Design of an Electric Furnace for a Brass Foundry Citation Bissiri, Alfio and Woodbury, Roscoe Errett (1920) Design of an Electric Furnace for a Brass Foundry. Bachelor's thesis, California Institute of Technology. doi:10.7907/64w0-w147. https://resolver.caltech.edu/CaltechTHESIS:07262024-162626612 Abstract No abstract. Item Type: Thesis (Bachelor's thesis) Subject Keywords: (Electrical Engineering) Degree Grantor: California Institute of Technology Division: Engineering and Applied Science Major Option: Electrical Engineering Thesis Availability: Public (worldwide access) Research Advisor(s): Unknown, Unknown Thesis Committee: Unknown, Unknown Defense Date: 1 January 1920 Record Number: CaltechTHESIS:07262024-162626612 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:07262024-162626612 DOI: 10.7907/64w0-w147 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 16594 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 26 Jul 2024 16:52 Last Modified: 26 Jul 2024 16:52 Thesis Files PDF - Final Version See Usage Policy. 10MB Repository Staff Only: item control page

THESIS Design of an Electric Furnace for a Brass Foundry by Alfio Biasiri R. E. Woodbury of the Class of Nineteen Hund.red Twenty Department of Electrical Engineering California Institute of Technology Pi;.oadena, California 1920 REFERENCES l. Rodenhouser, Schoenawa, Von Baur: "Ele ctric Furnaces in the Iron and Steel Industry" 2. St a nfield: 3. Iron Age, Ji. 26, 1919 . 4. Electrical World, Vol. LV, No. ,34 5. Electrica,l World, Dec. 20, 1919 . 6. Met . and Chem. Eng. l Dl 7. "The Electric Furnace 11 June 16, 1910. 17:461 Oct. 15, Design of an Electric Furnace for a Brass Foundry Table of Contents I PURPOSE II SELECTION OF TYPE III (1) Induction Furnace (2) Resistance Furnace ( 3) Aro Furnace DESIGN OF li'URNAOE ( 1) . Size of Furnace IV (2) Dimensions of Electrodes (3) General Features CONCLUSIONS I The Problem to be Solved It wa.s necessary to choose a design that would be pract1cab1e a nd eoonomica,l in handling small quantities of brass or aluminum castings. The attempt wa s made to cut dovm the almost prohi bitive initial cost and at the same time to offer a design th,;.i. t was efficient and com- paratively simple in its mechan ical and electrical details. Other details of the problen1 were rapid handling of the furna.ce and its contents, and easy access to these. II Selection of Type The Qualifications of 1nectric Heating The general cha racteristic a which make. the study of the ia.pplica.tion of elect1·icity to foundry. a practica l rind promising undertaki ng are of gre a t va lue. Hi gh t e m- perature a which a :l' e not po s sible with any syste m of gas heating , ar e obta ined in a. i: 1hort period of ti me with the use of electricity. A positive contro l of ~he tei~)erat ure of the b a t h is made possible thi'oug h e lee tr ic op er a tion and a ny definite temper a ture may be t~usta.ined t hroughout the time required by t he p e..rticu1ar de mands of the r,rooeas. in qUQBtion. Wit h the use of gas or oil as fuel in the proc~sa of melting the c ha.r ge, foreign and ot her ha rmful subat a ncea are introduced into the original mixture. Electrici.ty offers the cleanest he a.ting agent thua far _ known so that a ll the deleterious effect s whleh other heating processes have are entirely eliminated. Electric furnaces oper ate entirely without the need of a supply of air, wh.ich makes possible operation in any atmosphe r e and preventr:) reactions which may be caused by elements in the air, gases, or the products of' combustion. ~ The requirements which a practical ~urnace must pos sess; . We found.that the requirement s of a.ny furnace which shall give maximum resul ts are as follows: ( 1) The ability "t,o use any prevailing a.l ternating cu,,t'rent at any voltage and frequency. (2) The avoidance of any Budden change in the load. (3) Ease of regulating the i ncoming current. (4) High Electrical Efficiency. ( 5) The a.bi 11 ty to re ach any desired uniform tempera t i1.re in all parts of the l:ath 1 and at the same time avoiding any local under or over-heating. (6) Oorrip lete uniformity