Il II | | SS [Mili i an Thesis for Degree of C. E, 1970 FIRE BRICK AS MANUFACTURED AT CERRO DE PASCO, PERU AND SOME PAPERIMENTS WITH SILICA BRICK A, C. DODGE sHEaLS PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE VAXE DY 6/07 p:/CIRC/DateDue.indd-p.1 Ou FIRE BRICK AS MANUFACTURED AT CERRO DE PASCO, PERU and SOME EXPERIMENTS WITH SILICA BRICK. C; (- CS , a ee | 5 a May, 1910. A. Cy Dodge. New York. lo, — OT a, i i THESIS The Cerro de Pasco Mining Compeny is a New York corporation Owning or controlling nearly a11 of the mining claims losated at Cerro de Pasco, Peru, South samerica. The mineral mined at this point is a fairly hieh grade copper and silver ore being very rich in places and differing greatly in its physicsl properties. “he high grade ore is of sufficient richuuess to warrant ship= ping it to the United States or Sneland for treatment but the ereat bulk of the ore is not rich encugh to ship so far at a profit. It became necessary, therefore, to treat this ore at the mines, shipping Only the metals themselves and thus effecting & great saving in trais- portation charges. “or this purnose 2 snelter was erected about six miles from Cerro at a favorable point on the Cerro de Pasco 2.R. It was to supply this smelter with fire-brick at as low a cost as possible that the fire brick plant was constructed at vYerro. The Cerro de Pasco Mining Comnany also owns coal mines at three points in this district, the nearest being located at Vinchus- cancha, seven miles distant; the next at Goyllarisquisga, twenty-two miles distant and the other at Cuishuarcancha, twenty-five miles distant. ‘t each of these mines fire clay forms either the hanging or foot walls and ot Goyllarisecuisea and “uishuarcancha both walls. This is a perticularly fcrtunate circumstance as it enables the clay te be mined with the coal and at a comparatively small additional cost. The fire brick plent at Cerro is between these mines and the smelter and although it woulc probably have been an advantare as far ag transportation charges are concerned to locate the plant at one of (1) 95494a Cs Ze the coal mines where the raw material end fuel is obtained, the fact thet it ts difficult to secure encugh labcr at the co2l mines to keep them operating at their full capacity and that the topography of the country ig not suitable for such a plant probebly led to the select- lon of Cerro. S3esides, at the time the fire brick plant was construct- ed the reilroed was not completed ta the coal mines anc it would ac- cordingly have been necessary to delay the construction of the plant until the railroad was finished, so that the heavy machinery could be transported had the plant been located at any one of these mines. “his was not to be considered as it was very essentiel that brick be furn- ished as scon as possible for the construction of the smelter. The fire brick plant erected at Cerro cunsisted of one large building divided into three parts. Une part contained the boiler and coal bin, the central and larger part contained the engine and machin- ery while the third part was used as a drving room. (See plate I.) the boiler was of the locomotive portable type and rated at about seventyfive horse power. Thies furnished steam to the engine and tc the colle in the dryire room. The engine was a simple non-condensing slide valve horizon- tal engine with throttling governor, rated at about thirty-five horse power. The brick-making machinery proper consisted of a disintegra- tor, wet pan, hand moulds and three hand presses. This equipment is hardly sufficient, as withcut a dry pan it was almost impossible to erind the clay as fine as is necessary for the best results without greatly decreasing the capecity of the plant. Plate I gives an idea Se of the srrangaznent of this machinery in the main part of the building. In addition to the above, & supply of shovels, picks, sledges, wheelbarrows, &c. were required. The drying room was about one hundred feet in length and forty feet in width with four racks running its en- tire iength except for a trensverse isle in the centre of the building. These recks consisted of four by four posts from the floor to the ceil- ing, spaced about two feet apart with cleats one inch by three inches by two feet in length nailed on each post about eight inches apart from the floor up to about five feet. The floor of the drying room consisted of one inch boards laid on stringers with two inch planks on top of the boards. Every two feet & epace of about four inches was left vetween the two inch planks in which a two inch steam pipe was