CONSTRUCTION OF A RESISTANCE TYPE FURNACE Thesis for the Degree of 8B. S. re anil yoty WEP \a lat tela) ya tS wT CO CONSTRUCTION OF A RESISTANCE TYPE FURNACE. A Thesis Submitted to The Faculty of MICHIGAN AGRICULTURAL COLLEGE By . m4 No . “J , a. Ri Sarleon F. A. Johnsen - “ s Candidates for the Degree of Bacheler of Science June 1921. TReESIC Gq. | TABLE OF CONTENTS. Intreductlonscecccccvcscccccccccccccccs erage Study ef Materials and Type@e..cccvccces Study ef Local ConditionBesrccececcececes Construction Detallseccccccccsccccesecse Preliminary Tests. .cccccccccscccscevccee Comolusiericcscccccesccscsccccccssccceces PLOCUTOBs ecccccccccccccescvccccsneecsece Sectional View = Prontecescccccccscccces Sectional View ~ Bide... .cccc-rcccccscccs Connection Diagram....ccccccccccaccccces Bwitchboard Panel........ccsceccceeseece 9430% INTRODUCTION. In the fall of 1920 a course in Electre Chemistry was effered for the first time at M. A. C. All those who teok this course were given a rather hurried view ef a number ef uses for electric energy, aside from the usual applications fer pewermé light, which are not only interesting, but very useful and of great commercial importance. Among other things, a study was made ef electric furnaces of different types and their proper applicatien to different industries with some of their limitations and advantages. During the early spring the Chemical Experiment Statien tried to get an electric furnace with rather special char- acteristics for some of their work. A temperature range fren reem temperature te 1000° C., was wanted, no regular stock furnace with this wide variation sould be obtained. Dr. Rebinson, who is in charge of this work, censulted with ‘Prof. Sawyer and it was agreed te let the building ef such a furnace be a thesis subject for the coming spring tern. The problem as outlined was to build a furnace with a temperature range as given above, having a heating chamber about four inches high, five inches wide, and twelve inches leng and capable of running continuously for days at a time. STUDY OF MATERIALS AND TYPES. The choice of a type of furnace at once narrowed down te some type of a resister muffle furnace. An aro furnace, altho of rather simple construction, would have been out of the question, as the lower temperatures could not be obtain- ed at all, Successful laboratory furnaces have been built using earbon granules, or carbon plates as a resistance. By vary- ing the pressure on the granules, or plates, the resistance ean be waried and in this way considerable ranges in tem- perature can be had. However, the range of temperature from room te 1000° 0. could not be reached and it weuld be necessary to have a rheestat of large capacity. Also the earbdon plates er blocks gradually burn out and need te be replaced at shert intervals, In the granule type the gran- ules become packed and upon reducing the pressure it is necessary te shake the furnace to get the particles to loosen up and then return to original ocontaet pressure. The carbon resistance furnace is a complicated structure and while it works well for higher temperatures, it is not readily adapt- able to such a wide temperature range and due to the incan- venience of replacements, this type of fumace was abandmed, The use of a metallic resistor was then stu@ied. Re- ferring to a wire resistor, the property of high resistame alone, by no means maxes it suitable for this purpose. In addition, it murt have a sufficiently high melting point ana it must have durability. ‘ithout thie last property, the ade metal is practically useless for furnace use. Given qual- ity, the question of cost must of course be given considera- tion. The best commercial resistors for this use are some ef the nickel end chromium alleys invented and developed by Mr. A. lL. Marsh and known as "Marsh Alleys". They are controlled by the Heskins Manufacturing Co. and the General Electric Oe. and can only be obtained in limited quantities for experi- mental purposes in non-permanent devices, and not for resale All ether manufacturers meking these alleys mst be liscenced under these patents, as well as appliance manufacturers wing these alloys in their applianses. The resister material we used is Hoskins’ Chromal A wire. This alloy contains approximately 60% nickel ané 20% ohromium, has an electrical resistance ten times that of platinum and sixty times that of copper; a melting point of 1285° C. and shows remarkable resistance to oxidation at high temperatures, which gives it a leng life. Its cost is a very gmall fraction of tit of platinum. For the heating chamber iteelf, an Alundum muffle with inside dimensions of 3, "x 5i"x 12" was purchased. This is made largely of aluminum oxide, which is preferable to car- borundédum, which is a carbide of silicon. Carborundum, which is high in carbon and silicon, has a bad effect on nickel- eohromium alloys, oorreding them severely. Fire clay brick was used outside the element itself. This material is very cheap, will easily stand the required temperatures, eané has a lewer thermal conduetivity than either chrome or magnesia brick. Outside of this, ordinary asbestos pipe covering was used. This is the regular 85% magnesia covering ani has geod heat insulating qualiti es. STUDY OF LOOAL CONDITIONS. The cellege supply of electricity is distributed at 220 volte direct current. This at onee removes any possibility ef using transformers and getting lower veltages and higher amperages, Which are soe desirable in furnace werk. the furnace will be used fer laberatorzry werk and must be rugged eneugh to stand considerable misuse. Every part mast be simple enough and made of lecal materials ao that it gan easily be renewed or replaced when wom out, The wiring comections must be “feol proof", i.e., sueh that no pessi- ble combination of switches will injure the furnace er over- lead the incoming feed lines. The furnace mst be economi- eal, Gisipating as little energy as pessible frem contrelling rheestats, Charts must be previded showing the heat curves ef the furnace, and the constant temperatures that can be ob- tained from any given amperage input. CONSTRUCTION DBTAILS The actual construction of the furnace was seriously de layed until rather late in the term while waiting for the we heating chamber er muffle and the resistance wire te arrive. 