of the material in all parts of the molten metal and consequently a sufficient circulation in the bath. (7) The ready adaptability of the furnace to intermitters service. The Selection of the Type of Furnace Best Suited to E.tny Condit 1 one. #3 In seleating the type of :furnaoe we oompared the merits of ea.oh with the standard desoTibed above, that is, the ideal furnaoe and the typid which satisfied the larger nu.m·b er of the reg_u irements was adopted as a working bas.is. All of the various furnaces devised operate on the simple principle that when an eleotrio current flows th,:ough any oonduotor a certain work is is done which manifests itself as heat, the principle still being fixed the only work to be done re~olved itself into two main divisions. One was oonoerned .with the question as to how the large ourrents required are to be prod1.:oed; the other was concerned with the question as to the resistanae through whioh these currents be passed to produce the neoessary heat with the maximum efficiency. The two questions whieh arose were fully investigated by independent · workers; all the possible combinations were worked out and developed. '.l.'hree well defined combinations were estab- lished, u1Jholding methods entirely separate, but :fundamentally identical. These com·binations are: (1) The induction type of furnace. (2) The resistance type of furnace. ( 3} The arc tirpe of furna.oe. 1. ·a odenhs.user Sahoenawa Von Baur Electric furnaces in the Iron and Steel Indus'try. #4. Each has been investigated and compared with theoret1:_pal standards in order that the one beat aui ted to the cori&it.-~ons required miftht be selected. 1. The Induction Type of Furnace ; When f.m electric current flows in. a. condu.ctor a m"tznet:to ••.. field is gonero.t;ocl around the conduc tor, and if the al tei!i~~tinp; current te. maclo to f1ov, t 1ie · magnetic field will be in phase with t he current ~.nd altermi.te from zero to a ma.xir,um, duced upcn it a.r.. ,i tt.i1..'l c urre nt Nill be rr cY;,. ,crtic,nal to the mmber of line a of force out in u nit tir:iC. ci'plc is t·uil t !;hE:i°' t~,i-,e Up oE t ,,i s y;!'. in- of iiit~u.:: tion furnace. produced wbe rt t he i n(!. u<.H'.ld ou.r rEn .t t in t:<i.'1.de The ~te~t io to flow th.r ough F.t.. exce ssi vo ho ;c;..t. fol 1oYrn : turn fo~ the sceonda,ry winding , wtiich i ::'~ f or me d. by the rz;oJ.-t en metal. The prirnal"y winding, P (se c fig. l.) is wound <)n lf5. an iron core cansi stin g of oxie or two legs and h aving the bath of met a l in the form of a circle around .t l1e i r on core. 'fhe current in 1; he primary coil, P, genera.t e n lines of force, ,{,, which tr cwel i n t he iro~ core. The induced cur rent· in the secondary turn gczu r a tes line s of force which take a p a th in the iron in t he op r, o n it,e direction, J.,,, r, The curr ent strength· in the n'etal can -p 'be .varied thr<'>Ugh very wide ranges ,,- r=-t9.1 by inor ea.sin g or decreasing the number of turns ,, in the priniary winding a nd. by ch n.n g ine; the voltage .a.p p lied to the. pr i- mar y. Sp tc::cia l pro v isi ons r:nurrt be t Rken to p r otect t he primary winding frow t he i ntenae he at of the bath and a water coo line; system is used to keep the iron of the core at a low temperatu.r e. Co rrp a rison With St o.ndnrd Furnace This furna c e fulfills the fir at r· equirement since 1 t can be built :f'oi· single or three p has e cur rent c>,nd a.t a ny voltage ob1Hi'"in~;ble. The oper a tion of the furnace.cannot t a ke place if the char ge o f met 4l,l is cold scrap since ~- comp lete ring- is not formed rand there f or e the secondary winding is not short-circuited. For t hia reason all of the met &l cannot be poured, but a sma ll quantity formiri:.g a rin g is left to provide for #6. the resumption of the operati on with the next charge,- This p1•ocess is obj e ctionab le, since the remov a l of all qf the slag is made impossible. The second requiaite is only partly co mplie.d w,i th. The. inco ming current c a n be e & l. sily re gul a t ed, since ii0-1 that is needed is a var ia.b le rheost a t which vari e a the vo:i\;tage