The Iron Age 1921-09-22: Vol 108 Iss 12

THE IRON AGE New York, September 22, 1921 BLISHED 1855 VOL. 108: No. 12 Open-Hearth Furnace Design Reversing Valves and Their Effect on Efficiency—Chim- ney Areas and Heights —_—_—__— BY A.. D. WILLIAMS* —————_- naces have been a source of much trouble. Many different valves have been designed and placed the market, and a number of different flue ar- R EVERSING valves for regeneratively fired fur- — ws ————— EE don deanna tt ed 1 and 2—Pyrometer Diagrams of Stack Tem- ires Abnormal conditions showing sudden vard kick in Fig. 2 are due probably to air leak- d the combustion of checker gases passing along the flue toward the stack gements have been devised, to eliminate the revers- e and accomplish the reversal with a multipicity of mushroom valves and dampers. The simple Siemens butterfly valve was the first four-way valve used on engineer, Newark, N. J Copyrighted by h jumber of Furnaces of Each Size J 3224 a Me a Rte d cal | i -—o6 | 92+ - tt > ¢2 o. + * } > Pet ery yy Pr > i © } \ | Pere err } ToT $—F-— + 4+-—++ edit errr a ee eS es 1+} itt) = 1 TEP Te TTT tity _ 0 30 40 50 60 10 80 90 100 200 lons, Nominal Capacity of Furnace Fig. 3—Gas Valve Areas Spotted these furnaces. In use, however, exposed to the hot gases on one side, and the cool gas or air passing to the regenerators, it soon warps and becomes leaky. These leaks permit air to pass direct to the stack and cool down the waste gases, or if used for gas there is continual leakage of gas, which either burns in the valve or in the stack flue. While many reversing valves are water sealed, most of these, during the operation of the valve, in common with the butterfly valve, open a direct connection, prac- tically the full area of the valve from gas main or air to the stack. A few valves have been designed which Number of Furnaces of Each Size 6 3224 Re a " t + } }4} 5 13} | f oi+-* #2- o~—9 12} \ © IN} oe ic , on >_> +4 @ 10 oe +4 be aS, 3s 9 +9 Bnaidhml , ; ; 4 =." s > + . wD = 3 sa 8} ei § ? ws Th $ * * 4 a | of . > 2 } 3 a 5 + ‘ee > } = e 4} +? r ; . 3} j a Gan Gon oS ee ee ee eee ae ee ae a eee ae 2} + +-4 } A clhcaelbediiead 1} $44 4 t—$—-t-+$-++44 Quititi_t Hereinceg ipsa 4 iii iadia dae 0 0 20 80 40 50 60 10 80 90 100 200 fons, Nominal Capacity of Furnace Fig. 4—Air Valve Areas Spotted cut the furnace, gas and air entirely off from each other and the stack, but these valves have not come into extended use. Water sealed valves are used extensively. As long as the seal holds they are tight, but there are usually structural limitations to the depth of the seal. When exposed to gas pressure on one side and the stack de- pression upon the other the seal is unbalanced, and may readily be broken by surges or explosions. In some cases a‘considerable water area is exposed to the en- tering gas or air as well as to the waste gases. Producer gas and stack gases, being several hun- dred degrees hotter than boiling water, will absorb a 719 ewe é igh 2 Ta ee ee 720 THE considerable amount of moisture from a very small area of water surface. With all of these valves the sealing lip must be raised, to clear the water surface and the port rims, whenever the valve is operated. From the time the seal is broken until it is re-established the full suction of the stack acts to pull air or gas, and in some cases both, into the stack. TABLE 1 Reversing Valves NominakbDiameter Valve —Nominal Area of Valve In. Ft Mm. Sq. In. Sq. Ft Sq. Meters 18 1.50 $50 254 Rae 0.164 21 1.75 525 346 2.40 0.223 2 2.00 610 452 3.14 0.292 27 2.25 685 572 3.98 0.369 30 2.50 760 706 1.91 0.406 33 2.75 $40 S55 5.93 0.554 s6 3.00 915 1017 7.07 0.657 40 3.33 1000 1256 8.73 0.808 2 3.50 1070 1385 9.62 0.894 48 4.00 1220 1809 12.57 1.168 54 4.50 1370 »9G() 15.90 1.477 60 5.00 1525 2827 19.63 1.823 66 5.50 1675 3421 23.76 2.210 a2 6.00 1830 4071 28.27 2.630 ans tote With some valves the furnace itself is directly con- nected to the stack through both regenerators at re- versal, so that a portion of the stack pull tends to draw air in at the doors. This may or may not be seriously objectionable, according to the distance between the valve and the furnace, and the rapidity with which the valve may be operated. Effect of Absorbed Moisture Moisture absorbed from water seals is a direct loss of the amount of heat required to evaporate it and superheat it to the temperature at which it passes out Number of Furnaces of Each Size an , tte gs 2 4 3 1 | T 1 80 ° 08 + 10 + i - ° 4 60} Pe, e *3 of > * + 140 * | ¢ c t . 0 | + 5 < “- 5 x 3 | 4 +> . j 40 | 4 | | | { 80} | | | 4 —— 7. ag A a aC a 0 0 “20730 40 50 60 10 60 90 100 190 200 C Nominal Capac ty of Furnace Fig. 5—Stack Heights Spotted of the regenerator to the stack. In addition, its disso- ciation probably occurs in the checkers, which may release some oxygen to combine with other combus- tibles at this point. The reactions here are complex, as certain hydrocarbons dissociate in the checkers, as well as CO, A further increment of moisture occurs in certain elements of the charge, and an open- hearth furnace is not particularly efficient as a dryer. At the same time, moisture is carried in by the air supply. All of this water leaves the regenerator for the stack as highly superheated steam, and its amount is considerable, particularly with large furnaces. In “The Heat Balance of the Open-Hearth” by Sidney Cornell, (Chemical and Metallurgical Engineering, May, 1913,) the weight of moisture passing in the flue gases was about 26 per cent of the weigh* of ingots produced. The water seals in the valves increased the IRON AGE September 22, 192] amount of moisture in the producer gas 1 per cent As the water seal depends upon its water supply a very slight stoppage breaks the seal. Frozen water Water vapor absorbed by the flue gases reduces ‘ Number of Furnaces of Each Size $3933 BS 1 | 4 ns rir. tT a Tt hr hehe ne TITTTTT ES | } oF + + + -4-—4+—+4+—+ t ++ A 50 a 4 4 Jt d, + + 4 cage dni ‘it | Ree ry wy + a beard sr + + ae 4 $ ne. z }_} cy D 30 . < Be: x 5 r mH 20 10 0 20 30 40 50 60 10 80 90 100 150 200 é Tons, Nominal Capacity of Furnace BY Fig. 6—Stack Areas Spotted ee. oe ° ° . ° ci lines in winter frequently cause the superintendent 3 to consign the plant to the tropical regions, particu- larly when they occur on a cold Sunday morning. TABLE 2 5 Reversing Valves and Chimneys in Practice cS Nominal ———Chimney - 4 Furnace Valve Area Area of Sq. Ft Capacity, Gas, Air, Sq. Ft. per Ton Height, Bore, per Tons Sq. Ft. Sq. Ft. Gas Air Ft. Sq. Ft. Tor 10—1 3.12 4.87 0.3812 0.487 100 9.62 0.96 15 3.12 3.12 0.208 0.208 90 12.56 0.837 20—A 3.90 3.90 0.195 0.195 100 si Gin ? 