4s seen as the muffle arrived, work was started in melding a set of fire clay bricks, as shown in the drawings. The fire Clay was mixed with water and then tamped inte wooden melds ef the preper sise. These, as scon as partly dry, were burned to harden them sufficiently to stand handling. The skeleton frame of the furnace was built ef angle end strap iren welded together at the corners, giving a very strong ana stiff construction. The top, bottom and sides are covered with a leng strip ef galvanised iron belted with $/16" stove belts en each edge and at beth the beck and front te the fren skeleton frame. The front and back are made of panels of asbestes mill beard 3/8" thiek, drilled and out te proper sise as show in the photoes and drawings, With the back of the furnace in place, and the whole tipped om end, a 2$" layer of wet asbestes pipe cevering was packed in. The back brick was then plaeed in and on it the muffle and the four side bricks placed. Woeden spacers were placed between the bricks ané mffle so as te relieve a Winding spaee fer the heating element. With this preperly spaced, the 2+" left all arcund between the outer metal case end the inner part were tightly packed with the wet asbestos, A heating ¢011 of chromo was placed inside the muffle and run at about 1000° F. for three days and nights to dry it out and harden it an mch as possible. Doors of various types were planned and it was finally €eei ded to use a hinged type operating down. Such a doer when open, would serve as a shelf on which te place anything being placed in er taken out of the fumase. The tendency to slam it is also less than with one hinge at the top and as counterbalanced, it stays tightly shut when up. The counterbalance arm acts as a cool handle with which to epen and close the door. The only possible objection to this type of door is that it cannet be partly epened without exposing the entire opening to the air, The hinges used are ordinary strap hinges. Legs were made of four pieces of angle iron securely belted te each corner. Having decided upon a wire resistor fumace using Chrome "A" wire, the selection of the sise of wire depended upon several items. By preliminary experiments it was found that from 10 to 12 amperes were required to bring a furnace of this size up to high heat in a reasonable length of time, but after reaching this heat it could be held with consider- | able less current. For the lowest heat values about one- half an ampere is required. To compute a sise of wire and length needed to draw any given current is easy but there are limiting factors that must be met. If designed to draw 10¢ amperes a large and expensive rheostat would be needed to reduce the current to # an ampere, Hot only this, but at lew heats the major part of the power consumed would be wasted in the rheostat, - 7. In order te remedy this, a system ef three parallel circuits was decided upon whieh aould be operated in par- allel fer high heats and in series for lew heats. Te have used mere than three oirouite would have made the switching and constrotion toe complex without amy cerrespending in- crease in flexibility, A smell rheostat placed in one ef these circuits gives the necessary contrel,. The furnace was designed for a minimum amperage of 1+ witheut a rheostat. The sise ef wire used was a cempre- mise between several factors. Using #10 wire would have required 2700 feet te get the required resistanee, while #20 wire, the sise used, required £70 feet, while #230 wire would have required 27 feet. Any wire less than #22 would not have had the required carrying capacity er a sufficient Giameter te resist oxidation fer very leng. The use of the larger sises would have been mere costly as well as aiffieult te place on the muffle because of their length. A life test of small tube furnace using #20 wire, by Eimer and Amend, gave 1576 hours at 1900° PF, It was decided to use 270 feet of #20 B.& 8. gauge Chrome] "A" for the heating element, To wind all this aire on a twelve inch muffle in the ordinary way was impessible, eo the wire was would in a long helical coil using a 1/8" steel red as an arber. This helix was them womd around the muffle making twenty-four complete tums each with a very reasonable spacing. Taps were taken at each end and at twe peints in the center go as to include about 90 feet ef wire in each section. A geed grade of se called asbestos eevered copper wire was used te lead from these four con- nection points to the outside of the furnace. The muffle wound with the heating coil could, wth a little care, be slipped in and eut ef the fire clay chamber so that the condition ef the wire and leads could be deter- mined after each test. The entire fumace is so made that it can easily be taken apart whenever necessary to do so. The frent of asbes. tes millbeard with the door on it can be removed by un- screwing four belts. These come out ef holes tapped in the frame so there are no nuts to lese or held in place when either removing er replacing them. The leads from the wind- ing are attached te the wires from the switch board in a roomy outlet box, net attached to the front, with "Ne tereh" connecters. The muffle with its entire winding can be slipped out er in and splices made if the winding has been burned out at any place. The winding for the same reason is not cemented to the mffle. The follewing material was used in the constructien of the furnace proper: | 80 lbs. Fire Clay 40 " Asbestes Pipe Cevering 7.6 ft. Angle Iron 1"x 1/8" 12 ft. Strap Iron 1"x 1/8" 7.8 sq.ft. Galvanized Iron 5%’x 17" £2 pieces Asbestos Millboard 16"x 17" 5 dos. Stove bolts 1"x 3/16" 14+" Machine Bolts 1"x +" 13" Iren Kod 3/6" - | 1 - Iron Counterweight 1"x 2"x 8-7/8" 1 Set Screw 4"x + 1 Pipe Strap 3/8"- =9a 8 Washers +" a Alundum Muffle 54"x 3}"x 12" inside 270 ft. Chrome] "A" Wire #20 B.&S8&. afnre® 1 Fire Briok Door 64"x 4" S des. Porcelain tubes 1"x 37" 1 pair Strap hinges i Outlet box 3" £ Conduit fittings +" 30" Conduit +" ~ 4 switch board was made and arranged so that either one, twe, or the entire three of the sections could be placd in series or in parallel, depending en the ampere input needed. A fused line switeh was provided, the sctual switch. ing being dene with smaller switches. An anmeter shows the amount ef current being used at any time. PRELIMINARY TESTS, | Fhe furnace cempleted as abeve Was placed in the Mettalurgical labeoratery and teated for several days to de- termine and remedy any pessible weak spots and to determine its heating characteristics.