The Iron Age 1922-05-25: Vol 109 Iss 21

ESTABLISHED 1855 gms THE IRON AGE New York, May 25, 1922 VOL. 109, Ne. 2! Continued Progress of Business Recovery Shown by the Statistics of Financial Conditions, of Industrial Conditions and of Commercial Conditions—Data Classified in the Three Groups BY NATHANIEL R. WHITNEY* have improved, and unquestionably confidence imong business men has increased enormously. The most striking evidence of this change in sentiment . furnished by the activity in the stock and bond mar- ket. Does this improvement rest upon a solid founda- tion or is it primarily speculative? This article presents evidence which, to the writer, indicates that better nderlying economic conditions are responsible for this new confidence and that these improved conditions give assurance of a continuing upward trend of the busi- ness curve. F ROM all sides one hears that business conditions In a previous article+ it was suggested that the data used to estimate present business conditions and . forecast coming developments might be arranged in three main groups—(1) that relating to financial con- litions; (2) that relating to industrial conditions; and (3) that relating to commercial or trading conditions. It was pointed out that financial data are of much more mportance in forecasting future developments than the information assembled under the other two heads, nd that a change in the third group is usually pre- eded by a change in the second group. It is a commonplace of economics to say that there in be no prosperity without ample production; it is ilso trite to repeat that production depends upon the pply and use of land, labor, capital, and managerial abilit While these are commonplaces they require t repetition, inasmuch as the usual explanations ressions and the usual suggestions for improve- gnore these fundamental facts. If these are the requisites for prosperity, it is a relatively simple mat- ter letermine which of these is the limiting factor { any given time, and therefore to say which it is ‘ needs correction in a period of depression in order tore prosperity. The limiting factor which has ponsible for the current depression, and which nust be supplied in order to restore prosperity, is ndamentally, capital consists of material things nes, buildings, tools, and raw materials. These of finance, college of engineering and com- ersity of Cincinnati \ AGE, March 23, 1922. are commonly valued, just as are all other economic goods, in terms of money units, and we thus speak of the capital of an individual or a corporation as con- sisting of a given number of dollars. This leads to confusion, and a great many persons think of dollars as capital, whereas in reality capital consists of the goods themselves, and the valuation in terms of dollars is merely a convenient way of measuring their volume. These are the materials, then, which we say are needed with labor, land, and entrepreneurial ability for in- creased production. Capital goods are used in industry in two forms. Sometimes the materials and machines are embodied in a fixed form located at a definite place and used in connection with a specific type of industry. In that situation they may be used for a longer or shorter period, sometimes, as in the case of buildings, lasting for years. Other portions of these materials are not allocated at any given time to a definite industry to be kept for exclusive use in that industry. They are free and mobile; they are available to be used in various lines of industry. This gives rise to the distinction between fixed capital and free or circulating capital. The capital which is fixed is obviously useful only in connection with the industry in which it is “fixed.” The fact that there may be but slight use for it in that particular industry, and that there may be a big de- mand for it in other industries, is of little significance in attempting to estimate the available supply. Since it is more or less permanently tied up in the given industry, it cannot be shifted to the other lines in which activity may at the time be greater. In the case of circulating capital the situation is very different. These materials are like the mobile reserve units in an army. They may be called upon wherever there is need for them. The relations, however, between fixed and circu- lating capital are close. Most industry calls for the constant employment in a fixed form of certain of these materials, and this supply of fixed capital can be recruited usually only from the supply of circulating capital. Fixed and Circulating Capital It was the scarcity of capital in both forms which brought on the depression. One of the factors re- Y tabulating certain information falling into one of the three general divisions given below, the author analyzes the showing in business so far this year. 1.—Financial Conditions, 2.—Industrial Conditions, The first group of data helps to forecast future developments. 3.