0) 3 3.07 1.00 0.154 0.200 wise 20 D 3.90 5.90 0.195 0,295 - j 20—4 ; 5.85 0.293 25—4 1.90 4.90 0.196 0.196 125 12.56 0.503 25—3 5.65 6.90 0.226 0.276 125 19.64 0.7S8¢ 30—5 1.90 1.90 0.163 0.163 125 14.2 1.47 30 { 4.90 4.90 0.163 0.163 125 14.2 0.47 3 2 5.90 5.90 0.169 0.169 er ane 35—A 1.90 5.90 0.140 0.169 114 17.0 + Shot 1.90 4.90 0.140 0.140 130 15.9 474 % 35—~F) 1.90 7.10 0.140 0.203 4 10 7 ee 9.0 0.22 40—4 140 19.64 0.49 50—4 7.06 7.06 0.141 0.141 150 23.7 0.4 ‘q 50—10 7.06 15.00 0.141 0.300 aie 21.5 t 50—11 5.94 9.18 0.119 0.184 150 26.27 50—R8 9.60 12.50 0.192 0.250 150 28.27 t 50—13-15 9.60 9.60 0.192 0.192 150 25.8 l 50—16 9.60 12.50 0.192 0.250 160 20.0 { i 50—X 5.22 7.75 0.104 0.155 BY »0—4 8.13 8.13 0.163 0.163 7 > 50—19 9 60 9.60 0.192 0.192 150 20.0 1 50—5 7.00 9.52 0.140 0.190 eine 50—D 5.90 8.70 0.118 0.174 60—13 7.06 12.58 0.118 0.21( 180 33.18 0 60—3-5 7.06 ave 0.118 0.145 153 28.27 75 1 9.42 13.00 0.126 0.173 75 A 9.42 12.56 0.126 0.167 75 3 8.70 12.56 0.116 0.167 75 IY 7.10 10.50 0.095 0.140 adie er 75 2 ; 165 20.36 80 1 15.8 100—3 ai a bas sal ward 19.6 100—H 10.00 10.00 0.100 0.100 180 28.25 100—E 8.75 8.75 0.088 0.088 180 50.0 ( 100 $ 9.42 15.90 0.094 0.159 160 28.25 5 100—D 8.70 12.60 0.087 0.126 ; 150—1 ee ot es er Saad 150 23.7 5 200—4 7.00 12.56 0.035 0.063 180 38.50 ' not only their temperature, but also the stack draft and the amount of heat available for the waste-heat boiler to convert into steam. As to the valve area required for a furnace, is as much diversity in practice, as there is with Figure 3 isa graph- other proportions of the furnace. ical comparison of reversing valve practice, while T there the Table 2 shows the same data in tabular form. The reference numbers in the table correspond with those given in September 22, 1921 weding articles, which have appeared in Révue Vétallurgie in 1919 and THE IRON AGE in 1920.* Some of these valves are so heavy as to require eetric motors or some other form of power for their eration. In these cases the control is located on a ilpit at a central location on the charging floor in : rear of the furnace. Smaller and lighter valves ae operated by cables or levers led into the pulpit. Heavy valves have considerable inertia, and for this son cannot be operated as rapidly, even by power, ea .s the smaller valves which have lighter moving parts. Na Furnaces fired by pulverized coal, natural gas, oil, -e oven gas or tar require reversing valves for the - \ir only, the fuel being reversed by shutting off the ey t at one end of the furnace and turning on the jet Ee he opposite end. Some of these furnaces are sup- ; with one checker chamber at each end, while B. ers are so constructed, with two chambers at each e |, that they may be converted with little difficulty to roducer gas firing. Those furnaces in which the air only is preheated ' a slight advantage over those in which both the and air are preheated, in that no unburnt gas has e wasted up the stack at reversal. The amount iel lost in this manner depends upon the gas-filled ime between the reversing valve and the port, and frequency of reversal. When the gas is preheated loss cannot be avoided. This gas likewise creates an explosion hazard When the conditions are right this gas burns and passes ) the stack as a puff of flame. Again, if it becomes xed with air, the mixture being below the ignition perature, and a portion of this mixture is suddenly ted, an explosion of more or less violence occurs. [hese explosions damage the walls of regenerative hambers, flues and waste-heat boilers, opening up acks, so that large amounts of air are drawn into the system, reducing the stack draft and the output of the boiler, by reason of the lowering of the tem- perature of the waste gases. Velocity of Gases Through Valves One factor in regard to valve area that meets with le consideration is the velocity of the gases pass- ¢ through the valve. In addition, most valves involve Y hange in direction of flow, totaling 360 deg., 180 deg. c the valve and two 90-deg. changes in the flues. re When a stream of flowing gases goes through a pas- involving changes in area, velocity changes are , nvolved. The velocity of flow in the normal section of the flue may be represented by Vmin., and in the con- icted area of the valve by Vmaz. The corresponding ty heads will be hmin. and hmer. That is, a veloc- ead = hmar.—hmin. will be required to produce the ise in velocity. The pressure in inches of water lired to produce the increase in velocity will be: a? vt ive p 0.192 (hmaz — Rmin) de at weight in Ib. per cu. ft. of the gas in motion be temperature ¢ deg. : rhe coefficient of contraction varies from unity, E en the areas of the two passages are same, to 0.83 n the area of the smaller passage is 0.01 of the f the larger passage. This slightly increases sure required, but ordinarily a large margin ible to cover this. The loss of pressure due to ¢ s of direction may be expressed by the formula: td p = 0.192r dt — 29 pressure in inches of water column; velocity in ft. per see. of the stream of gas; sravitational constant — 32.2; veight in lb. per cu. ft. of the gas in motion at the temperature ¢t deg. ; function of the angle through which the stream s deflected, Owing values are given by Weisbach for short I es -Odeg. 40 deg. 45 deg. 60 deg. 80 deg. 90 deg. 0.046 0.139 0.188 0.364 0.740 0.984 i S last loss, due to directional change in passing — the valve, amounts to 180 deg. in the valve “set and 90 deg. in the flues each side of the valve. tnr Fi page 35; Jan. 8, page 119; Jan. 29, page 317; I : March 18, page 805; April 29, page 1225; 0; Aug. 5, page 319. THE IRON AGE 721 For this reason changes of diameter of valve of 6 in. or so are of comparatively small effect upon operation. These losses vary with the square of the