THESIS YX 102 oY tah om vs « The Design of a Semi-steel Puddling Furnace A Thesis Submitted to The Faculty of MICHIGAN AGRICULTURAL COLLEGE By R. G. Bigelow Candidate for the Degree of Bachelor of Science June, 1916. THESIS PREFACE. In this THESIS the autnor has endeavored to explain the term, "“Semi-steel", to explain its uses, to give the results of tests on this particular material, and to de- Sign an oOil-burning furnace of the mecnanical puddling type, such as can be used for the manufacture of semi- steel. In the design of tne furnace, parts of trade size nave peen used wher-ever possible in order to make tne construction less laboricus. While some of the parts were designed by the application of tneoretical formulae, Many were chosen from sizes sanctioned by current practice, The autnor wisnes to acknowledge his indeptedness to Mr. George K. Elliott of Cincinnati, Onio, for valuable information on semi-steel, to Crane Company of Cnicago, I1l., for valuable information and data, and to Edwin G. Greenman of Cincinnati, Ohio, for the use of blue prints, 94065 DIVISIONS OF CONTENTS. PART ONE. Part One deals with seml-steel entirely, giving a detailed account of wnat it is, how it is made, purposes for which it is used, tne place it is gain- ing in the engineering world and a table of results from tests on the metal. PART TWO, Part Two d@éals entirely witn tne design of an oil- burning furnace for tne manufacture of semi-steel. PART THREE. Part Tnree consists entirely of the Blue prints made from the drawings based on the design. This part will be found in the pocket on the inside of the back cover, PART ONE. In the first place, so-called “Semi-steel" structurally and actually is nothing more than gray cast iron, the ordinary forms of which are very familiar to us all. However, it is a cast iron naving unusually close or fine grain and great strength. Ordinary grades of cast iron rarely nave a tensile strength of more tnan 25,000 pounds to the square inch, while real semi-steel snould not be weaker than30,000 pounds to tne square inch and may run as high or even nigner than 40,000 pounds to the square inch. Semi- steel partakes of none of the cnaracteristic qualities of steel and it is entirely lacking in ductility. It must be thought of as cast iron and not as steel. It is made on the plausible theory that a very impure form of iron (pig iron) diluted witn a compara- tively pure form of iron (steel) will give an intermea late compound. The process is in most cases carried Out in tne common foundry cupola furnace, scrap ends of low carbon steel veing mixed with the cnarges of pig iron, The amount of steel varies from 5% up to ten times that amount according to tne size and nature of tne castings to be made. Tnere are several opvjections to this method: the additional amount of coke necessary to bring about the fusion and absorption of the steel makes the sulphur content of the semi-steel ratner too hign for tne vest strengtns; the presence of the coke in contact with the metal seriously reduces the possi- bility of attaining a much lower content of carbon which is one of tne cnief results claimed for the process, The truth is that the great majority of tnis semi-steel is no better than ordinary cast iron and not nearly so good as some of the better grades of cast iron which are made by intelligently blending selected grades of pig iron. In contrast with the semi-steel wnicn has been des- cribed and which is representative of a great percentage of the semi-steels on the market, is the Puddled Semi- steel made for use in valves to be used on superneated Steam and saturated steam at nhign pressures. Tnis pud- dled semi-steel is not made in a cupola but ina specially designed oil-burning furnace of tne general type of the air furnace. By this metnod as practiced by the larger companies using semi-steel, no steel is added to tne charge but the pig iron is purified by burning out a portion of the carbon and silicon while molten in the furnace. In other words, the effect is the same as though part of the pig iron were reduced to the purity of steel during the process. The advantages of this process are that no sulphur is taken up by the