—Commercial or Trading Conditions. A change in the Second Group usually precedes a corresponding change in the third. Ay eren enema vonrveneaeennearene vens 190) cA Maerrs eed tere reer renenrener seentenees: 4417 Se ee tt a oh ith eat 2A ears man 1 ey ER or o yeaa Se Pe le! MPR aN SEP ae By ee - Prices tg A A OO ante eect eee: EERIE ee = eh i a ¥? 1418 THE IRON AGE May 25, 1929 Pet! quired for production was not available in sufficient on a narrow margin could not longer afford t - ‘ amount. But while the total amount was insufficient it. Unfortunately for both groups, the owners xed ; ; some industries possessed more than their normal pro- capital in excess of their needs could not transi portion of the total capital supply in a fixed form. At those who had need of free capital. Hence jy , the same time the supply of circulating capital did not activity on the part of such producers had to a i keep pace with needs. Therefore, the cost of capital accumulation of capital and the lowering of to producers mounted until those who were operating to a point where they could use it profitably. 4 1 e XY . 7 - ] I di - ‘ re > ‘ . . ~ € | Forecasting General Business from Financial Indication: ’ S has been said, there is need for capital in both The San Francisco Reserve Bank listed say e forms. Since in almost all cases capital appears posits in 74 banks as follows: first in the circulating form, we can obtain a clew to Maes 81, BO886. Sacissswncnasae $704,737 ’ he possibilities of an increase in this necessary factor RE 5 EG REEM oh cic a. nclnwiom denen 728,524 " the possibilities of an increase in this necessary acto} M h 31° 1922 73112 Maren ol, VEGreveere ee. see ° iol, l . of production if we examine evidence which shows the . F : ; E , . On April 26, 1922, time deposits for 801 tendency in the production and accumulation of circu- : + lating capital banks in the reserve system showed an incr: n ce . . $246,000,000 over April 27 of the previous yea 1. One of the best clues as to the accumulation of s : ; ; oe f : The evidence thus presented from these capital is the trend of savings bank deposits. They : ae ; : ae 7 sections of the country indicates that, while indicate roughly the excess of production over consump- ; ce . , : s deposits have made no startling gains, they h . ‘ tion so far as these depositors are concerned. It is : bi: A ee A Te the most part made some headway, and under e» this excess of goods not immediately consumed that ' ; z ; , 6 is s - circumstances, with a large amount of unemp| ‘ replenishes the capital supply. Unfortunately no : : : , ria ao and dullness in various lines of industry durin; oe figures are available for savings deposits since the 7 Y past year, the mere fact that they have not lost ¢ first of the year for the country as a whole. The =: : : ; is encouraging. However, we must look furth nearest approach to such figures is furnished by de- : ; : ; > Th of evidence as to the accumulation of capital, since there e posits in postal savings banks. trend such ; : ; ear : i are many forms in which free capital is found hx deposits is indicated in the following table: : ; : savings bank deposits. Postal Savings Deposits 1921 / 1921 ; Free Capital Expansion Shown by Bond Sales January $163,656,000 October $151,150,000 ? February 163,356,000 November 149,400,000 9 ——, . savere kaa 3 Se cE aaeae. . Manca 148'000,000 2. A growing number of savers keep their accu- April 158,096,000 mulations in the form of invesment securities. Th May 155,394,000 1922 . : ; : June 152.389.000 January . 146,500,000 Volume of capital accumulations in this form may be ul 15 O.000 “ebr ¢ \ 7 ) . . ° . ao Try Toye oe lis'ononon gaged by ascertaining the course of bond prices and Septembe1 152,400,000 April ........... 144,500,000 the interest yield on good bonds. The volume of bonds ‘ This table shows a steady decline, though to a very sold on the New York exchange during March and moderate extent, throughout 1921 and the first four April was the heaviest ever reported. The transa months of 1922; but the declines in total are getting tions in listed bonds other than United States Govern- smaller and many offices are now reporting an upward ment securities in March amounted to $237,000,000. al tendency. Furthermore, note must be made of the This was more than double the volume of sales during » fact that postmasters have reported that many de- March, 1921. During April the total was even large: positors are converting their postal savings deposits and prices reached the highest levels attained since th we a into treasury savings securities which offer a higher close of 1918. During March and April the 8 per cent rate of interest. rate on corporation bonds practically disappeared, and A more favorable showing is obtained when we ex- an increasing number of bonds was brought out at a amine the figures for savings deposits in the banks in rate ranging between 6 and 7 per cent. During Apri various parts of the country. The Boston Reserve’ the city of New York sold $45,000,000 worth of 41,4’: = Bank reported time deposits in member banks in Port- at a price to yield about 4.11 per cent. This