904 and 1905 were furnished by the first two railroad companies mentioned above. Table No. 4 gives information concerning these tubes. Fig. 2 shows some of the tubes tested. *These experiments were conducted by the following men as theses for graduation: Series of 1901, by F. L. McCune; Series of 1904, by W. A. Miskimen and C. N. Stone; Series of 1905, by H. F. Godeke and A. A. Hale. 1 7.70 r i Fl. iNo. 2 SCALED BOILER TUBES EFFECT OF SCALE ON BOILER TUBES TABLE 4 THE TRANSMISSION OF HEAT THROUGH SCALE-COVERED BOILER TUBES RAILWAY ENGINEERING DEPARTMENT- UNIVERSITY OF ILLINOIS 8 V REMARKS • '0 *3  General Character of Scale, Etc. 1 2 3 4 5 6 7 SERIES OF 1901 1 I. C. R. R. 311 10.5 2 0.06 Even, hard, dense 2 P. & E. RY. 526 13.5 2 0.04 Soft. porous. Removed in places 3 P. & E. RY. 536 5.5 2 0.02 Hard, dense, white 4 C. M. & ST, P. 126 .... 2 0.03 Hard, dense, white 5 C. M. & ST. P. 1337 .... 2 0.13 Hard, dense 6 I. C. R. R. 820 5.5 2 0.07 Mileage during service, 19690 7 P. & E. RY. 513 37.5 2 0.04 Hard, dense, rough, one end. Soft, porous at the other 9 C. B. & Q. 1179 .... 2 0.11 Hard, porous, gray. Mileage, 50889 11 I. C. R. R. 110 21. 2 0.09 Soft, porous 14 P. & E. RY. .... .... 2 .... New and clean tube SERIES OF 1904 1 I. C. R. R. 41 16 2 0.04 Hard, gray. In bad condition 2 I. C. R. R. 41 16 2 0.07 Loose, gray 3 I. C. R. R. 41 16 2 0.08 Loose, gray 4 I. C. R. R. 141 15 2 0.05 White, porous. Removed in places 5 I. C. R. R. 141 15 2 0.04 White, porous. Removed in places 6 I. C. R. R. 141 15 2 0.08 White, porous 7a G.C.C. & ST. L. 540 .. 2 0.06 White, soft, irregular 7b C.C.C. & ST. L. 540 .. 2 0.06 White, soft, irregular 8a C.C.C. & ST. L. 540 .. 2 0.05 Hard, white, irregular 8b C.C.C. & ST. L. 540 .. 2 0.04 Hard, white 9 I. C. R.R. .. 2 0.06 Hard 10 I. C. R. R. 440 .. 2 0.03 Hard, gray 11 I. C. R. R. 440 .. 2 0.09 Gray, porous 12 I. C. R. R. 440 .. 2 0.03 Gray, porous 13 I. C. R. R. . ... .. 2 .... Clean tube SERIES OF 1905 3 I. C. R. R. 136 18 2 0.07 Medium 4 I. C.R. R. 802 8 2 0-05 Hard 8 C.C.C. & ST.L. 533 10 2 0.03 Soft 9 C.C.C. & ST. L. 233 14 2 0-09 Very soft 10 I. C. R. R. 1424 10 2 0.07 Soft 11 C.C.C.&ST.L, 233 14 2 0.04 Very soft 12 I. C, R. R. 140 21 2 0.07. Hard 13 1, C, R R. 303 18 2 0.02 Hard 14 I. C. R. R. 1004 21 2 0.04 Medium 15 I. C. R. R. 1012 12 2 0.03 Very hard 7 ................ .. .. 2 .... Clean tube 10 ILLINOIS ENGINEERING EXPERIMENT STATION These tests were made as laboratory tests on account of the desire to make comparative tests under entirely similar conditions except in regard to the scale itself and in order that more exact measurements might be made than were found possible with road or roundhouse tests. The apparatus used in all of these tests has been practically the same from year to year. It is shown in Fig. 3, 4, and 5 and consists of a long water chamber through which the tube to be tested was passed, and in which water was circulated. On one end of this water chamber was fastened a combustion chamber, at the forward end of which was placed a burner. This burner was supplied with gas and air. Combustion took place in the chamber, which served the purpose of the firebox. The hot gases passed through the boiler tube to the air. The water entered the water chamber at the right, leav- ing it at the left as indicated in Fig. 3, at both of which points its temperature was read upon the thermometers there shown. The water tank received the water from the city mains. It was pro- vided with an overflow, as shown, and the water was led directly down from the bottom of this tank to the water chamber. This was used in order to give a constant