laid. Across each end was a four inch pipe with four by four by tvo inch tees every two feet to which the two inch pipes running lengthwise were connected. At the lowest point in the farthest corner a steam trap was connected to take care of the condensation in this floor coil. (See Plate I.) During the day tre floor coil was furnished with stean from the exhaust of the engine and the four inch gate valve near the stean trep left open. £t night this gate valve was closed and live steam from the boiler at about fifteen pounds pressure was turned on. the two kilns were of the down draught type with a sixty foot stack between. They were twenty-six feet in inside diameter and seven feet six inches in height to the spring line of the roof. The walls were twenty-six inches thick, lined on the inside with “nglish fire brick. These kilns had six fire boxes ang two doors. The main flue extended entirely across each kiln under the floor snd 4 transverse 4. flue extended at right angles to the main flue in 4 ling with the doors dividing the kilns into four segments under the floors. The floors were supported by fire brick arches over the flues ard rested or lateral flues six inches in width connecting with the main flues. (See Plate II.) In the manufacture of fire brick it is necessary to use some kind of binder to hold the fire clay together in the moulds especially where the grinding is not fine as was the case at Cerro. Yor this pur- pose a blue clay found about a mile from Cerro was used. This clay being off the line of the railroad, it was necessary to cart it to the plant, one dump cart with a boy driver being used for this purpose all the time the plant was running. TO determine whether or not this binder olay was chemically suitable as well as to determine which of the coal mines furnished the beet fire clay, analyses of these clays were made. A good fire clay should be high in silica anc alumina and low in fluxing materials sush sas iron oxide, lime and mariesia. The sane applies to the binder, though in a less degree as not so much of it is used in making the bricks. In analyzing these olays the writer proceeded as follows: The sample was first dried and pulverized on the bucking board at the aseay office. Cne gram of the clay was then weighed out and three grams of sodium oarbonste Nao Oz were added for flux. This mixture was placed in a platinum crucible anc heated at a high temperature in a muffle for thirty minutes, or until fusioa was complete. The Crucible was then cooled, placed in a number two beaker and covered with hot water. This sclution was then acidulated with a few drops of H. 01 and allowed to digest for ten or twelve hours. usually over night. In the morning the crucible was taken out, washed well and the solution poured into a casserole and evaporated to dryness. This renders the alumina Al, 03 sOluble. A few drope of Ho S CO, should be added to the solution while evaporating to clean the Si Oi. Cool the crucible, add three co of H. Sl and then enough hot Ho O to just cover bottom of dish. This was allowed to digest for several hours. Fifty o oc of hot H, 0 were then added and the solution boiled for a few minutes. Si Op» the silica is insoluble and the other salts are now in solution. The solution was next filtered through ashless filter paper, the filter paper placed in a platinum crucible, dryed and incinerated in the muffle. The residue in the crucible is Si Oo and its percentage of the oricinal clay is readily determined by weighing. A few drops of H NO, were added to the filtrate to oxid- ize the iron. The solution was then neutralized by aduing TH, O# end a few drops in excess were addec to make it elightly alkaline. This solution wss then boiled under a watch glass for ten or fifteen micutes. This precipitated the ircn and alumina which was filtered cut while hct And washed with hot Hp 0. Sufficient H Cl was then added to the precipitate tc dissolve it and Ho O was added to mxke up One hundred oc of the solution. One helf this solution fifty o c was then taken and to this was acded fifty c c of hot Ho O and 5 ac of He Cl. Freshly prepéred stannous chloride was then added drop by drop until the solution becane colorless. It was then cooled quickly and twenty cc cf & saturated solution of mercuric chloride was added. This eolution wac then titrated with a standerd solution of bichromate of potassium the condition of oxidation being tested by means Of spots on @ spot plate. (These spots consisted of a weak sOlutica of potassium ferri cyanide K, Fe Cyg-) “hen the soots became colorless the titration was