velocity in the valve and the flues. Many furnaces are, undoubt- edly, choked by the small area of the valves used, and much operating trouble is doubtless due to lack of con- sideration of these details. A valve small enough to choke the furnace is an expensive luxury, as it exacts its toll every minute the furnace is working. Added chimney height or forced draft must be provided to overcome its resist- ance. There is very little doubt that the erratic work- ing of some furnaces with different weather conditions is largely due to the close balance, between draft and resistance, being disturbed by barometric and temper- ature conditions. These various losses will be gone into in more detail in the design computations. Figs. 5 and 6 give a graphical comparison of chim- ney areas and heights, which are tabulated in Table 2. Open-hearth practice in the United States tends to steel self-supporting chimneys lined with fire brick. Even with waste-heat boilers and induced draft the straight chimney is used, while abroad many of the venturi cone (Prat type) chimneys are used. The chim- ney is required to carry the waste gases to a sufficient height to prevent their becoming a nuisance, not only in the works, but to its neighbors. Another factor which must be considered in con- nection with the draft required to operate the furnace is that the gases must be pulled out of the furnace through the ports, down through the checkers and through the valves and flues. That is, the waste gas end of the furnace is below atmospheric pressure, and there is a constant tendency for cold air to be pulled into the system. When the brickwork is new and tight the leakage of air into the system may be slight, but after a few explosions have shaken things up, this brickwork is liable to leak like a sieve. For this rea- son, gas samples taken at the base of the stack are liable to show a condition of affairs quite foreign to that which actually exists in the gases leaving the heating chamber or the regenerators. Many tests and considerable investigation of open- hearth furnaces have been made from time to time at different plants. Tests of this kind cannot be per- mitted to interfere with furnace operation. They must be carried on night and day over several melts. When it is decided to run such a test, the question of cost must be considered, not only in apparatus but in per- sonnel. A large number of observations and chemical analyses must be recorded and analyzed. Very slight details are liable to vitiate the value of such tests, and it is extremely difficult to impress upon the assistants available the factors that are really essential. Frequently, in fact, the busy executive is unable to devote to the test scheme the preliminary attention he desires, and as a result only a portion of the data required are obtained. Comparatively few of these test results are made known, hence there is much repetition of the work of others. Inquiries now before structural steel mills call for 280 tons of steel to be used for the initial unit of the factory-warehouse building to be erected at Youngs- town, Ohio, by the Paul J. Kalman Co., a Chicago fab- ricating interest. The plant will manufacture rein- forcing steel and when completed will have an annual consumptive capacity of 70,000 tons of bars. The American Rolling Mill Co., Middletown, Ohio, has made another cut of 10 per cent in salaries and wages. The new schedule was effective Sept 16. The base rate for common labor is now 27c. per hour. The announcement was made that this will be the last re- duction in wages to be made this year, regardless of the price of steel. Molders of the American Steel Products Co., Ma- comb, Ill., recently struck as a protest against a re- duced rate on piece work put into effect by the com- pany. The molders had been earning from $8 to $9 a day and in order to meet competition the company found it necessary to reduce this amount to $6. : F | tg TOOT Pt sen, af, 722 Electric Crane Safety Switch A new crane safety limit switch, designated as type L C, intended to prevent overtravel of the hoist brake and the consequent damage, has been brought out by the Westinghouse Electric & Mfg. Co., East Pittsburgh, Pa. The safety limit guards the equipment irrespec- tive of the kind of control or of the position of the con- troller handle or brakes. In addition to its safety fea- ture time is saved within the normal operating range of the hoist, as the switch obviates the necessity of ap- proaching the limit of travel slowly and cautiously. Under normal operating conditions, the limit switch is held in the operating position by the counterweight as shown in the illustration, which overpowers the tor- sion operating springs, also shown. In case of over- travel the counterweight is raised by the hoisting hook or other moving part, which permits the torsion spring to operate the switch. This disconnects the motor from the line and establishes a closed dynamic braking cir- In Case of Overtravel, the Counter- weight Is Raised by the Hoisting Hook Pormitting the Torsion Spring to Oper- ate the Switch cuit through the motor armature, motor series field and resistor, which stops the motor, irrespective of the position or type of the controller or brake. The opera- tion of the switch also releases the series magnetic brake which holds the load until the controller handle is moved to the lowering position. As soon as the hook has been lowered beyond the limit the switch is again ready to function as before, as it re-sets itself auto- matically. The quick-make and quick-break feature of the switch throws it rapidly from the normal to the braking position when approaching the limit and back to the normal operating position when backing out. It will not remain in a partially open or partially closed position and destroy the contacts or fail to function. The safety limits may be used also on metal mixers, bridges and other applications employing series motors, where it is desired to limit the travel in one direction by applying dynamic braking. Industrial training, by co-operation between school and shop, has been inaugurated by Pennsylvania State College. A supervisor in each shop gives individual aid to the students working