pressure at the inlet and thus to avoid variations in the rate of flow of the water. The gas and air tanks were arranged to give constant pressures of gas and air. The inner vessels, open at the bottom, float in water contained in the outer tank and confine the air or gas in the space above the water level. These inner tanks can be weighted at will to give any desired pressure to the air or gas contained within them. During the tests of 1901 a copper ball pyrometer was employed to obtain the temperatures of the gases entering the tube being tested. For the series of 1904 and 1905 a Le Chatelier pyro- meter was employed for this purpose. The location of the pyro- meter is shown in Fig. 3. The temperature of the gases as they left the flue was read on the thermometer shown at the end of the tube. The purpose of the tests was to measure the number of heat units transmitted per hour through the different tubes. This was accomplished by weighing the water which circulated around the tube in the water chamber and measuring its rise in temperature. The attempt was made to maintain a constant furnace temperature at the entrance to the tubes throughout all experiments of each EFFECT OF SCALE ON BOILER TUBES 12 ILLINOIS ENGINEERING EXPERIMENT STATION series and thereby have available for transmission the same amount of heat, kince the amounts of gas and air supplied to the burner were continually the same. TABLE 5 THE TRANSMISSION OF HEAT THROUGH SCALE COVERED BOILER TUBES RAILWAY ENGINEERING DEPARTMENT-UNIVERSITY OF ILLINOIS SERIES OF 1901 5 13 14 Average 1595 266 1690 267 1693 268 1659 267 931 979 981 964 64.0 67.8 68.5 66.8 109.4 114.0 114.2 29830 29463 25746 26981 27123 27212 29812 30609 26060 28330 30916 30427 28274 30194 24351 22776 27167 26809 30063 29021 25999 27331 26987 27162 30304 31522 27923 27606 30494 29101 27961 29802 26459 23794 27167 26321 86.7 45.4 644.0 844.3 29238 Clean 90.9 46.2 650.2 888.1 30039 Tubes 91.4 45.7 664 0 889.6 30345 *Increase i 112.5 89.7 45.8 652.7 874.0 29874 14 14 1i 1 1 1 Sz 0 !z 0 0^ 50 0 20 0) 6 0 EFFECT OF SCALE ON BOILER TUBES SERIES OF 1904 13a 13b 13c 6 5 4 1432 648 1439 659 1436 667 1418 515 1438 523 1439 511 1439 647 1437 523 1443 623 1414 593 1423 547 1431 517 1415 536 1426 551 1441 554 1439 563 1440 569 998 1001 1005 56.5 56.7 56.3 82.5 85.1 82.6 69.5 70.9 69.5 27.9 829 26.1 850 25.2 895 22.7 975 26.3 814 24.4 865 24.9 875 25.0 803 32.4 662 32.4 654 26.2 767 25.7 808 26.3 735 23.9 896 26.0 891. 28.4 825. 26.3 900. 928.5 930.1 935.5 23166 23430 23670 22227 21128 21405 22958 21895 21734 20845 20525 20812 21247 20471 21442 19916 21727 Clean Tubes 5.1 9.8 8,6 2.0 6.5 7.2 11.0 12.4 11.1 9.3 12.6 8.5 15.0 7.2 Average 1440 563 1001 56.5 83.4 70.0 26.9 872. 931.4 23422 SERIES OF 1905 3 48 1 1783 873 1328 62.4 103.6 83.0 41.2 717 1245.0 29540 28722 2.8 3 49 1 1781 879 1330 62.4 102.1 82.3 39.7 753 1247.7 29894 29002 1.9 3 50 1 1798 867 1333 62.0 99.2 80.6 37.2 800 1252.4 29760 28764 2.7 4 45 1 1816 733 1275 62.2 102.6 82.4 40.4 699 1192.6 28240 28663 3.0 4 46 1 1809 715 1262 63.2 102,0 82.6 38.8 707 1179.4 27432 28155 4.8 4 47 1 1806 705 1256 63.3 102.4 82.9 39.1 692 1173.1 27057 27920 5.6 8 24 1 1788 740 1264 59.7 102.7 81.7 43.0 641 1182.3 27563 28220 4.5 8 25 1 1790 743 1267 60.1 102.4 81,3 42.3 660 1185.7 27918 28502 3.6 8 26 1 1805 743 1274 60.1 101.7 80.9 41.6 656 1193.1 27290 27688 6.3 9 27 1 1752 791 1272 59.9 101.1 80.5 41.2 683 1191.5 28140 28588 3.3 9 28 1 1753 800 1277 60,0 100.6 80.3 40.6 682 1196.7 27690 28009 5.2 9 29 1 1763 793 1278 59.7 100.6 80.2 40.9 695 1197,8 28426 28727 2.8 10 30 1 1772 786 1279 59.0 101.2 80.1 42.2 674 1198.9 