app lied to the p ri m&ry coil. The hi e h cff ic iency of the furnace is due tor 't})le absence of electrode los s es a nd also to the -fact tht.1.t the l:O~·.aes which take p l a..c e in the transformer oc ••;ur in _the for m 01; heat, a nd since the -r,urpose of ,: the furnace is to produce _b~_a.t, tht:H-ie losses are utilized. • Requirements five and six are satisfied by ,~a!1a of the same peculiar conditions which a re c aused by t he'. fi{lfluence of the thf;-l..gnetic fiel d s gener a t ed 'by t he primary a nd B'~u'Ondary. i'' The p ri ir1a ry ,;...nd. seconda ry are not; in direct c onta c t c,ut have a.n a ir sp ace intervening . Some of t .1e 7: .lines of force wil l not interlink t he coils but for m close d p1:1 tl ! s around the coil, ;:~:ene r r,ting the flux <// a nd c/:/'. 'I'hi(t"' \ 1.' .Lux path s are in direct ions ·opposite to e ach other but ~, 0 mi the accomp a nying di a gr fam it c tm be aeen tha t the oppoHit ,, direo- \ ,' tion of these linea of force makes the d1_rect1on of ,':fvf~re,nt flo w the same between t h e primary winding and the ·b~~h. Since i ·1 lines of force of the s a me direction repel e c1ch other"~ f'orcea ~ . .. must appear whic;h r epel the molten metal from the p ritta.ry ' to · th~ outside. This repulsion does take place in practice and the level of the metal is that produoed by the combinations of the gT H,Vity force ( g) and t he repelling force ( r.) at right angles to each other. The level of the .b ath is at rigtlt :• angles to ·' the resultant of the combination of the two forces{ ( . At the same time a. flow of the metal takes place _from the outsi~e edge to_; I wards the inner lower one providing for il unifor m And intimate temper '}ture and mixtur e of the bath. The last re.quirement is not · fully met, since · ~( :lith'e r some ·\! part of a previous charge must be left in. the furn;ce , in which , \ case a different composition of met a. 1 cannot be ~e1~ed, or separate ring~ must be provided, made of the m,:t$.l Which it is desired to fuse. While the fundamental :principles m.:re more or · less ful- filled, another cond:i. tion must also be oo r, sidered :\\; ~'his is the power factor of the system. Experiments show tht t with the •. 1, ' . 1· . .\ . • same frequency th'c p ower f a ct or decre ases with an amount of metal charged. inoteased ; '.~ Thio is exr<lainecl when the ':\: cqndi. • tions that go to fi,r:~ke up the quantity repreaenting the po.wer factor are examined. cne - -f . i: == Ji--~_,_.z.; r= l( ,::l e¾ ,-rrf L Therefore, for a cert r.d n furnace 1 , f r r~.d and L a.x e fixed\ que,nti ties. =/f8 .• The value of r decreases with . increase in charge or orosis-- sect ion. of metal; therefore the only way to imrea.Qe the power factor is to decrease the value of the frequency. Z or ( Fig. 3. ) decrease This meana t'hr t a different frequency would . have to be used for different· quantities of oharg~.s , so as to make the change quite impossible. Conclusions "'' While thia tyi;e of fu.rn<),ce haa many good fea.ti,.,.re s, yet it is net aday,ted for a brass tound,:y, because 1n t\, brass · foundry different mixtures o:f met;:il are poured. every day, necessitating new lining for every charge. The present f~e- quenoy of the city system 1s sixty cyc.les. which is much highex than is ret}uirtHt to give a good ;)ower ta.ctor at muimurn charge for the fur.n:~..ce. The d.epreoi htion on a small fur- nace is much higher per ton of t11etal t>roduced because of the small amount of . ni.: tal oha:i:ged. at every run, while the whole appa:ratus is cub1-r>1.tted. to the extreme heat to ilhich· a 1,a:cge furnace would be En:.bjectecl 1;1-.at prod.uced mlimy ti n,e~ the am.c unt not t\do-r.ited. • The Resistance Type of FUl'rtrwe When an electric ou.rrent f,lows in a. conductor, heat 1a developed and the quantity is: Q = 0.24 it r I cq/4rt'a.s trom this 1 t is evident that if a. large current 1a pasaed through a small resistance, as for example, a l tu-ge volume of molt;en metal, a, high temperatu;re ia obtained. In order to maintain the current at a reasonably small value, the rea1a·~ance of t he bath must