compares land, Fall River, New Bedford, Springfield, Worcester, with a yield of 4.33 per cent on a large issue sold by Providence, Hartford, and New Haven, as follows: the city in December. Even at this low yield the Apri! April 15, 1921 _ $137.703,000 issue was oversubscribed about eight times. conn ca. anes -: 550739 600 The fact that the volume of free capital is increas- . : . ing is further indicated by the declining interest yiel i The Philadelphia Reserve Bank reported time de- & : ee oe d ing oa tl ag yh: ' on Government borrowings. The following table shows posits in member banks in Philadelphia, Camden, t] ‘eld : i ial ele ‘ ie 1e yield at vé s s as years ." Scranton, and Wilmington, as follows: yie at various periods during the last four yea 7 3 Issuing Rate on Certificates and Notes April 15, 1921 $40,950,000 : ; : March 15, 1922 18,236.000 About About 3—4 Yi } April 19, 1922 49'003.000 6 Months 1 Year Year Liber - : ‘ F Maturity Maturity Notes Be It reported also that the savings deposits in 80 re- January, 1918 { : ° or Ji ary, 919 i, ae 4.¢ porting banks on April 1, 1922, were 11.1 per cent january eae i? 1 4 greater than on April 1, 1920; 2 per cent greater — a: 5% 6 - ecempber, 920 o™% h : than on April 1, 1921; and 0.1 per cent greater than March, 1921.. 514 5% ms f 9° June, 1921 bly a4 d s are Q99 , 2 4 on March 1, 1922. September, 1921 5 5% 5, * The Atlanta Reserve Bank quotes savings deposits November, 1921 114 11% ae i 9 ‘ 7 February, 1922 Sas ‘ 4% + from 82 reporting banks as follows: April, 1922..... . sx an ee { : ren 3, 2223 - saws Cer aan aes A similar tendency is illustrated in the increased Teh 28, 1922 52.298, 006 > - } nd March 3i, 1922 . 154,562,000 demand for investment stocks, both preferred ané . The Kansas City Reserve Bank reported savings common. This demand has carried the price of man) deposits in 62 reporting banks as follows: of these stocks to the highest point touched for fork L, 1921. $74 $24,696 number of years. April. 1, 1922 . 78,411,736 8. Although the supply of money for short periods At the same time the number of savings accounts in of time is not determined by the same conditions tna‘ these banks increased between these dates from 222,730, determine the capital supply, there is some relation- to 267,578, to 276,069. ship between the two. While for the country 4s 4 ” May 25, 1922 THE IRON AGE 1419 an increase in the supply of bank credit and a_ banks were $616,654,000 as compared with $2,160,547,- in the discount rate may occur even though the 000 in the corresponding week of 1921. Reports from fund may not increase, nevertheless from the all sections of the country indicate liquidation of loans int of the individual business man a supply of in both the member banks and in the Federal reserve redit usually serves him in the same manner as_ banks. Loans of the New York City member banks y of free capital. Furthermore, bank credit, have declined $1,204,000,000 since they reached their he banks are on a sound basis, must bear some maximum in October, 1919. On April 26 loans of Bos- to the capital supply. This proper relation is ton member banks to their customers were smaller in ‘o be ignored in a period of inflation, but in a total than at any time since 1919, while borrowings ‘f depression when both banks and business of these banks from the Federal reserve bank were at » much more cautious the trend of the discount the lowest point since 1917. On April 19, 1922, the r commercial paper is likely to be comparable borrowings of member banks in Philadelphia, Camden, eld on capital. The accompanying graph shows Seranton, and Wilmington were $21,140,000 as com- irse of the interest rate during the past two pared with $108,748,000 a year earlier. In the San 1r 60-90 day prime two-name commercial paper. Francisco district borrowings by member banks from the Federal reserve bank declined 19.7 per cent between errr T i Crore March 8 and April 12, 1922. This is all the more re- markable in view of the fact that at this season of the year the demands on the country banks are usually heavy. | 5. The decline in the volume of Federal reserve notes outstanding suggests also a decrease in the volume of credit. The following table gives the aver- errr re age value of Federal reserve notes in circulation for ee pase We. il RR > Fe Raglan Ric igs a, Bigatti er ak a, ST 9 A ag eS aoe MN Oo itn. te each month: e Federal Reserve Notes in Circulation Jaf 1921 1921 1 { L l i January $3,159,000,000 October $2,476,000,000 og s SRBZSSE = * : PB SESE February 3,037,000,000 November 2,398,000,000 if : ages oe F ae March 2 963,000,000 December 2,394,000,000 ai 7 920 92 922 April 2 869,000,000 1922 $e a . — , . ata May 2.767,000,000 January . 2,230,000,000 ; B3 tverage of 60-90 psy a e Two-Name Comm ial To 2°674.000,000 February 2°170.000.000 i 1 ; pe July . 2,604,000,000 March 2,189,000,000 ‘% 5 tate, Rat August 2.504,.000,000 April 2,200,000,000 ’ i Per Cent 1921 Per Cent September 2,492,000,000 May 3. 