28443 28718 2.9 10 31 1 1782 788 1285 58.8 101.8 80.3 43.0 663 1204.7 28509 28646 3.1 10 32 1 1783 792 1288 59.1 101.4 80.3 42.3 671 1207.7 28383 28449 3.8 11 34 1 1776 805 1291 67.1 103.1 85.1 36.0 767 1205.9 27612 27717 6.2 11 35 1 1785 803 1294 67.0 102.4 84.7 35.4 753 1209.7 26656 26674 9.8 12 37 1 1776 804 1290 64.5 102.1 83.3 37.6 747 1206.7 28087 28176 4.7 12 38 1 1785 790 1288 64.0 102.1 83.1 38.1 721 1204.9 27470 27598 6.6 13 39 1 1766 740 1253 60.2 101.4 80.8 41.2 679 1172.2 27975 28889 2.3 13 40 1 1754 748 1251 60.7 102.9 81.8 42.2 668 1169.2 28190 29190 1.3 13 41 1 1747 736 1242 61.2 101.5 81.4 40.3 695 1160.6 28009 29213 1,2 14 43 1 1805 808 1307 58.5 102.1 80.3 43.6 657 1226.7 28645 28267 4.4 14 44 1 1792 803 1298 58.8 102.9 80.9 44.1 645 1217.1 28445 28290 4.3 15 51 1 1804 759 1282 61.1 102.1 81.6 41.0 688 1200.4 28208 28445 3.8 15 52 1 1808 736 1272 60.8 100.9 80.9 40.1 701 1191.1 28110 28568 3.4 7 21 1 1792 794 1293 64.5 102.1 83.3 37.6 782 1209.7 29403 Clean 7 22 1 1796 787 1292 64.2 103.6 83.9 39.4 754 1208.1 29708 Tubes 7 23 1 1808 784 1296 63.3 101.4 82.4 38.1 776 1213.6 29566 ) Average 1799 788 1294 64.0 102.4 83.2 38.4 770.7 1210.5 29559 The method of conducting a test was as follows: The burner was first lighted and the gas and air pressures ad- justed, then the flow of water through the water chamber was reg- ulated and the apparatus allowed to run until all conditions had become uniform. This usually occupied about one hour, at the end of which time the test was started. 1 1 14 ILLINOIS ENGINEERING EXPERIMENT STATION At the beginning of a test for the series of 1901 a determination of temperature was made by the copper ball pyrometer and read- ings were taken on all three thermometers, which readings were also taken at intervals of five minutes throughout the test. At the end another determination was made of the furnace tempera- ture, and the water which had flowed through the chamber was weighed. Observations for the tests of 1904 and 1905 were taken in a similar manner except that all temperature readings including that of the furnace were taken at regular intervals of 10 minutes. Table No. 5 gives a summary of the data of the various tests and also the calculated results. The furnace temperature was not maintained quite the same throughout the tests, as an inspec- tion of Table No. 5 will show. It was likewise impossible to maintain the average temperature of the circulating water the same during all the tests. Consequently the range of tempera- ture between the gases in the tube and the water varied some- what. Since the rate of transmission of heat through the tube varies directly with this range in temperature it is necessary, in order to compare the conductivity of the different tubes, to reduce the actual amounts of heat transmitted to what they would have been for one standard range of temperature. This standard range was assumed the same as the range existing during the test of a new clean tube, such a tube being tested with each se- ries of tests. These derived figures are given in column 14, Table No. 5, and they show the amounts of heat which would have been trans- mitted in each case had the difference between the temperature of the gases and the temperature of the water been the same in all tests of that particular series, i. e., the same as during the tests of the clean tube then tested. It is from the figures in this column that the losses due to scale are computed. This loss ex- pressed as a per cent is exhibited in column 15 of Table No. 5. Table No. 6 gives the chemical analyses of the scale found upon the tubes tested in the series of 1901, 1904 and 1905. The constituents of the scale are calculated to compounds and ex- pressed as per cent. These analyses were made by the Chemical department of the University of Illinois. EFFECT OF SCALE ON BOILER TUBES TABLE 6 THE TRANSMISSION OF HEAT THROUGH SCALE-COVERED BOILER TUBES RAILWAY ENGINEERING DEPARTMENT-UNIVERSITY OF ILLINOIS CHEMICAL ANALYSES OF SCALE 2 c/I '5 92 1 2 Constituents of 66 O0: 0 0) '5 '5 0 Scale Amount in per cent SERIES OF 1901 1 7.00 8.32 16.78 47.16 .... 