necessarily be increased. The resistance of any material is expressed. by the formula: 1 = length f ,.. s pecific reaist a.noe a ; croas-aection::1,J. @.rea, It is evident tl";.u.t the only tern;s ·which c&.n be varif3d are 1 b dicates t i, f.l oh,,.n nc.:l :J :i.:rl .hh;h. t he 1:et a l in confined; c 1 in- is the e 1jj4 in~ to the c r es s Be ction of the bath. cauae flnctua.tions of power input. The third re(rui~:ement is ectsily s at isfied with the help of externa l resistances. l!:l·Q •. ti.... The electrical eff1o1enoy is very high, since no loss-es. a.re encountered; but thii'3 ia offset by the large rad.1.a.t\on losses which are brouc:ht about by the very long exposed bath .. ., The f i f' th r e quirement ls :fulfilled by the air(1p le adjuet- ment of the impressed voltage. The heating produced 1e entire•'." ly uniform throughout the b ath. Girc ·ul i~. t1on of the metal is not produced without the a.id of external rnethod.s. Thc:? furnac e in c-"Ldapted to intermittent service, since oper- Co:nclutdc n a The good points of thiB furnrwo a.1·e the u.1~ifo1·m hern.t 1ng of ser ioua.. ! 1:1 0:rc.er \: o. \ •roduco t ··c b e x t eleGtric ::: :L oond. i tions, long anct n a.t·row ch . .nnelf3 rirun~c be constructed.. expor~c · ) . ~;:-J.•• !fit} ?, .,_._,_.f',_·_1.{·•., "-'i'." 1.~~~ ~~r. .. - ·~ These ohllt:rmels c•,1" ·,·.,,,,~,t ,";,,,-'·1 '.'.~ nd the:r.efo:r-c o r:.u90 exooaaive ~ .. '-' _ ¥ or inriut. 1mr,ossil)1.e. ,. eci).use B r- o ~~ th·"~e ., o ,., • ,,,..-· ri··)u"' .~,j.. ,. _,,, 4.,_tJ.• :• l t 13 •. -~ u. th.i n t,r:;e of ftt:rmwc wae ., 1 not ad.opt~d. The Arc Type of Furnace .#11. When the ends of t wo conductors ·carrying current are brought in contact and slightly sepaxa.tod, the current continues to flow, forming an a.re. When the arc is formed t o c11rrent is not carried across the gap by the intervening a ir but by the gaaea v:hich ema na te fr om the metal constituting .t he conductors. The theory of the way 1 n which the arc occur a ia very simp le. At t he inst ant when the conductors ar e slightly aoparated . a high resistance is offered to the flowing current. This gener ates s.n int e nse heat which gasifies the metal a nd ~ills the gap. Al- thcugh the current ca.n pasa fro r.\ one · conductor to t i'.:.C· other with the z~id ot' the s mc.11 t:iotal pa1·ticles, it is opposed by i9.. very h i gh re s ist 0.noe, s,nd thi s io tl:'hJ 0 ,1.u se of' the high temp •a r a tures The c t;;.:c ;:;on arc gives tho hi ghe13t te mpera.... produced by the arc. tUre so f e.r re,'.'.ched, ab out 3 ! 50{) C, . this being the temp,e ratu.:re v.t wh ich o;.,~rbon r;aeef ies; and t >.is i s t he o ne uaed in the ar c . type o:t' t' urna.ce. One of t he ai rny,l est arc fu:rna.cea c onsist a of &.. he a rth (a), a b a.th of metal (b) and t he carbon electrodes for ming the arc, ( c). Comparison with St an dard Fur nace Th it1 type of furnace c tu 1 u se any cnr ro nt wi-t;h .e.ny number of phases. Sime the :metal does not come in C()ntact with th.e ,J...~o, t he operation of the o..:rc is not d ist u rbed by t Le proe(:;;SS of melting #12 . • large pieces of scrap. The incoming current is easily regulated by varying the .• gap between the electrodes~ and this adjustment can be made within na r row limits. The electriC fJl efficiency ia hi gh, since the only loss e n .., countered is th.si.t due to electrode losRes. Wit h mecha.nic a.l mod ific a tions the he a ting of u.11 the p arts of the bath can be made uniform. With ~he a ai!;e improvement the. metal is mixed thoroughly into a h:rn1ogeneous .trt13-ma . . Beo auae of the ai mple conatruc t ic,n of ·t M.s furns:.ce t;he lining c a n be cleaned so t hat different mix't t-1.res of metal mfJ.Y be p1 aced in t 11e f urnace in succeasivf~ runa. Conclusions ;ia t h t h e mec llani cal niod i f 1c ~•-t ions t hat must be n:-1'1,de, the e..rc ty p e of f'u:rna ce fulfills all of the r e qui r ements ap oc tfied. Ttds ty:p e wa s t herefore Rd.OI.Jted as :a, working ba.sis, and a part of t Le task of the t hesia oonsiste,i in the wr;rk1nr: ou t of the · rne oha.ni cn.. 1 fe at ur es of t his fur ni1ce. Design of