metal; that the carbon and silicon can be reduced with accur- acy and to a desired figure. Semi-steel made by this process has an average tensile strength of about 35,000 pounds to the square inch, and it is unexcelled for handling nigh pressure and superneated steam. Warping, distortion, and cracking accompanied by disintegration of the metal, which nave been obdserved so often in cast iron and ordinary cupola-made semi-steel valves and fit- tings wnen exposed to high temperature steam, are en- tirely unknown to this semi-steel known as puddled semi- steel. It is a known fact that many tneoretical metallur- gists deny tnat there 1s any such thing as semi-steel, prand it as a misnomer, and maintain that it is only cast iron. The last part of their contention is true, but the name, “semi-steel¥ is very firmly entrenched in modern iron foundry lore, in mechanical engineering and in steam engineering, and by it is meant extraordinartly gocd cast iron. That semi-steel is of great value tc engineering in its different brancnes i8 proven by tne fact that two of tne largest valve and fitting manufacturers in America have gone deeply into its manufacture and use and are making all their representations for the arti- cle prove out. TOPE. Z00°R, 450°F. 30660 31450 32€5C 33970 31090 31280 35240 34750 32850 o_o 32692 600°F,. 750°F. 873°F. Results cf Tests cn Semi-steel. 900°F., LOOO°F, Tensile Strength in pounds per square inch. 30940 30530 31640 36400 34450 35450 33600 32440 33720 32200 32980 30620 33780 32710 35880 33600 31530 29980 30960 33960 335490 35310 34800 Average: 33110 32860 2/7200 24350 20530 26420 28190 26540 29770 28 380 28740 28280 27920 25450 335420 25780 -—— eee - 27310 PART TWO. Tne Center cf Gravity of the Cross Secticn of Drum. In order to have the drum revclve easily, the sup- porting shafts are placed approximately at the center of gravity of the cross section. Following is a sketch of tne section and the computaticns for finding the pcsition of the center cf gravity with respect tc tne axis, XX, through the center cof the upper arc. It might be noted nere that all angles used in the computaticns were care- fully measured with a protractor. OA l2 x Cos 52° 45’ = 12 X .605 = 7.26" AB = 12 x Sin 52°H5° l2 x .796 = 9.55" Area Of Sector CEBDC <= 2 18 x 3.1416 x 64.25 — 360 161.5 sq.in. Area Of triangle CEB = 9.55 X (8 t7.26) we 145.7 sq.in. Area of segment CBDC s 35.8 sq.ir. 161.5 ” 145.7 Area of sector CGBO ° - Tee x 3.1416 x 254 = 320 84.in. Sketch used in finding the center of gravity of the section and the length of the Turnace | . io w ee _ .? 2a . - ae ams a, ~ elon Ee: TPS oe Area of triangle COB = 9.55 X 7.26 = 69.3 8q.in. Center cf Gravity cf segment CEDC = 3 a 3 C- | 19.107 es 16.25" from E. loa = 12 x 35.8” ? Center cf gravity cf triangle COB = £,20°< 2.42" from A. Center cf gravity cf sector COBFC-s 2xX1.91 x 12 x 360 . . 3 x 24 x 3.116 x 10505 7 0°92 " From 0. Center of gravity cf sector COB e« 105.5 X or 08 (144 x 3.1416) x 12 - (144 x 3.4 b6 x 360 “326 14.86" from F, Center of gravity of section witn respect to XX thnrougn O. Y - 320 X 2.86 - 69.3 X 4.84 - 35.8 x 8.25 420 - 69.3 ~ 455.8 z 1.865" from QO, Calculations for inside length cf drum in order to held 1000 pounds of iren. To begin with 5" is allcwed as the distance from the gravity axis to tne surface of the metal. Some areas used are taken from above computations. Area of sector HOIFH = To" x 3.1416 x “35 = 188 8q.in. KI = 12° - 3.137 © 11.3" ” area Of triangle HOI : 3.13 X 11.3 = 35.4% 8q.1n. Area Cf sector COBFC; Lo" x 3.1416 x S257 = 132.5 s6q.ir. Area of porticn HKIDH: 188 = 35.4% - (132.5 - 35.8 - 69.3) = 125.2 &q.in. Volume of 100C lvts of cast iron; 2000 x2 28 = 3840 cu.in. Inside length: 3840 _ = ” 125.2 29/9 In order to accomcaate the brick size, we will use 32¢" as the inside lengtn. Cailculaticns for the weignt of the assembled drun. All dimensions are taken from the drawings. The diameter is taxen as 42", which is a fair average, and the section treated as a circle rather than a combination cf two arcs. Calculations for the welignt of the hneada: Approximate weight of the outer rim: _.32 X 3,1416 X @.75 X 8.75 X 450 = 63 ycunds,. 