1.30 11.20 1.18 7.06 2 5.68 8.98 27.30 22.27 10.45 .... 15.82 1.14 8.34 3 9.24 10.98 2.04 59.75 0.62 .... 10.18 0.64 6.55 4 9.80 15.92 25.86 23.20 3.02 .... 13.16 1.42 7.62 5 10.00 7.00 14.35 50.30 .... 8.40 7.30 0.84 1.81 6 7.42 4.26 16.08 51.44 .... 1.53 11.25 1.19 6.83 7 12.22 5.38 17.70 37.02 .... 0.79 16.76 1.63 8.50 9 12.46 11.24 36.85 21.37 .... 3.13 0.72 1.02 13.21 11 17.82 10.32 5.06 37.50 6.70 .... 13.92 2.25 6.43 SERIES OF 1904 1 26.04 10.80 1.36 27.50 3.84 ..... 23,07 7.39 2 17.21 8.02 18.50 11 52 .... 15.61 21.74 7.40 3 15.10 3.95 21.93 10.48 .... 11.06 25.28 12.20 4 8.99 1.90 22 81 49.11 .... 14.11 0.50 2.58 5 9.11 2.20 47.96 16.85 .... 2.46 11.76 9.66 6 12.10 2.60 54.99 10.78 .... 4.63 9.57 5.33 7a 12.93 1.35 21.83 31.14 ..... 5.14 15.94 11.67 7b 12.93 1.35 21.83 31.14 .... 5.14 15.94 11.67 8a 12.70 2.83 35.05 17 03 .... 3.99 17.85 11.05 8b 12.70 2,33 35.05 17.03 .... 3.99 17.85 11.05 9 23.52 7.20 6.56 37,56 3.48 . .... 5.59 16.09 10 10.05 6,47 11.71 50.98 .... 7,33 4.94 8.52 11 9.75 2.08 6.05 54.71 .... 5.72 6.43 15.26 12 7.98 4.68 8.89 55.61 .... 11.36 3.75 7.73 SERIES OF 1905 3 7.09 5.05 17.16 19.45 0.77 34.10 0.58 15.80 4 6.92 3.57 21.57 3.61 25.61 .... 1.85 0.56 36.31 8 6.61 1.34 0.62 74.26 ..... 0.15 10.87 0.68 5.47 9 8.44 2.52 12.59 56.80 ..... 0.33 11.43 0.87 7.02 10 3.33 1.43 5.82 67.99 1.64 .... 11.11 0.48 8.20 11 7.09 2.30 14.87 58.18 ..... 2.22 8.26 0.75 6.33 12 27.72 9.53 12.11 10.05 0.24 .... 29.90 1.07 9.38 13 9,54 10.38 8.41 9.67 3.87 .... 39.09 0-71 18.33 14 16.87 4.73 2.22 40.30 6.06 .... 19.25 0.83 9.74 15 24.03 12.69 1.46 31.37 0.35 .... 20.91 1.45 7.74 16 ILLINOIS ENGINEERING EXPERIMENT STATION Table No. 7 gives a summary of the data and results. In it are given for each tube the corresponding average loss as deter- mined by the several experiments, as well as the thickness and some of the results of analyses. From this table- there have been plotted five diagrams-Fig. 6, 7, 8, 9 and 10, which exhibit the loss due to the scale with reference to thickness, hardness and chemical composition. Fig. 6 shows the loss due to scale plotted with reference to its thickness. Fig. 7 is identical with Fig. 6, except that the letters H, S or M have been added at the various points to indicate the scale as being either hard, soft, or medium. In Fig. 8, 9 and 10 the loss due to the scale is plotted with refer- ence to the amount of its chemical constituents; in Fig. 8 with reference to the sum of the percentages of calcium carbonate and magnesium carbonate; in Fig. 9 with reference to the percentage of calcium sulphate; and in Fig. 10 with reference to the percent- age of silica. In the series of 1901 there are a few tests which indicate an increase of conductivity of the scaled tube as compared with the clean tube. These are perhaps to be accounted for by errors in conducting the experiments, although they could not be detected at the time the experiments were made. The apparatus used in 1904 and 1905 was improved in some particulars, the most import- ant change being in the means for the measurement of furnace temperatures. Such discrepancies disappear in the