Ji"'hrnace The · requil' c rnent was for a desi gn of furnace sutt ab l¢ to ac c o r:irnodri.te omall amounts of 'b r e oa metr.,l f :requ~m tly and r ap idly. The mechan i c '.1 1 • det a il s t he r efore, h;:1.d to b e i:ie si e;nJitl "'d th ·t hese object a ip view. The desi g n WH.s exp ected to .comb ine s i mp licity and economy with ease and r ap idity of h@,ndl 1ng. #13 Size of Furnaoe The size of the furnaoe, it was agreed upon, should be large enough to hold 250 lbs. of molten brass or copper. Sinoe a oharge of copper requires B. T. u. per pound to produoe a melt than brass the amount of current required was aaloulated taking the value of B. •1i. lJ. required to melt a oharge of 250 of oopper. Since 85 watt-hours are required to melt one p'ound of copper; 250 x 85 = 21250 watt-hours The electrode loss is; 400 x 2. 2::. 880 watts per eleotrode 880 x 2 ~ 1760 watts total 1herefore the amount of eleotrioal energy required · is 23010 watt1 hours. Assuming a total effioienoy of 75~ 23010 ,,. .75 30800 watt-hours. This amount of energy will meet 250 lbs. of oopper in one hour. The time required for each melt was desired to be cut to 45 minutes and the amount of energy must necessarily be inoreased. 30800 .75 1: 41000 watt-hours. 400 ampetes at 110 vol ta .... 44 Kw. hours. Dimension ·of ~leo~rode Table of Electrode constants. Carl Hering. .c Quirent Sg. au !Qllii . ln. Units cm.Unit;(! .000036 36 ~6 36 '70,300 71,800 '12,800 '13,400 ''3, 900 74,300 .000034 0.18 .22 .26 .30 .34 .38 .42 74,600 21. '100 28,300 28,700 28,900 29.100 29,300 29,400 .000100 100 .000099 99 98 98 .000098 ,00026 25 25 25 25 25 25 .29 .30 .43 .49 .66 .63 .10 10,000 10,000 10,100 10.100 10,200 10,200 10,200 3900 3900 4000 4000 . 4000 4010 4010 .00025 23 22 22 2l 21 20 .00020 .00063 59 57 1.3 1.4 55 1.5 1.5 1.5 1.5 Lb 4100 4300 4500 4600 4700 4800 5000 5100 · 1600 1690 1760 1810 1860 1890 1960 29-10 Section S Tenm. Drop. 0 F Sg. in. 0opper 800 1000 1200 1400 1600 1800 2000 1472 1832 2192 2552 2912 32'12 3632 .000014 14 14 14 13 13 .000013 Iron 800 1000 1200 1400 1600 1800 ·2000 1472 Graphite 1000 1200 1400 1600 -],800 2000 2500 3000 4532 5432 34 34 54 53 51 .00060 1.5 i ,' #14 ::section s Temp dl'Op C F Sg! in ~g. OU current 0 i Volts Units Units 1.9 2.1 1360 1570 530 620 690 750 Carbon 1000 12 • 000'/4 64 57 .00lt3'/ .00162 16 52 .00145 132 18 48 123 14 20 45 25 40 116 102 .00037 .00093 3000 2.3 2.5 2.1 2.9 3.3 3.7 1'750 1920 2070 2200 2600 810 2700 1080 8'10 980 The eurrent required, the temperature differenoe between the aro-end and the oool part of the ~leotrodes and the length of the eleatrode must be predetermined by the requirements of the :tlll'naoe" For the drop in temperature required the value of (S) is found in the table for the ma.tei-ial. which is to. be used. '.J.lhis numbe:r (S) is the speoifio section for that drop in temperature. it is analogous to specific resistance that is, to ohms per oubic inoh or per unit-foot, eto., at a high temperature; it represents the section per ampere and per unit or qen\imeter · 1ength. _The proper size for the electrode is then found by multiplying this faotor (S) by the amperes required and the length in inohes or aentimeters. Oirou- lating the cross -section to: the various materials which may be used the oomparative oost of eaeh is found and a seleotion made rel,.. atively- simple. The dimensions ot the electrode thus foroed will be so that th~ heat lost through conneotion by the eleot,:,ode will .be 1,.: ' the minimum in conjunction with the minimum loss due to the r ·e sistano, of the el~ctrode. To determ~ne the loss due to the resistanoe of eleot:,:ode the value of (C) from the table ie found which oorrespond!ii to the oonditions decided upon. th~, ourrent will give the Multiplying the value of (C) by loss due to each eleotrode. If it is required to find the proper ourrent value for existing eondition and size of eleotrode, the value of corresponding to the temp. drop. as taken f~~m the table ~!.'his value is again a speoifio quantity, being the ourrent per square inah seotion for an inch length. Thus number is multiplied by the existing o~oss seotion and divided by the length. In this particular oase the ou.rrent re- quired is 400 amps. and sinoe the melting temperature of copper is l083 t C a temperature of 1400ba may be taken for the ara temperature. The temperature of the oold end of -the eleotl'ode is kept at 100" C ' 0 giving a temperature drop at 1300 c. Carbon was ohosen a.a the material for the eleotrodes and from the table the value of "S" 0 oorrespond.ing to a 1300 . tem p. diff. is .000675. The propel' oross-seotion is: .000675 x 400 x 20 5-4 sq. in. or 2.6 in. diamete~. A standard size is chosen . 