there, on the points where they show weakness. All papers are corrected at the college, and all lessons are sent from there direct to the students. Special text books have been prepared for students in textiles, automobile parts, electrical equipment, iron and steel, machines and tools and other subjects. THE IRON AGE September 22, 1921 Industrial Cost Association At a meeting of the board of directors of the Indus- trial Cost Association, held on Sept. 8, the resignat of J. W. Stannard as president and director was ; cepted and Horace S. Peck, comptroller, S. K. F. Ind tries, Inc., 165 Broadway, New York, was elected to the vacancy. Christopher Haigh, General Electric ‘ West Lynn, Mass.; Joseph P. McLean, Pittsburgh Fo & Iron Co., Pittsburgh, and Ernest J. Wesson, W. 1 Raleigh Co., Freeport, Ill., were elected directors the association. It was decided to hold the fall conference at Pi burgh, Nov. 2, 3 and 4. The association now has 224 members. The organization has no salaried officers The board of directors at one time passed a resolut to the effect that at such time as the receipts of th: association exceeded disbursements, the secretary-treas urer was to be placed on a salary basis but at no tim has this condition existed for a long enough period | justify such action, and it was therefore impossib| this officer to continue to devote his entire time to association. In a letter to the membership President Peck ex- plained that for a number of months A. A. Alles, J the secretary-treasurer, through the courtesy of Fawcus Machine Co. and the Schaffer Engineering (o devoted his entire time and attention to the affairs of the association and that he still continues to devote as much of his time as possible to this work. New Plans to Study Drill Steel Breakages WASHINGTON, Sevt. 20.—Plans have been made by the Bureau of Standards and the Bureau of Mines to intensify further study of breakage and heat treatment of drill steel and invitations have been issued to out- side experts on the subject, who, it is hoped, will co- operate with the bureaus and act as a consulting com- mittee. Those asked to serve in this capacity are President J. A. Mathews of the Crucible Steel Co. of America; Dr. Van H. Manning of the American Pe- troleum Institute; B. F. Tillson of the New Jersey Zin Co.; W. L. Saunders, chairman Ingersoll-Rand | and F. W. Benton of the Copper Range Co. The Bureau of Mines has been conducting research on drill steel at its experimental stations in Rolla, Mo., and Minneapolis, and, having a lack of appropriations to continue them as extensively as necessary is seeking aid through experts in the trade and in the mining 1! dustry. The Bureau of Standards also has been co! ducting a research with the Bureau of Mines. It is stated that improvements in drill steel have not as yet come up to the standard of those made in the drill then selves and that there has been considerable breakage, causing a cessation in operations in drilling at mine: and resulting in heavy costs. Gain in Employment at Youngstown August wage distribution by Youngstown, industries of $3,349,974 compares with $3,323,982 dis- bursed in July. This is the first gain in payroll at Youngstown in several months. According to a stat ment issued by the United States Department of Lao", Youngstown leads other cities in the increase 1! ber of persons employed in industry in August as pared with July. Youngstown’s gain is 27.7 per whereas the average increase in 65 principal trial centers is 1.08 per cent. The Departmen! Labor bases its figures on reports from 1428 fim the various cities which usually employ 500 or persons. During August the Ford Motor Co. made 109,!' passenger cars and trucks in the United States, break: ing its former record for a month’s output >; 211 Total production of Fords in this country since i beginning of this year is 645,192, an average of 80,65! a month for each of the eight months. The average output for each of the 27 working days of August Ww approximately 4050. oe juan ae Electric Furnace Progress in 1921 Discussion of the Merits of Dual Voltages. for Melting the Steel and for Refining —Heat Losses and Electrodes Considered Vig ia ——BY E. T. MOORE* UCH of the work outlined as a program for re- Mf scares work of necessity involves continuity of 4% operation, a condition which has not been pos- to obtain. On this account it has been impossible nplete the work, which anticipated extensive re- h and the recording of a mass of secured data. ever, it is hoped much discussion will result from suggestions outlined herein, and a much broader given to the subject. The plans called for the ving lines: A thorough study of the heat losses from electric Presentation of operating data on furnace elec- , according to the electrode specifications drawn this committee and presented in a paper before ast convention. \n investigation of the merits of dual voltages, the deleterious effect on the metal, if any, of a tively high voltage, for melting. Investigation of electric furnace phemonena. Heat Losses From Electric Furnaces nsiderable attention has been given to the study losses through walls, roof, electrodes, cooling r, and escaping gases. (A bibliography of the ature on the subject is given in the report.) Messrs. Wolfe and Wysocki in “Heat Losses Through trodes of a 6-ton Heroult Furnace” give data for osses through the cooling water, but obviously of this kind are not complete, since a large amount is lost at the electrodes which is not carried vy cooling water. The figure of 18.7 per cent for power loss of the electrodes did not represent mplete loss at this point. This is especially true e furnaces where there is considerable clearance the roof and the electrodes, which permits ape of gases in large volume, carrying a tre- amount of heat with them. The economizer { to in our last report will prevent the escape irge amount of heat, and lengthen the life of trodes as well. e Stobie furnace, which was developed in Eng- 1 used quite extensively there, the makers claim which greatly reduces the heat losses at the . between roof and electrodes. This device con- f a series of telescoping tubes placed around the les above the roof, and quite effectively seals