2,173,000,000 ie } : 6.0 March 7.62 j . i” 6.4 a 4.00 6. The increase in the ratio of total reserves to le 6.7 May 7.12 . . ° onitn @ ’ ft 6.8 June 675 deposit and Federal reserve note liabilities combined oa} is es eo during the past sixteen months further supports the a 7.8 ounewabes 5.88 assertion that the use of credit has declined and that j 8.0 October 5.75 ‘ . . ; 80 November y pari passu a constantly growing amount of fixed cap- 8.0 Decembe! 5.12 ite has | . rerte ; fre . ; ] ! 4 a2 1992 ital has been converted into free capital. tap 7.88 January 1.90 7. The value of securities issued for new supplies 1 oe February 4.88 ; ° ' eh 7.75 March . i73 of capital throw light upon the amount of capital 1.75 April 4.65 which is being accumulated, since the issue of such securities implies their purchase by savers and in- vestors. Domestic capital issues for April, 1922, were paper 4% to 4% per cent. In nine of the $445,196,060, as compared with $380,668,000 for April, twelve Federal reserve districts the reserve bank dis- 1921, and $471,725,600 for April, 1920. For the first nt rate is now 4% per cent. four months of each of these years the totals were, i. The supply of free capital is being augmented 999 ¢1 141,331,430; 1921, $1,075,500,200; 1920, $1,288,- time to time by funds released from other em- 449 509. During the early part of May, Dun reported com- New foreign securities offered in this market ments. Large amounts of capital have been tied from Jan. 1 to April 25, amounted to $514,657,100. ring the past few years in raw materials and jy, spite of these heavy supplies of securities, the mar- shed goods in inventories which were not being used et for new bond issues still seems far from satisfied; vely in the form in which they existed. Thus the total of new bonds taken during the last week in he time being a large fund of capital and its con- April was $230,000,000. into circulating capital makes for prosperity, thus becomes available for entrepreneurs in ries where it can be used productively. For ex- hundreds of millions of dollars have been tied up ntories and in farm products which were not ised productively. To hold these commodities it essary for the owners to borrow current funds the banks. Therefore, if the evidence shows that ins throughout the country have declined ma- 8. No estimate of the supply of capital now avail- able or likely to be available in the immediate future is complete without taking into account the extent to which Government finance is likely to draw upon the funds out of which savings must be made. A compar- ison of ordinary disbursements with ordinary receipts of the Federal Government is presented in the follow- ing table: ee, ; Disbursements Receipts ry will indicate that some of these temporarily April. 1921 $494.091.189 $296.170.665 A tive goods have been disposed of to persons March, 1922 325,954,936 550,757,683 ii make use of them. In other words, these aaene Salas eects aes of fixed capital have been converted into cir- Of the large receipts indicated for March, 1922, si apital. What evidence is there of any con- internal revenue receipts made up $477,053,845 of the 4 liquidation of bank credit? total. Secretary Mellon estimates that there will be a f treasury deficit for the fiscal year ending 1923 of ap- proximately $484,000,000. There is little prospect of April 26, 1922, reports from 801 members of payments on account of European debts, and it is eral reserve system showed loans and discounts therefore certain that heavy taxation will have to ‘02,000,000 less than on April 27, 1921. On continue. If Congress should vote a cash bonus the 1922 22, the bill holdings of the Federal reserve drain upon the treasury, and therefore upon the cur- Sta a died Proofs of Conversion of Fixed Into Free Capital 1420 THE IRON AGE May 25, 1929 rent savings of the people, would be increased to that extent. The Government is making progress in funding the short-dated debt of the country. During March nearly $650,000,000 worth of Victory notes were retired. While the payment of these bonds represents merely a substitution of other obligations for those paid it is nevertheless a gain in so far as the lower rate of in- terest paid on the new obligations will ultimately make possible a lower rate of taxation and will thus release a larger share of current income for addition to the stock of capital. Way Prepared for Industrial and Commercial Activity On the basis of the information here given we have warrant for concluding that the accumulation of cap- ital, which was hampered for several years, is now gaining headway. Production, therefore, is not so likely to be retarded by limitations in the supply of this necessary factor. The supply of capital, however, is only half of the problem. Supply means little in itself except as it is considered in connection with the demand for capital. Indeed, one explanation of the recent accumulation of free capital is that business men have reduced demands for capital during the period of busines. 