latter series. When the experiments were planned it was considered prob- able that the transmission of heat through the scale was princi- pally dependent upon two of its characteristics, namely, its thick- ness and its mechanical structure and that probably, for such thick- nesses as are usually met with, thickness had greater influence than structure. Thickness was therefore carefully determined and structure approximately designated as in Table 4, as hard, soft or medium, no more exact characterization of structure being possible with tubes collected from different sources as these were. It was hoped that the experiments might develop, if perhaps only approximately, some law of variation of conductivity with thickness. After making allowance for probable errors due to the method of conducting the tests, consideration of Fig. 6 shows perhaps a decrease of conductivity with thickness; but certainly no regularity of variation. In Fig. 7 the loss in heat transmis- sion is again plotted with reference to thickness; and the struc- ture of the scale, in so far as it was determined, is indicated as pre- viously explained. No regularity of variation is observable with respect to hardness or softness. EFFECT OF SCALE ON BOILER TUBES TABLE 7 SUMMARY THE TRANSMISSION OF HEAT THROUGH SCALE-COVERED BOILER TUBES RAILWAY ENGINEERING DEPARTMENT-UNIVERSITY OF ILLINOIS SERIES OF 1901 1.2 H 0.06 10.8 S 0.06 9.4 H 0.02 *3.5 H 0,03 7.1 H 0.J3 0.3 / 0.07 3.3 M 0.04 15.9 H 0.11 10.5 S 0.09 *Increase SERIES OF ] 5,1 .... H 0.04 9.8 .... S 0.07 8.6 .... S 0.08 2.0 .... S 0.05 6.5 .... S 0.04 7.2 .... S 0.08 11.0 .... $ 0.06 12.4 .... S 0.06 11.1 .... H 0.05 9.3 .... H 0.04 12.6 ... H 0.06 8.5 .... H 0.03 15.0 .... S 0 09 7.2 .... S 0,03 c3 04 I '4 0 u a' C3 + 4) '4 6' '4 0 z E4 z 4) 4)1 I 0 4) '4 0 4) 0 4 0 fr 5 '4 0 0) '4 .0 SERIES OF 45 0 C5 a 1905 20.21 3.61 74.41 57.11 67.9c 11 12 16 18 1 2 20 21 6 22 3 23 8 9 4 15 24 24 1 2 2 3 3 4 4 5 5 6 6 7 7 9 11 11 4) 17.16 21.57 0 a' 17.16 21.57 0.62 12.59 5.82 2.5 4.5 4.8 3.8 M H S S 48.46 16.78 22.27 27.30 59.75 2.04 23.20 25.86 58.70 14.35 52.97 16.08 37.81 17.70 24.50 36.85 37.50 5.06 904 0.07 0.05 0.03 0.09 0.07 1 15 2 16 3 17 4 18 5 19 6 20 7a 13 7b 14 8a 12 8b 11 9 10 10 7 11 8 12 9 18 ILLINOIS ENGINEERING EXPERIMENT STATION hi w I F 0 Z z 0 (flJ fiJi 3z I--1 'Ii' 0: 'IP 0 Ja u 0 tld IL 0 V) In E r- F.- - U -4 ~- s- - 02 "N l NI w o *J-N30 b3< NJ &sO-1 4 4- 4 -4 -4 -f -0- -1 -C c C C me -e I L m EFFECT OF SCALE ON BOILER TUBES z m J 0 Id) I I-l -J -J IL 2 0 z < z I- x 1- U) Ii z In 4. U) I- ILN3 '.3d N I SO-1 U) 3. 4 I .4 In I In 4 U) -4: U) In .4 I 0 2 I I e 0 U, I 3. I -c I 0 In '3 r .J* 20 ILLINOIS ENGINEERING EXPERIMENT STATION 0 L I e LI) u u - 3 0 Z r -U4 05 O zfl 0 - V U - I- J JF z oIi -u J 4- 40 * j V-m 4 4- I;- -9 i 0 -I, 4 I U (0 z 0 CD ff U N.LN0 2d NI .0 0 0 4 0 a I- 7 FZ 0 0 t+ S n 0 10 CM l m N I W eso-1 EFFECT OF SCALE ON BOILER TUBES hi r F *1- 0 z 0 (fim U< 0^ 'Ii a fto 0 -I w .j 0 lal I- z 1 ki z U I- D 8t Ij 0 h4 !c I - - I 6 4 U i- I ".ILN30 3d NI ~0-1 a I~. -D 4 r- I 4 Q. I-. w o 10 t 4. 4 4' 0 '9 1U f0 4. i - < u. 0 I-. z bi I a. - 2. .- 22 ILLINOIS ENGINEERING EXPERIMENT STATION Li w Sz i L 0o OJ Z) F z lNe £ *JLN30 h- F z 0 3 w. Ii Ii- 0 1. 