2.5. The electrode is therefore 2.5 · 1n. by 30 in. fhe Design of Fu.rnaoe Electrode by Carl Hering Elec. World Vol L N No ~4 June 16, 1910 #16. General Features The furnaoe is oylindrioal in form, the axis of the cylinder being horizontal. axis. The furnace revolves about this The oylinder is sup_portcd by rollers whioh transmit the motion to the furnaoe, the rollers being in tu?n revolvt'H\ by f'rioti on oontaot with a pulley fastened to a motor shaft. There is ·1 door at o i ther ond of the i'urnace. electrodes pass through the oenter of ea.oh door. The oarbon Power is transmit'ted to the furnace by means of oopper a.hoes which slide along a eopper band around the turnaoe. ally oonneated to the eleotrodes. Thia band is eleotrio- The electrodes are cooled by water ooils surrouncUng tnem, there being a steady oil'~ula.tion ot water kept up at all times when the turnaoe is in operation. The Furnaoe Proper The furna.ee proper is m9 de up of a steel cylinder. 1'his 1 oylinder ls lined with fire briek 5 inches thiok, the bl'iolt growing thicker toward the ends to give the inside surface a more nearly spherioal form. ~•he inside of the :rurnaoe is 18 inches in d.iam.eter and 20 inches in length. :rhe ends of the cylinder are alosect with oast il'on plates whiah are riveted to the steel oylindel'. the enct plates are lined with 4 ineh fire briok, except at the openings for the doors. oast iron lined with brick. i 1he doors are of 'l1he doors swing on one large hinge, whioh isr bolted to the end plates. The doors are fastened by means of a lever oatoh. 1~rough one end plate and the fire briok back of it, a hole 1t l7 is drilled. This hole opens at the innet edge of the inside. of the furnace. The outer opening is pluggeG. with a movable fire clay sto,p per. Around this hole, on the outside, a metnl spout is bolted to the end plate. To empty the furnaoe of its bath, the furnace is stopped so that the apout is at the lowerside ot the c,ylinder. When the slug is removed it leaves a free path for the metal bath to run out. Around eaoh end of the cylinder a circular ring of 4. inah ohannel steel is bolted. These rings al'e entirely insulated from the rest of the furnace by placing heavy press board insulation material between the rings and the cylinder. insulated from the ring and the oylinder. Every bolt is also The weight of the furnaoe is carried by these rings, wh:i.oh rest upon two parallel rollers. These rollers are driven by a friotion pulley on a motor shaft and they in turn transmit the movement to the furna~e by bearing friotion with the channel rings around the cylinder. Power Transmission and Commutation Around the inside of the outer flange of eaoh channel ring. a band of copper is fastened, but this band is insulated from the rings. Copper shoes are suspended and held against the oopper band by slight spring pressure. These shoes are connected by wires through the proper switches to the transformer. The current from the copper bands is carried by wires to metal olamps which surround the oarbons. 