4, pression. When they again come into the mark.: ¢ supplies of this factor of production the dema: be increased, with the result that the supply relatively diminished. This explains why p toward a normal condition of business is certaj) forward at uneven speed. In a period of dullness circulating capita! creased. This offers opportunity for industria! ity. The new activity uses up temporarily a | able part of the capital accumulated, and f progress is then halted or retarded until gradua mulation again puts ample supplies on the yn This causes another burst of industrial activit, the process is repeated. At present, and for th month, we seem to be in such a period. The for high grade bonds have not increased not} in the past three or four weeks, and it is repo financial circles that there are large blocks of m curity issues which have been prepared but whi being withheld from the market until there is evidence that investors have digested the pr large issues and have accumulated surplus liquid ital for the new offerings. Data Bearing on Revival in Industrial Situation VIDENCE furnished by an examination of the financial data given clearly shows, however, that the way is prepared for industrial and commercial re- vival. If the above reasoning has been correct, we ought to be able to observe from statistics of industrial activity signs of progress in this direction. As indi- cated in the article before referred to, which appeared in THE IRON AGE for March 23, the evidence upon which a judgment concerning industrial conditions may be based can be examined under four main heads. The Showing of Unemployment Figures 1. Unemployment. Reports of the departments of labor of the United States and of New York State show continued gains during March and April in the number of persons employed throughout the country. Records of the United States employment service for 65 indus- trial centers showed an increase in the number of per- sons employed of 2.5 per cent between March 1 and April 1. Similar evidence is obtained from statistics for wage payments. The total payments to employees of 1600 representative industrial establishments in New York City show a marked increase during the past few months. This was due both to an increase in the num- ber of employees and to an increase in the time of em- ployment for each worker. The average weekly earn- ings in the factories in New York State as a whole in creased 2 per cent in March over February. Along with the subject of unemployment, wage con- troversies or other labor difficulties which hamper pro duction and increase unemployment must be considered. Happily, the situation throughout the country is less serious than it was at about this time a year ago. There were only two industries according to the figures of the Department of Labor in which there was any con- siderable decrease in the number of persons employed. These were cotton manufacturing and coal mining. In both of these industries strikes are in progress. Reports from various cities indicate that there is still much friction between building contractors and building laborers; but the phenomenal increase in the demand for building laborers for the new construction program throughout the country has prevented this dis- satisfaction from becoming acute in most centers. Of course a part of the increased employment results from seasonal demands, but there can be no question in the mind of the observer that over and above these demands there has been a real increase in industrial act which has resulted in a larger amount of employment for wage earners. Industrial Activity as Shown by Production 2. Volume of production. Of the basic industries agriculture is undoubtedly the most important. While it is still too early to obtain close estimates of produc- tion, forecasts are already being made which are mod erately favorable. As to wheat, the condition of the growing crop was, on May 1, rated from 81.7 to 85 pe! cent as compared with 88.8 per cent last year at time, and with 87.1 per cent for the ten year averag for this date. Visible supplies of wheat in the United States and Canada are, however, about double what they were last year at this time. The outlook for wi ter wheat is for an average yield. These figures be improved as the season advances. That there w be a large export demand for our crop is cert inasmuch as the wheat crops of last year in Argentina and Australia were small. Russia, which is normall) a large wheat producing nation, did not produce enoug! for her own needs last year. As to oats, reports indicate that there will be a a 10 per cent decrease in acreage. Prospects for are very uncertain inasmuch as planting in the belt has been delayed by wet weather. Howeve! position of the corn belt farmers has been improved because of the rapid distribution of last year’s The receipts of corn at Chicago between Nov. May 1 were 103,821,000 bu., but by the end ot period the stock on hand in that market showed a! crease of only 2,049,000 bu. For the same period year the receipts were only 66,023,000 bu., and the on hand increased 3,164,000 bu. In the cotton belt heavy