4- 416 S S * - -A - -< 83d NI SO-l 0 D 9 .9 0 I Il u1 -6 0 .4 0 ot CD 0 It E4 * U IL a F* . o EFFECT OF SCALE ON BOILER TUBES In considering Fig. 6 and 7 it must be borne in mind that the tubes tested were taken from locomotives which had been in ser- vice in different parts of the country and that the scale on each tube was made up of the mineral constituents of many different water supplies. What is designated as hard scale in one case may be very different in structure-in porosity, for example-from what is designated as hard scale on another tube. Fig. 7 cannot therefore be considered as providing conclusive evidence concern- ing variation of conductivity with structure. The results may properly be interpreted as indicating that mechanical structure is at least as important a factor in the change in heat transmission due to scale as is the mere thickness. Such a conclusion is, of course, in accord with the facts concerning other heat insulators. Fig. 8, 9 and 10, in which the loss in heat transmission is plotted with reference to the principal chemical constituents of the scale, do not warrant the conclusion that its chemical composi- tion has any direct influence on its conductivity. From the point of view of the physicist the experiments are open to objection as to method. From the engineer's viewpoint it is believed that the possible errors of the experiments do not, by any means, account for all the irregularity in the plotted results, and considering the controversy upon this subject and the com- paratively meager information available, it is deemed proper to publish at this time the results as they stand in the hope that they contribute additional information which may be of interest in some quarters. Conclusions: In so far as generalization is warranted we may sum up the results of the tests in the following conclusions: 1. Considering scale of ordinary thickness, say of thicknesses varying up to i inch, the loss in heat transmission due to scale may vary in individual cases from insignificant amounts to as much as 10 or 12 per cent. 2. The loss increases somewhat with the thickness of the scale. 3. The mechanical structure of the scale is of as much or more importance than the thickness in producing this loss. 4. Chemical composition, except in so far as it affects the struc- ture of the scale, has no direct influence on its heat transmitting qualities. 24 ILL[NOIS ENGINEERING EXPERIMENT STATION PUBLICATIONS OF THE ENGINEERING EXPERIMENT STATION Bulletin No. 1. Tests of Reinforced Concrete Beams, by Arthur N. Talbot. 1904. Circular No. 1. High-Speed Tool Steels, by L. P. Brecken- ridge. 1905. Bulletin No. 2. Tests of High-Speed Tool Steels on Cast Iron, by L. P. Breckenridge and Henry B. Dirks. 1905. Circular No. 2. Drainage of Earth Roads, by Ira 0. Baker. 1906. Bulletin No. 8. The Engineering Experiment Station of the University of Illinois, by L. P. Breckenridge. 1906. Bulletin No. 4. Tests of Reinforced Concrete Beams, Series of 1905, by Arthur N. Talbot. 1906. Bulletin No. 5. Resistance of Tubes to Collapse, by Albert P. Carman. 1906. Bulletin No. 6. Holding Power of Railroad Spikes, by Roy I. Webber. 1906. Bulletin No. 7. Fuel Tests with Illinois Coals, by L. P. Breckenridge, S. W. Parr and Henry B. Dirks. 1906. Bulletin No. 8. Tests of Concrete: I. Shear; II. Bond, by Arthur N. Talbot. 1906. Bulletin No. 9. An Extension of the Dewey Decimal System of Classification Applied to the Engineering Industries, by L. P. Breckenridge and G. A. Goodenough. 1906. Bulletin No. 10. Tests of Plain and Reinforced Concrete Columns, Series of 1906, by Arthur N. Talbot. 1907. Bulletin No. 11. The Effect of Scale on the Transmission of Heat through Locomotive Boiler Tubes, by Edward C. Schmidt and John M. Snodgrass. 1907. Bulletin No. 12. Tests ot Reinforced Concrete T-Beams, Series of 1906, by Arthur N. Talbot. 1907. Bulletin No. 18. An Extension of the Dewey Decimal Sys- tem of Classification Applied to Architecture and Building, by N. Clifford Ricker. 1907.