stooved and the Iron celculated. To the second hrld of the solution which contained iron and alumina a drop or two of H NO, was added and then it wes made 3 slightly alkeline with N Hg 0 H. ‘This solution wes boiled for fif- teen minutes, Ciltered, washed with hot Ho U, dried and incinerated. The residue contained Feo 0. and Al, O Having determined the Fe 2 3° by titration, the Peo 0. was calculated ard this deducted from the weight of the residue gives the percentage of aly Ose The filtrate remaining after the Fe and Al has been removed was usually about one hundred fifty c c which was alkaline. “his was brought to a boil. In another besker a solution of smonivwm oxalate was then brought to a boil and twenty c c of the boiling oxalute were added to the boiling filtrete. This was stirred weil and allowed to stand for ten minutes on a hot plate. ‘tt wes then filtered through ashless filter paper, dried and incinerated. This residue was Ca U and the percentarce in the oririnal sample was de- termined by weigning. The fire clays of the Cerro de Pasco district contained scarcely a trace of ma;nesia and we did not as a rule attempt to determine its percentage in the sample. “he following tables give the analyses of averare Cerro binder and fire clays and English fire clays. Clay Si 05 Al, C3 Feo Og Ca 0 Vinchuscancha 44.2 29,3° 205 A.2 RA Te soyllarisquisga F604 16.5 eT 5.8 " 56.2 20 02 3.8 6. Cerro Binder 57 4 24 6 54 207 New Castle-on-Tyne 51.1 51.3 4.63 1.4 " " 48 26 302 4.0 1.7 " " 51.1 30 «4 4.9 1.8 The chanisal analysis of the fire slays from Vinchuscancha and Goyllariscuisg2 showed thet the Goyllariscuisga clay was the better and this accordingly was the clay used at the plant except occasionally when a carload was shipped in from Vinchusceancha to fill in when we were short of the Goyllarisquisga clay. The cley received in carload lots was unloaded outside the plant neér the disintegrator and allowed to seascn @s mach te pos- sible before using. By exposine the fire clay to the weather it is softened and becomes much easier to handle in the disintegrator. Before runnine through the disintesrator the clay was sorted by hand and all lumps contsining pyrite or stained badly by iron were thrown out. The larger lumps were broken up by sledges until they were about the size of an egg and then put through the disintegrator which recused them to about pea size. Having passed through the disintegrator the fire clay was then mixed with the binder in the proportion of four cue ft. of fire Clay to One and one-half cu. ft. of binder in the wet pan where at the eame time it was ground finer cand moistened. At the end of about fifteen minutes the prepared clay was turne*? out of the wet pan. This wet clay was then carried by boys to the tables where it 8. was violently thrown into wooden moulds by the moulders. “he moulds were Of several shapes snd sizes depending on the dimensicas of the brick recuired. They have a shrinkare allowance of three-quarters of an inch to the foot which is about the amount the brick will shrink in drying and burning. From the moulds the bricx were carried to the drying room and racked. At night these fresh brick were covered with canvas su that they would not dry too rapidly and be too hard to press. The next morning the brick moulded the day before were pressed and returned to the drying room where thoy remained for at least two deys before being set in the kilns. As only one kiln wes burned at a time and as the burning took nearly a week, most of these bricks re- mained in the drying room six or eight days. Yhile one kiln was beiug burned the other was cooling and being emptied. As soon as one kiln was empty and while the other was still burning or cocling the setters began setting the dry unburned brick in the empty kiln after first repairing any dwnages to the floor or furnaces. The bricks were set five over six beginning at the outer side of the kiln and workim tovard the cextre. Between each tier of brick a space of from 2; to 1" was left for the down draught of the furnaces. This space was decreased toward the centre where the draught was great—- est and at the same time the height of the tiers of brick was inoreased. At the centre the brick were set six over six and this was the highest point of the brick in the kiln. “he kilns were filled from one door only on the side tovard the drying room. -fter the kiln was completely filled both doors were closed by wallizg them up with brick and mud and the kiln was ready for firing. 