rains have delayed planting, but it is exp that a larger acreage will be planted than last Reports from Texas suggest a 20 per cent incré acreage. The volume of production in the iron and st dustries has far-reaching influence on business dev‘ ment. Striking gains have been made during th: few months in these industries. The production of | e iron increased from 864,555 gross tons in July, 1%-': to 1,649,086 tons in December, and these figures %4"' further increased in the first four months of this as is shown herewith. LURE os feeder eteredans deawee cums 1,644,951 es MPO TRIER TT oe --» 1,629,991 Mareh Terr rer ee Te : 2,035,920 | eee caneenestece ae de 2,072,114 increase in April is all the more surprising in f the existence of the coal strike. production of steel ingots, as indicated by re- from companies representing about 84 per cent total 1920 production was, in July, 1921, 803,376 n December, 1921, 1,427,093 tons. This output en tremendously increased as the following fig- for the first four months of'the year 1922, will i 1,593,482 ary 1,745,022 n 2.370.751 cee wawe en rk 2,439,246 ther evidence of improvement in the steel in- is furnished by the reports of increased opera- d growth in the volume of unfilled orders. Al the United States Steel Corporation was oper little more than 50 per cent of capacity in ary, and is now operating at approximately 8) ent of capacity, its unfilled orders have been in- y On Dec. 31, 1921, it had unfilled orders on for 4,270,000 tons; on Jan. 31, 4,240,000 tons; 28, 4,140.000 tons; on March 31, 4,490,000 tons \pril 39, 5,100,000 tons. production of coal, large gains were made it hree months of 1922 as compared with 1921. s set forth in the following table. foal Fioductioi , fle i'n i Sta Bituminous Anthracite 1922 1921 192 192 6,615,000 S.570,000 1.607.000 ] 1 706,000 8,132,000 L.S1LL.OOO 1.985.000 10,309,000 7,859,000 1,822,000 2 O48 000 10,285,000 7.489.000 1.703.000 a 10,402,000 7.432.000 1,701,000 1.816.601 10,541,000 7,278,000 1,913,000 1.902 11,102,000 6,900,000 1.982.000 1.92 10,843,000 6.512.000 1,907,000 1.68706 11,448,000 6.457.000 2 095.000 1.564.008 1 463,000 5,822,000 1,896,000 1,1 figures since the first of April are of little sig e as showing the trend of the industry because has been cut down very markedly by th: lumber industry more activity is reported any time since the boom period of the spring 20. The San Francisco reserve bank reports that reat lumber section of the Northwest, produc March was 14.8 per cent greater than in Feb and 60.9 per cent greater than in March, 1921. present rate of production was reported to equa t 90 per cent of normal. In spite of increase in out uilding demand resulted in orders received by lucers in the San Francisco district 22.4 per cent than actual production during March. During | the beginning of May orders throughout yntinued to pile up, and some of the larges rs instructed their salesmen to advance p) $3 a thousand on various lines. the railroad equipment industry further evidenc: 1 of increased production. Up to the middle of lers had been placed for a total of over 67,000 ars, or nearly three times the number ordered entire year of 1921. The locomotive plants have yrted that they are doing a larger volume of than they have done for several years. 1utomobile industry is operating at high speed. ss is particularly active in pleasure cars, some companies having established new high records roduction and sales. The production of pleasure March totaled 152,920—40 per cent more than lary and the largest monthly total since last ist. Furthermore, factories are said to have orders -h on hand to maintain the present rate of output everal months to come. The Ford plant has made THE IRON AGE 1421 preparation to turn out 1,150,000 cars this year as com- pared with 1,038,000 last year. In the manufacture of woolen goods great increases have been registered. The following table, giving the consumption of wool monthly during the past two years, shows that in March, 1922, a record for consumption of wool was made which has not been equaled since January, 1920. Volume of Production Wool Consumption 1922 1921 1920 Pounds Pounds Pounds January 94,550,000 26.300.000 63.064.000 February 25,882,000 31,449,000 55,249,000 March 62,585,000 40,828,000 58,349,000 April 46,253,000 7,889,000 May § 5.000 50,650,000 June 000 10,681,000 July 000 32,274,000 August 5,000 32,848,000 September 000 30. 932,000 October 59. 