the ‘ against escaping gases and heat. The main- of tight joints between adjacent sections is which should be investigated. nsiderable amount of heat loss occurs through doors, which should be designed to give as ef- eal as possible. Electrode Specifications tial specification on any material generally complete, and the committee has realized ; first attempt to draw up an electrode speci- such as the one presented at our last conven- be far from satisfactory. hoped, after a year’s use of the suggested ‘n, to send out another questionnaire to manu- ‘, and record the results of such use in this (he industrial depression has caused the shut- n of most electric furnaces, and operating data new electrode specifications are therefore not It is hoped the specifications will be criti- ‘ eventually corrected and elaborated. engineer, Halcomb Steel Co., Syracuse, N. Y. s the report, substantially complete, submitted “s committee to the Association of Iron and al Engineers, at the Chicago convention. In the committee report before the last convention reference was made to “a well defined belief that the high rate of melting, obtained by high voltages, is de- trimental to the quality of the product, especially to tool steel.” A considerable amount of discussion by correspondence has taken place as to the relative merits of dual voltages, the matter dividing itself into three features: Effect of relatively high voltage during melt- ing, on the quality of steel. Effect of relatively high voltage during melting, on the refractories within the furnace. Effect of relatively low voltage during re- fining, on the refractories within the furnace. High Voltages During Melting There seems to be a general opinion that equal re- sults as to metal quality can be secured by either single or dual voltages. Some users feel that there is prac- tically no change in the quality of steel when melted and refined at one voltage anywhere between 100 and 140 volts, such as is used in the Heroult, Greaves- Etchells, Ludlum, and similar types of furnaces. How- ever, there seems to be a decided antipathy for volt- ages, say 160 to 250, as the extremely long are with such voltages is detrimental to the metal, particularly high grade tool steels, and also very severe upon the refractories, especially during refining. On the other hand, while it appears reasonable that detrimental effects to the metal are produced during a high voltage melting period, yet these effects can largely be eliminated by a relatively longer refining period. It is also evident that better steel is made by preventing contamination of the metal, than by al- lowing contamination and then attempting to eliminate the injurious elements. In other words, we return to the old adage “an ounce of prevention is better than a pound of cure,” and whether or not this is worth con- sidering, from the standpoint of commercial steel pro- duction, is more or less debatable. As to just how the steel is affected no one at pres- ent, it would seem, can answer, for this brings in such obscure phenomena as the effect of oxygen and nitro- gen on molten steel at high temperatures. In a mass of solid steel scrap, having a source of intense heat at its center, with considerable space for gases to play between the fragments of material, we have an ideal condition for the thorough mixing of oxygen and nit- rogen with solid, molten, and even vaporized steel. That such conditions permit of the most active chem- ical action between the gases and the steel needs no comment. The more intense the action is, the longer must be the period when these gases, oxides, nitrides and other combinations must be eliminated from the final product. This action is present in a greater or less degree in every arc furnace, or open-hearth fur- nace for that matter, and it might be pointed out now that this may be one of the reasons why the induction furnace is said to produce a higher quality of steel. Effect of a Long Arc Assuming, however, that nitrogen is injurious to steel, it would appear that nitrogen fixation would in- crease when a relatively longer arc with high voltage was used, than with a short arc when lower voltages are used. In other words, the principle of a steel] melt- ing furnace with such a long arc is very similar to the Berkelund and Eyde or Schonherr nitrogen fixation furnaces. The fact that the arc is long gives a much greater path of air for the arc to pass through, and it is reasonable to assume as a hypothesis that the greater air gap the are traverses the greater will be 723 724 THE IRON AGE the quantity of nitrogen or nitrous oxide produced. You will readily recall experiments made in the laboratory with an ordinary static machine, and how the length of the are was varied by increasing the air gap. You also probably noticed that the longer the arc the more perceptible the presence of ozone became, indicating a greater conversion of oxygen to ozone. While ozone is an allotropic form of oxygen, one and one-half times as dense—a much more powerful oxidizer than oxygen gas, and nitrogen a very inert gas combining only at high temperatures, and then only with a limited number of elements, it would ap- pear consistent to expect similar results with both as expressed in our hypothesis. The fact that nitrogen combines only at high temperatures is a fair indica- tion of its presence in steel, particularly when a long are is used, as with higher voltages. As a result of this supposition, it was thought de- sirable by the committee to investigate thoroughly the phenomena within an are furnace, by taking photo- graphs of the arcs with a multi-exposure high-speed camera, and simultaneously recording the phenomena of the circuit with an oscillograph. Also a close an- alysis of the steel, both physical and chemical, was to be made with particular reference to the occurrence of nitrogen in more or less degree, as the value of the melting voltage and consequently the length of the arms were increased. Microscopic examination of steel has revealed a fine irregular needle structure, which has generally been accepted to be typical of iron nitride. This struc- ture has been studied by a number of investigators, as outlined in the report. In an