9 in setting the brick in the kiln 4n opening about five feet above the floor and four inches by four inches was left without obstruction from One door to the other. This was called the peep-hole 4:4 was used to Gevgimalue the tenperature of the mass of brick from the centre to the outside tiers. ‘Cezar cones were set in this peep-hole to measure the temperature at various points. “hese segcar cones are mace of porcelain and have known melt- ing points accordiage to their mmvers. Thus the number 11 cone melts at 2000° F, the number 12 at 2250° F. Having filled the kiln and closed all oneninczs, a slow fire is started in each of the fire boxes the damper in the main flue being left wide oven as at first there is very little draught throuch the damp bricks. “his fire drives out the last drop of moisture in the bricks in about two days. “his process is called water smoking. Fron the second day on the fire is gradually increased until at the end of the third dsy all of the furnaces are going at their greatest capacity. four men and two boys are required on each shift of twalve hours to keep up these fires at this stage. As scon @s the bricks become hot or sbout the fifth day the damper ie closed down half way as by this time the draught is very great in the stack due to the intense heat. abcut the fifth day 2 look into the peen-hole was taxen to see the condition of the bricks. Fefore opening the peep—hole the fires were allowed to die down somewhat so that there might be no smoke in the kilr to obstruct the view through the Kiln. If the segar cones in the peep hole showed no sigms Of melting the fire was renewe! «10 increased if possible un- til the number 12 cone fused in the outer tiers cf brick. This ususlly occurred about the sixth day. 10. Owing to the way the bricks were set the centre does not become as hot while firing as the outer tiers but as soon as it is a brirht red and the cone in the outer tiers has fused the kiln is finished. in cool- ing the kiln the heat in the outer tiers is driven to the centre which is really burned after the fires are out. as soon as the kiln is hot encvugh as shown by the segar cone or inspection of the peep~hole, a liberal supply of coal is thrown into each furnace, the danper is closed to within about one inch of the bot- tom of the flus, the ash doors and furnace doors are closed and the kiln is left for twenty-four how's. At the end of twenty-four hows the ash doors are opened. At the end of the next day the fire doors are Opened end possibly one or two bricks removed from the tops of the kiln doors. This drives the heat to the centre and burns the centre bricks. At the end of three days half tne bricks ix the doors are teken out sncd the next day the kiln doors are cpened entirely and the kiln cooled 2s fast as possible. Ey the fifth day the bricks were be- ing taken out and loaded on cars for the smelter. “he capacity of each kiln was from 35000 to 40000 brick, devend- ing ou the sizee “he standerd squere brick menufactured at Cerro were 9" x 43" x 25", These bricks weighed about 72507 per li. while the Englisch bricks purchased before the brick plant was constructed weighed about 7100# per M. The Cerro bricke were not of es high @ grade as the tnglish either chemically or physically. As stated above, oving to a lack of proper machinery they could not be ground fine without greatly reducing the output of the plant. This necessitated the use of more bincer clay ll. than wes used in the “ngelish brick which in turn decreased the refract- Ory and increased the fluxing qualities. A chemical analysis of the “uglish brick showed that it con- tained Si 6, 61.3% Al, 0, 35056 Fe U 1.54) Ca 0 1.45 vith traces of magneslA. The average Cerro bricks contained Si Oo 54 4c Al, 0. 36 64% Fe 0 4.39% Ca 0 3622. The percentage of fluxing materials, viz. Fe 0 and Ca 0 it will be noted is much ¢ereater in the Cerro brick than in the English brick. Yor this reason whiie the Cerro brick was good enough for boiler settings, flues, coke ovens, Xc. it would not do for the reverberstory furnaces which the company proposed building soon after the writer left “erro. In order to see if it were ncssible to secure a high grade brick with 9 minimum of fluxixng material we undertook some experiments with silica bricks. Near the co0ai wiies at Coyllerisquisga the railroad outs through some sandstone bluffs which seemed to contain very little if any iron. cCamples of thie sand were secured, analyzed aiid found to be Buitable for silica bricks. The analysis of this sandstone wes made in identically the same way as the analyses of fire and binder clays. The Goyllarisquisge sendstone contained 93.6% Si Oo, 1.56). al, Ue and <7 Fe 3. “he problen was to mix the sand obtained by putting the exnd- stone through the disintegrator with sufficient binder to méke a brick which would not be too friable to bo handled or burned and at the same time not to increase the fixes to an extent which would impair the value of the brick. 