836.000 33,700,000 November 9.000 24,144,000 December »4.950.000 "0.864.000 -” Total »7 2,499,000 00,744,000 In building operations activity is very marked. Dun reports that information gathered from 40 centers in — (50, 7 . itiininaninaipnaiienemiiiines 50) | T T aki hut NORMAL YU NWO OM y oth) 7 a A * ee. Ji / PRODUCTION Q ‘ 4 4 2 > Sen v OV & €0h a p AD NU? o r Lr, iA ) Le SS widely separated sections of the country shows new building being carried on in tremendous volume. For the first three months building permits in approxi- mately a hundred selected cities were $470,196,656 as compared with $240,511,410 for the first three months of 1921. The Annalist reports new building authorized in 142 cities for April at $192,184,353. This compares with $112,373,483 in April of last year. A composite picture of the progress made in pro- duction is presented in the following table which is taken from the monthly bulletin of the New York re- serve bank. It is of interest to note that of eleven (Continued on page 1458) et ee ee at ee ot Nef Ry tn aok nom » cs aaa sian Sa ow Bry The Arc Welding of Structural Stee! Tests Made to Determine Reliability of Welded Joints and Limiting Conditions of Various Types of Joints BY E. S. HUMPHRYS, JR.* APID growth of the application of are welding to structural parts has inspired the writer to -arry on a series of tests which would tend to show what reliability could be attached to this method of construction. Naturally, one of the first questions was, “Why should the parts of a structure be electri- cally welded rather than riveted 9 Analysis of this subject showed that great economies can be effected by the arc welding method and that, in addition, the necessary high quality of workmanship can as easily be obtained. To visualize the tests as they were conducted, it may be stated that first consideration was given to the lay-out of a building, 40 x 40 x 40 ft. In this connection it was found that the amount of time consumed in the drafting room, and the drafting room costs, would be approximately the same for the structure whether welded or riveted. In designing structural parts, when they are to be riveted together, the size and position of rivet holes are shown; in designing a structure for 2 ' 3-2 ’ 7 . >! o” uw , "uv iny f* sia ‘ He --G < 29 >< 7 ~-->4 Ticsiliceentinimnmiaeiat i =» XT a | wy | | j i ¥ i | i . Leon . “i! ; - iicemaiaincenceiminey | — - a — —_ = Pl. =x /-4x45 (L 32 x Shx2 240” o c c’3 Fig. 1 Angle Lap-Welded in Shear to Two Flat Plates welded joints, all such connections must be designated by shaded portions. Because at least two holes must be punched for each rivet, great saving is effected in the use of arc welded joints, where the cost of punching holes is almost entirely eliminated. Those familiar with this type of construction will appreciate that the cost of drifting enters also, to a greater or less extent, and the elimination of this variable expense is an addi- tional argument in favor of arc welding methods. Let us consider what would have been the cost of riveting the structure 40 x 40 x 40 ft., then let us com- pare, with the figure thus obtained, the actual cost of construction which was done by welding. Structural steel used Welding Cost 165 welding hour t (average) $90.7 150 lb. welding wirs : 7 00 $105.7 Riveting Costs (estin Riveters, 78 hr. at $175.50 Rivets, 1800 at 1.2 ye A A 197 7 Notes Wages of four-man riveting gang: $2.25 per hr. per gang. Estimated, 4500 holes for 1800 rivets On a structure of this type it is estimated that there would be driven no more than 200 rivets per day by one four-man gang. *Union College, Schenectady, N. ¥ In this comparison no mention has been m the following items, some of all of which would into the final cost of the welding or riveting ope Riveting: Cost of template; of prick punching holes: of holes, or of drilling and reaming them. Percentage of rivets lost of those ordered Cost of fuel consumption in the rivet heate1 cu. ft.; of making a hammer test on all rivets. Welding: Cost of fuel, in this case electricity, figured at per kwh.; of sand blasting surfaces about to be w: free them from rust or scale, a practice sometimes Observations taken through the investigatior convincingly that at times fabrication can bh. ducted more easily and at greater convenience by welding rather than riveting methods. Labor cha are reduced and time saved because no material be handled twice; everything can be done in such a that there is a minimum of lost motion. More fully to appraciate what constitutes a | manipulation of the arc, the writer enrolled arc welding school of the General Electric Co., Sch tady, there taking the practical course in welding. 1 enabled him to assist in supervising the welding test pieces, in detecting any intentional carelessnes the part of the operator, and at the same time in s curing a full appreciation of the labor involved in weld ing a structural joint similar to one which ordinarily would be riveted. All the test data, which will be incorporated further along, relative to tensile strengths of various typ: and sizes of welds, were obtained in the testing labora tory at Union College, using for the purpose a 200,000 lb. tensile machine. One important feature, which should be given co: sideration, is the relation between the total dead we of a truss wholly welded and that of the same tr riveted. A number of tests and computations | shown that a welded truss of the same strength is preciably lighter. Is not this fact alone sufficient son why executives who are interested in struct work should give careful attention to this fast-grovw and dependable method of