article in THE IRON AGE, volume 108* on “An Occurrence of Nitrogen in Steel,” A. A. Blue presents data showing evidences of direct combination of ordinary inert atmospheric nitrogen with iron at moderate heating temperatures. Presence of Nitrogen at High Temperature If evidence of nitrogen is found in iron at mod- erate heating temperatures (not over 1800 deg. Fahr.), it is reasonable to assume that at temperatures several times these amounts, such as exist in the electric fur- nace, a much greater combination will result. To our knowledge this type of investigation has never been made, although J. Kelleher, in his paper “Some Phe- nomena Observed in Electric Furnace Ares,” -before the American Electrochemical Society, projected images upon a screen through an aperture in the walls of a small single are de experimental furnace, from which sketches of the are could be made. These sketches showed that the flame of the are apparently flowed from the electrode to the slag, de- pressing the slag and flaring out at all sides when the electrode was the negative pole. An arc length of 3 in. was easily maintained, and under these conditions the arc was silent. With the electrode as the positive pole the arc was very unstable, starting below the sur- face of the slag, the flame moving away from the slag surface and projecting particles of slag into space with considerable force. The arc could barely be main- tained, at a greater length than 1 in., and was loud and spluttering. These observations were of course made in an ex- perimental furnace using direct current, and may have no bearing in a commercial furnace. However, the obscure information obtained certainly warrants fur- ther extended investigation and discussion. In carrying out this work, it was proposed to use a new type of portable oscillograph to record the wave form, building up of current, instantaneous current and voltage wave distortion, high frequency, har- monies and other phenomena. As a number of phe- nomena in an electric furnace circuit are over in 0.001 second, the ordinary moving picture camera, capable of from 15 to 35 exposures per second, is hopelessly slow, and it was proposed to use a polar multi-exposure high- speed camera capable of taking pictures at the rate of 3000 per second, thus being comparable to the oscillo- graph in recording the details of transient electrical phenomena. For some purposes, of course, such speed is not *July 7, 1921, page 1. September 22, 192) necessary, but for many of the extremely rapid change in the configuration of visible forms the finely defin: well distinguished, high-speed pictures Are necessa if data of any value and utilization are to be secur: The stereoscopic exposures obtainable, when prope: mounted, give, with unusual vividness, a view of what is occurring in the arc, as the flame or objéct appex to stand out in space, disclosing the shape in thre dimensions. The oscillograph and high-speed camera are sy) chronized so that, by pressure of a single button, the mechanism of both is operated so as to give simulta eous records. Effect on the Refractories Irrespective of the merits or demerits of high vo!t- age during melting on the quality of steel, the effect upon the refractories within the furnace is about the same in each case, for the reason that the charge is cold, and most of the heat generated goes into the cold metal. Also, at this stage of operation, there is no blanket of slag to reflect the heat of the ares in such manner as to cause damage to the refractories. Unquestionably the use of a relatively lower volt- age for refining is desirable, since the amount of power input can be much reduced. With a lower voltage and reduced power input a much shorter are can be main- tained, and this alone will give a prolonged life to the refractories of the roof, and to the walls as well. With a single voltage available, this must be high enough usually around 100 volts—to give suitable power input to the furnace for reasonably rapid melting, and yet for refining purposes this is too high, because excessive heat radiation from the longer arcs will greatly cur- tail the life of the roof. On the other hand, a voltage of 100 for melting, while entirely satisfactory, can reasonably be increased to at least 120 volts without injuring the refractories a bit more than with 100 volts, and in addition more rapid melting is secured, better regulation from the regulators, and the kwh. per ton input reduced. In some instances, a value of even 140 volts would be ad- vantageous, although above this point joint trouble is liable to develop in the electrodes. The values given above are based on 3-phase 6-ton Heroult furnaces. For refining, a voltage low enough to give sufficient power input into the furnace to maintain heat bal- ance is desirable, and yet not low enough to allow the electrodes to dip into the bath, or the metal and slag to be splashed against the electrodes, by the action of the ares, as contamination of the metal will result A value of 60 volts on a 3-phase 6-ton Heroult furnace installation, with 1500 kva. connected capacity, was found ideal from a power input, refractory, and elec- trical efficiency standpoint, but undesirable because the electrodes would occasionally drop momentarily into the bath, and thus increase the carbon content of the steel. A voltage of 65 was likewise found undesirable. At 70 volts, however, the are length is greater, and the vertical distance from electrodes to bath level suff- cient to give nearly as good electrical and refractory performance as at 60 volts, but with much greater safety from contamination due to dipping electrodes. Values of 75 and 80 volts are also desirable, but with a slightly decreasing electrical efficiency and refractory life, but greater safety from contamination. Individ- ual installations will require separate consideration, with the selection of a value between 70 and 80 volts, which will best meet the specific case. General Conclusions To recapitulate, it would appear that as good steel could be made using dual voltage as with one voltage, provided the melting voltage is not too high; and, in addition, better economy may be secured electrically and from refractories. This economy amounts to quit an amount, so that the use of dual voltages is well justified and, we believe, will pay handsome dividends. On Jan. 1, 1921, the total number of electric ste! furnaces in this country was°356, and we believe this number has not been increased during the first six months of this year. There has been very little a¢ tivity in the non-ferrous industry, although there has been a small increase in the number of brass furnaces: Three Types of Alloy Sheet Steel—III Electric and Acetylene Welding Details and Their Effect on Physical Properties —Chrome-Vanadium Steel Adopted BY HORACE C. KNERR® T\LECTRIC spot welds, although not used to trans the joint being a continuous weld 7% in. in length. 