12. lire Henry Vogel an exnerienced fire brick men had preceded the writer @s superintendent of the Verro brick plant and had made ecne experiments with this send. Mr. Yorel suzceeded in making a brick which met the recuirements and seemed very satisfactory. “hen the writer tock charse of the plant ir. Yorel ecve him his formula for these briczs °:d explained how he had made them. “re Vogel stated that he mixed eight parts of dry sand or silica vith two serts of finely srceund dry bincer clay and wet this mixture down with a sclutioa of one part lire slacked in six pails of water so that it was just damp enough to te hancled in the press. 4 ve attempted to make these bricks accordiiz, to this formula several times and thouwrh we used a snecial dle and press were unsuccess— ful in getting a brick thot was not very soft avd crumbline. 3ricks made in thie manner contained Si 0 90.1% Al,, 0, 6.45, Fe 0 1.8% Sa 0 1.9%, so that in addition to poor physical characteristics the chen- ical iugredients were not un to the etandard. A ¢o00d silica brick siould cont:in not less than 92%) silica. after trying in vain to get satisfactory results by mixirg with very little water as sug;ested ty ure Yogel ve determined to try mixins in the wet pan and proceeding the eare as with ordinary fire bricks. For cur first experiment with this method we used 400% of ground sandstone 754 of binder clay and 10# cf clncked lima. This was thoroughly mixed in the wet nan, the lime being added in the form Of paste. The bricks were moulded, dired for six hours and then press- ed. “hese bricks driec well after pressing and seemed hard snd firm but when burned they turned out too soft to handle. ‘Tre chemical analysis too showed too much lime though the brick was hieher in gilica. SiO 91.5, Als C. 6.6% Fe O led Ca 0 3.9%. (Note this analysis shows over 100); due probably to the fact that H, O was absorbed by the Ca O before it was weizhed). “he secord experiment wes much more satisfacwory and we secured 2 brick which while not up to the Singlish standard wuld be without doubt entirely satisfactory. For this brick we used 4007 of silica and 754 of bincer clay ground fine mixing them in the wet pan but not adding lime at all. This brick was moulded, pressed and dried as with the first brick mixed in the wet pan. it dried well and burned mich better than any we had tried tefore. ‘ie consid- ered this to be a fairly successful brick and the writer believes it to be better than those itr. Vogel made first because mixing in the pan was much cuicker and less expensive than Mr. Vogel's dampening process, and seccnd, because the chemical analysis was better. “his brick contained Si 0 91.7% Al 0, 609% Fe 0.5%; Ca 0 =, 2 In these experiments the writer was very efficiently assist- ed by the foreman of the factory who wes en exceptionally intelligent Cholo or native. The entire factory force of laborers was composed of Cholos. ‘Nile these natives had some very bad cualities, yet if handled with a little consideration they were not bad workmen. ‘early all of them are netural thieves and unless watched closely will] steal any tool they use. At first the writer had some difficulty on this score but finally fot the men to be fairly honest by the rather unjust method of deducting the velue of any article or tool stolen proportionately from the wages Of all hands. After losing a little of their wares once or twice the 14, innocent ones soon brought suf“icient pressure te bear on the guilty to keep them from stealim from the fire brick plant though all of them had no hesitation in stealing from other d3pcrtmente of the cunp- BY “he wages of the men varied from forty cents 4 day for the boys to four sols (1 sol equals 50 cents ‘merican) a day for the fore- mane “he ceanacity of the “erro plent when operating sterdily was about 2509 bricks per day. “he average cost of these bricks covering @ period of six months was about 12 E Sterling per Lh. Compsrin this to the cost cf “nglish bricks delivered at Cerro for 49 Zz Sterling, it is easy to see why the fire brick plant was constructed snd what an exceptionally payinc investment it really is. _ " eo = Se = Ax a ora _ a. MICHIGAN STATE UNIVER: win 3 1293 500 SaREEREEREEEEa@ se Hitt ageanE “ss eee