1728 Arproximate weignt of the face inside of tne rin; —2 31 x 0. [8a4 x 1.25 x 450 - 246 Lounas. Avy rOxinmate welgnt cf the spokes: 6 x (1.5 - 3.5) x 0.625 x 12 x 450 = 39 pounas. 2 x 1728 Tctal apvroximate weight of one end cr tne drum: 348 pounds. Total approximate weight cf both ends of the drum: 696 pounds. Approximate weight of the snell; 33 xX 3.1416 x 0.3125 x 46 x 490 1728 = 423 pounds. Arrroximate weignt of brick lining: 2 —2 (33° - 24 ) xO_ 16° X 5205 X 150 2 i135 pounds. Arrroximate weignt of end bricks: —-2 2 x x OQ. ux 4 xX 150 - 43 he 29 =- 594 pounds. Approximate welgnt of empty drum; 43195 pounds. Weignt of maximum charge... ..... 1000 pounds. Total weignt.........6. veeeee eee 4195 pounds. Tne design or the supporting shaft on the control ena. This shaft being a sort of a cantilever beam is designed by the use of the formula, M = SI/C, where M is the moment, S, the allowable unit stress and I/C is the section modulus. 1/C for a circular section = 3.1416 x py? - 0.098 Dp? 32 Tnen M= S x 0.098 x Dp? For 8 we will take 2000 pounds per square incn, which will be well on tne safe side. Taking the weight as computed anc allowlng 6" as the distance from the center of the besrirg roller to tne end of the head, we nave,- 4200 x 8 = 2000 x 0.098 x D° Fron which we have, Dy =z 64.2, and D = 4", approximately. Tne design of tne nollow supporting shaft for the purn2>r end. Here we use tne same formula as for the aolia shaft, but I/C has a different value, 4 Then, M- SI/C=S x 4:1 118 (He 5a) where Rk, is the outside radius and Ro 1s tie inside radius. The irside radius is 4" as determired vy the Burner. Sub- stituting values for M, S and Ro, we have,- 4% _y 4200 x 6 . 2000 x 3.1416 (= - i) m 2 (+ 2 R} 4h 4h or, Rh - 2 —L_— = 4200 x 6 x 4 ® 2.025. Ry 2°x 2000 X 3.1416 > Here we have to assume scme value fcr R, and try it in the ejuation to see if the equaticnwill balance. First, we will take R, z= 2.75", Subdstituting and solving, we nave 212-48 = 8.025, which is nct true because R, nas been chosen too large. We will row take R, = 2.25" Tren, east 55° = 8.025, which is not true again, fcr Ry has ceen cnosen tco small. We now know that the exact velue lies between 2,25" ara 2.75". To be on the safe siae, we will use 2.75" as R, or 5.5" as D. The design of the operating gears. By planning, it was found that for recking the drun, three gears, one 8" and two 4" could well be used. A diametral of 4 was assumed to start with. Tne number of teeth, N= Dx p’, wnere D is the pitch diameter and p’ is the diametral pitch. Subdstituting and solvirg, we have for tne large gear N= @xX 42 32 teetn, and for each of the small fears, we have Ne +x 4 ze 16 teeth. » The outside diameter, OD = (N+2) pb’ Then for the large gear, OD = 228 = 8,5" and for each of the small gears, oD = 16t2 2 y 5 Tne Roilers. The rcllers on wniehn the drum revolves are chosen of such size as to make the drum revolve as easily as possible. These rollers are designed to be placed in positicn or. the shaft with a press fit, as shcwn by the dimensions on the drawings. The Snafts. The shafts used in this design are all cnosen from Standard sizes. This will save much machine work in the . construction of the furnace. As noted on the drawings, all snafts are toc pe of cold-rolled steel. The Legs. Tne legs of this furnace are cesigned sc as to be Simple in construction and very rigla. They are of tne wed type which increases the strength for the amount of metal used. The webs are of sufficient thickness so that, should there be a small flaw in the casting, the strength of the leg will nct ve impaired. Tre Control Wheel. The control wneel is made of cast iron and steel as indicated on the drawing. The steel spokes are cast in Place which insures a rigid constructicn and simplicity. The wneel is made of such size. that the task of rocking the furnace will not ve nard. The Burner. Tne burner is not designed, vut the furnace was designed to use a burner made by the Gilvert & Barker Manufacturing Company of Springfield, Mass. trnis firm Makes a specialty of oil burners for this and similar types cf work, 928 ) Ww e—— > Ls —} —S== — —_—— ==>, — i | viii il ” w ec < « a a os e ” ec Ww 2 z = w me < —=nN Uy