4 mit heavy stresses, play an important part in One set of sheets was welded with copper-plated 4 the manufacture of sheet steel fittings, often Norway iron welding rod. The other set was welded ised to hold parts in place for brazing and, in with the parent metal, strips about 1/16 to % in. wide ases, to replace rivets. Tests were therefore being sheared from the \%-in. sheet of each grade for to determine the comparative value of the three n spot welding. J - ecimens of %-in. sheet were sheared into strips ne I » in. wide and 4 in. long. A pair of strips was a g” c) Single Spot Weld ipped about % in. and a single spot weld applied oe! . in them as in Fig. 1. One set of 5 specimens of ' ; 4 a! alloy was welded as received and another set was d before welding, to remove scale. Electric Spot Welds 3 . . : \fter spot welding, the specimens were heat treated i coTealealecte J ally with the corresponding tensile specimens en tested in tension, the breaking load in pounds letermined. The results are given in Table 6. ne of spot welded specimens show that: eels spot welded satisfactorily | of scale by pickling greatly improved h 2 iniformity of joints in each case ind = the { ee ide much more rapidly | MII e pickled, the greatest average breaking strength wa mm the 3.50 per cent nickel steel Unpickled rei idium steel gave best results The unpickled } f E t nickel steel showed the greatest variatior , marked “X,” Table No. ¢ e sp weld Thess Ww f ed |} a J he eee ! 1 ter ) anger { I i sep! Shap Shap! shepl Shep lS in tensile strengt p Tig Tig Tig Ti varied from it Ss I lla 5 4 im specimens howed tendency ti Fig 1 and Upper and Left) Show Method of Welding sed. Fig. 3 (Right Lower) Size and Appearance of Tensile » . ind He i Specimens Cast from Origin: Th ooate Acetylene Welding : om Original Sheets eets, 74% in. wide by 10 in. long, were taken welding rod. No flux was used in either case. The h erade and thickness of steel. These sheets specimens were supported so as to allow free expan- in half and the cut edges milled to a 45 deg sion during welding. The sheets were laid on fire- to give a 90 deg. V when joined (Fig. 2), brick and preheated with the oxyacetylene torch to a dull red heat for about one inch back from the joint. irgist Naval Aircraft Factory, Philadelphia The Tha oa . » tas . s ‘4 % ” es 1 instalithenta aepeared’ in Tues Yaow Aa The ends of the joint were first “tacked. The %-in. sheets were welded in one run; the %-in. specimens Fig. 4) ; t | . . ee ne Back View of the Welded Sheets. This particular one is the bead side of welded sheet ' t ~ nickel-chromium, %4-in. thick, parent metal melt bar, which developed a shrinkage crack (arrow) 5 725 aot required a second run to fill the joint. The weld was little, if any, thicker than the sheet. All sheets warped slightly in welding. For uniformity, all the work was done by one welder, the most skillful operator in the shop. Fig. 4, shows the “bead” side of welded sheet No. 12, nickel-chromium, 4-in. thick, parent metal melt bar, which developed a shrinkage crack (arrow). The corresponding %-in. sheet (No. 10) developed ? aiso Fig. 5—We'ded Tensi'e Specimens After Rupture, Chrome Vanadium Steel Sheet small cracks in the weld. The reverse side of sheet No. 15, 3.50 per cent nickel, %4 in. is shown. No cracks were found in any of the nickel or chromium-vanadium welds. After welding, the sheets were cut into tensile and bend specimens as shown in Fig. 3. These specimens were heat treated identically with the corresponding standard tensile specimens before testing. The joint was not “dressed” or ground off, but was tested in the condition as welded. Average maximum and minimum tensile values for Welded Tensile Specimens After Chromium St>el Rupture, Nickel- Sheet 1% and \%-in. specimens of each type appear in Table No. 7. A comparison of the results obtained in the three steels, using parent metal and Norway iron melt bars is given in Table No. 8, with a final comparison of parent metal vs. Norway iron for all steels, in Table No. 9. Photographs of welded tensile speciments after rupture are given in Figs. 5, 6 and 7. The following points are noteworthy: All three steels can be welded either with Norway iron or with parent metal welding rod, Although the latter gives a higher maximum tensile strength, the minimum strength may be as low or lower than the minimum obtained with Norway iron rod. The latter is easier to manipulate, shows less tendency to burn and to develop shrinkage cracks, and is therefore preferable When welded with Norway iron, chromium-vanadium steel nickel-chromium gave about the tensile strength, per cent elongation and variation from minimum to and steel same maximum of these properties, chromium-vanadium being slightly superior Nickel steel 3.50 per cent gave slightly higher average tensile strength and lower elongation with greater uniformity (less variation). THE IRON AGE September 22, 192) When welded with parent metal the 3.50 per cent nic! steel is distinctly the best in all respects, having a hig! tensile strength, greater ductility and greater uniformity these qualities than the other alloys. The nickel-chrom steel is next in order and the chromium-vanadium lowest The welder found the nickel-chromium welding somewhat more difficult to manipulate than the others, a had a tendency to become spongy and “burn.” The difference in strength of made with Norw iron and parent metal is much than the differencs tween the physical properties of these materials, heat tr: welds less ment being understood in both cases. See Tables No and 9. It is noteworthy that the minimum strength values parent metal welds were less than those for Norway welds in the chromium-vanadium and nickel-chromium st: This was no doubt due to the better quality of welds n with the Norway iron and that the latter more reliable the conditions of present test. indicates medium, at least under Bend Tests A few bend tests on welded (heat treated) joir were made. Results are given in Table No. 10. T! welds were not dressed, but tested in their natural « dition as with the welded tensile specimens. Note: The operator who performed these be: Fig. Welded Tensile Specimen After Rupture, 3.50 | Cent Nicke) Steel tests recorded which was the convex (tension) side i: bending, but failed to arrange this variable in a sys- tematic manner, placing the “bead” or built-up side sometimes up and sometimes down, so that the results are not always strictly comparable. The 3.50 per cent nickel steel specimens showed dis- tinctly the best bending qualities, with the chromium- vanadium second and the nickel-chromium last, whicn corresponds to the order of merit in the welded ten sile specimens. The angle of bend withstood by the parent meta! CUvenepnNeeaNNsonNNenONAenOauennaONRErOONERDONEEDTOEEEDUEEDOOOOUOOEDEEENAEDOUNCEDONEDEONEROOUTRDOOOPOLENEDOUARESOYNEEGOOEEO ERED LONSEDEOAUDOODDDOGD boesHONDIOnABEOHTT Table No. 6—Electric Spot Welds, Tested in Tensio (Values Give Breaking Load in Pounds) Chromium- Nickel- 3.50 Per Ce Vanadium Chromium Nickel Spec. Not Not N No, Pickled Pickled Pickled Pickled Pickled P 1 4,650 2,745 3,940 2,055 3,585 1,9 2 3,755 2,080 3,740 2,455 4,750 2,5 3 3,715 4,855x 3,160 2,510 4,385 ovo 4 4,770 2,445 3,130 2,490 5,675x } 5 4,105 3,860 2,280 2,095 3,560 Aver. 4,199 3,197 3,250 2,321 4,391 “ Heat Treatment: Spot welded specimens were g'V' same heat treatment as tensile specimens, aimed 150,000 lb. pr sq. in. in the steel. (See Table 3.) Resuis E checked by BPrinell hardness readings. b Remarks: Superiority of the specimens of all typ ; which scale has been’removed by pickling is notewor'! q The chromium-vanadium steel gives the most satis! Pe results, considering strength and uniformity. : By Specimens marked “x” broke in tension, tangent ' : weld. All others failed by shearing through the weld , welds was practically the same as that of the Norw@: iron welds for each given alloy and thickness. The results obtained from these few tensil: bend tests show great irregularity, due chief! sponginess and defective contact in the weld, but believed that a careful welder could obtain more °°” sistent results with practice. Too much depencence should not be placed on welded joints in any case, 10W- and + ) September 22, 1921 THE IRON AGE 727 Table No. 7—Ory-Acetylene Welds, Tension Tests Be Chromium-Vanadium Steel ea Rating Elong., Elong., Per Cent Thick, Ultimate, Lb. Per Sq. In. Per Cent in 2 In. Per Cent in 4 In. of 150,000 Melt Bar In. Ave. Min.-Max. Ave. Min.-Max. Ave Min.-Max Average Norway 1% 83,200 75,400 2.6 1.0 1.3 1.0 55.5 , 94,700 3.5 ye Parent 14 47,700 25,800 0.5 0.0 0.2 0.0 31.7 85,500 1.5 0.7 Norway A 59.000 49.000 6.0 1.0 3.0 20 9.2 66,300 7.0 3.5 Bs Parent “4 101,300 90,300 3.2 1.0 1.9 1.0 67.4 115,200 7.0 3.5 Nickel-Chromium Steel Norway 14 74,400 50,200 2.0 2.0 1.0 1.0 49.5 90,500 20 1.0 Parent 1g 83,500 40,700 0.2 0.0 0.1 0.0 55.6 129,600 1.0 0.5 Norway A 58,800 46.100 6.3 6.0 3.5 3.0 39.2 68,500 7.0 - 4.0 Parent A 102,200 66,100 2.5 2.0 1.2 1.0 68.0 135,500 3.0 1.5 Nickel Steel, 3.50 Per Cent Norway 14 87,250 80,100 2.5 2 1.2 1.0 8.1 97,100 3.0 1.5 Parent 1g 132,200 120,400 2.5 2.5 1.2 1.32 88.2 136,800 2.5 1.2 Norway \Y 69.050 50.800 4.6 3.5 2.3 1.7 46.0 95,800 6.0 3.0 Parent 4 105,050 72,700 3.8 3.0 1.3 1.5 69.7 137,400 4.5 2.2 All breaks occurred at weld. Each average value represents four tests peneoersaneeny SUS PROALEREEUEAERO ENE OREN TA u ‘ ine ten OUD UCLONCSEREODOORELDAEODSOREEOELORPOROR ONT ING |) (OEbs S080RET EO ORRSTE DRED ERORO REDON TERS REODRRERSROERDO NRE DENEREeE EEG. MET KIT TT TY , and they should not be required to transmit high es. A breaking strength of 25,000 lb. per sq. in. i seem to be a safe allowance for welded joints, treated. In special cases, this might be increased {0,000 lb. per sq. in. Bs Since starting this investigation a considerable “ nt of chromium-vanadium steel of similar com- tion has been welded in the shop in the manufac- i‘ ture of fittings, with satisfactory results, Norway iron 3 velding wire being used. No high stresses are trans- P tted through welds. Brazing and Machining : \ny of the steels may be brazed by the ordinary y ethods. The heat of brazing would, of course, remove / effects of previous heat treatment, and the latter therefore follow brazing. Brazing material of er cent copper, 20 per cent zinc composition has a ng point between 1650 and 1750 deg. Fahr., and withstand the temperature of heat treatment of hromium-vanadium steel. Although in the lat- ase, the margin of safety is rather small, satis- y results have been obtained for a considerable in production with brazed chromium-vanadium ngs which require quenching at 1625 deg. Fahr. treatment tends to strengthen the brazed joint, are often held together by electric spot weld- efore being brazed. This also prevents letting go the brazing material softens in heat treatment. | joints develop a tensile strength of about 40,000 lb. per sq. in. and are remarkably uniform. There is less tendency to damage the steel by overheating in brazing than in welding. All the steels were found more difficult to machine than low carbon steel sheet, but could be sheared, milled, drilled, etc., with ordinary tools in the annealed state with reasonable facility. Tests made at another time on steels of similar analysis, heat treated for 125,000 to 150,000 lb. per sq. in. showed that they could be drilled and machined on milling cutter or lathe with ordinary drills or high-speed cutters, al- though with some difficulty. Impact and fatigue tests, at least of a comparative nature, on these three steels would be interesting and valuable, but lack of time has prevented including them in this investigation. Chrominum-Vanadium Steel Adop