'Phus the 01.urent oomes :fxom the positive lead of the transformer, t~rough the -switches to the shoes, whioh are in aontaet with the band of copper a:round the furnace . .From the band the current travels by wire to tl1e positive eleot:rode; aras over to the negative eleotrode .and on baok to the transformer by a similar path. The Electrode Carriage The oarbons are perperidioula.r to the plane of the doors. Three equally spaoed rods extend from the doors for a distanoe of about 2 feet. the doors. :ehese rods are also perpendioular to the plane of At the ends of these rods an ifon plate is bolted. The whole structure is strengthened by six braees. braces to eaoh rod. A pair oi These braoes are bolted to the plate, each pair diverges to the outer edge of the doors. where they are bolted. Along these three rods slides a spider, but it oan be olamped in any position. Through the eenter of this spider passes the oarbon and above that a threaded shaft. The shaft is held in position laterally with respeot to the spider by ootter pins. :t.ihe threaded part of the shaft screws into~ olamp whioh oan readily be fastened to the oarbons. is a orank. At the other end of the shaft there By turrdng this orank, therefore, the oal'bon is moved in or out at will. If a greater and quieker ohange of the aarbon is desired; all that is neeeasary is to unolamp the spide?' and slide it and the oarbon in either direction and ala.mp again in the new µosition. To remove the aarbon entirely all that ie nea- essary is to unclamp the threaded clamp which holds the carbon. The Cooling System At tl{e point where the carbons pass through the doors, they are surrounded by a eoil of pipe which is oast in a blook of metal that fits in the oenter of the doors. The oarbons are insulated 11=19 :from this blook by a thin graphite sleeve. 1'he water aomes into one ooil, passes through it and is aarried out by a flexible hose to a pipe whiah runs horizontally along the outside of the cylinder to the other side. At the other end of this pipe another flexible hose aonneots it to one end of the aoil surrounding the other electrode. The other and of the ooil ie aonneoted to a pipe whioh runs along the electrode oarriage. At the ends. of each oarriage the water 11ipes ara aonneated by flexible hose to elbows whioh are held on pivots, the pivots being fastened to the end plates of the carriages. The centers of one side of the elbows exactly ooinoide with the oentel' axis of the furnace, when the elbows a.re pi voted into position. '.!.'he elbows are pivoted so that they may be swung out of the way when a oarbon is being removed from the carriage. It must be :remembered that the furnace iB revolving and that the pipes and other attachments are revolving with it. Thus it was necessary to bring the two water pipes out with their oenters coinciding with the axis of revolution. After this was aooomplished it was neoessary to oonneot pipes which would be revolving with others whioh were stationary. This was done by designing a short pieoe of pipe whioh screws into the ?'evolving elbow. 11he outside of this _pipe was to be polished to a shafting finish. ary pipes. Unions were threaded to the station.. Over the revolving pipes paoking rings were placed. I ' The stationary pipes were brought up until their ends almost touohed :ff 20 the ends of the revolving pipes. The puoking rings were then screwed into the unions, whioh were on the end.a qf the stationary pipes, until tight. 11he revolving pipes were then free to re- volve, and yet the water would pass from the stationary to the revolving pipes, and vioe versa, with little or no leakage. The stationary pipes are held in plaoe by braoes whieh run slantingly down to the stationary hose of the furnace. Hose or other pipes may be attached to these stationary pipes of the furnaoe. The water goes into one of these pipes to the revolving pipe, on to one coil, out, and over to ·t he other eoil. The water passes out ot this ooil and on to the revolving pipe on this side. From here it passes to the stationary pipe and is carried away wherever desired. uono lusiona: The outstanding features of the furnace are: (1) fhe aomplete rotation of the :ru.r·n aoe gives a longer life to the lining beeause the bath absorbs the exoesa heat whiah the lining absorbs when it ia exposed to the heat and not covered by the bath. ( 2) 1'he lining is evenly consumed and a longer life made possible, sinoe only two joints are present and these ean be filled with oement. (3) The electrodes oan be moved out quiokly, and in this way the doors oan be opened with little delay. ------- --- ---------