construction? The intention of this article is not to point out definite or standard method of fastening together st tural parts preparatory to welding, since in the jority of instances the local company—assuming t! this work is carried on under a structural superinte! ent—will be in a much better position to recog) the most efficient ways of assembling jigs and clam} For these reasons the endeavor has been to ascerta only the true value of a welded joint as compared that of a similar joint, riveted. An additional fact which should be mentioned fore the test data are illustrated is this, that a w and one helper together are capable of producing mor: work in a given time than that which could be turne! out in the same time by a riveting gang, the latter co! sisting of from three to five men. Broken Specimen Corresponding to Fig. 1. The angle broke; the weld held 1422 re undertaking this investigation, the writer n under the impression that in all past struc- welded members far more electrode material n applied than was necessary to make the joint and one which would withstand safely the which it was subjected. This impression was ated later, as the results of test showed. welded specimens were prepared at the General Co., under the supervision of W. V. Lewis, al supervisor, and welded by the departmental _W. Kutassy. The latter has been welding for ten months only, yet great dependence is placed fforts. I cite this fact, that it may be realized tual welding operations were carried on to give age rather than a maximum value. structural members used in the tests were taken tock, and no especial preparation of the sur- 5% “ s ci * a“ P gxl x43 OAs. 2x6 53h ove). Two Cover Plates Welded to Two End Plates Side of the Halftone Is the Broken Specimen ng to Fig. 2. Again the plates were too weak fort The right specimen in the halftone represents Here the combined angles proved too strong for the h, as shown in Fig. 3, below, were not continuous be welded was made other than the cleaning surfaces with a small wire brush. What scale her surface impurities resisted the brush were to remain. It is not meant to infer, however, ere should be no treatment of such surfaces; a small, portable sand blasting equipment is ended, with which may be removed the tougher re objectionable impurities. electrode or filling material used in joining ictural members was a standard, commercial, re, the approximate analysis of which is as Carbon—low Manganese—high. Sulphur—low. Phosphorus—low Silicon—trace sts may be divided into two groups: Six preliminary tests, made to determine the ate strengths of the different types of welds, that future joints might be designed with the of each fully developed. Five tests, following these, all being designed alike as possible, with the exception of the ross section, which was varied with the in- test. first six preliminary tests may be further di- nto two additional groups: Joints composed of two plates and two angles, two plates and one angle. Joints composed of four plates, as_ illustrated accompanying diagrams. Description of the First Six Tests angle of dimensions shown in Fig. 1, was ded in shear to two flat plates, one plate at nd of the angle. All welding areas or cross sec- are indicated by shaded portions. This specimen eripped in the tensile machine and pulled apart to THE IRON AGE 1423 determine the ultimate strength either of the welds, the plates or the angle. In advance of the test the prediction was made that the weld would go before either the plates or the angle. At 89,230 lb. the ultimate strength was reached. With the exception of a few small cracks, the welds were unharmed. Fracture occurred in the angle on a line approximately perpendicular to the line of stress and was seen to be one of shear; in addition, the speci- men bent in two directions, as would have been sup- posed. This gave no numerical indication of the weld strength in shear, save that it was greater than that of the angle. The specimen in Fig. 2 was made up of four plates (two end plates and two cover plates) welded as before, in shear. Cover plates were used instead of one or of two cover angles because it had been ob- served in the preceding test that there was induced too much bending in the specimen to give accurate results. Under stress, the two end plates elongated and be- came deformed, until the ultimate strength of the two cover plates was reached, at 158,420 lb., when the two latter broke in shear, perpendicular to the line of pull. Again it had been graphically demonstrated that the plates were too weak for the welds. The test specimen shown in Fig. 3 was made up similar to the one in Fig. 1, except that in this instance two angles instead of one were used to connect the end plates. As had been foreseen, there occurred bending in the specimen to an amount sufficient to be readily noticeable and, quite probably, to affect the results ob- tained. In this case the strength of the plates and angles proved to be too much for the combined strengths of the welds, and the latter failed at 114,750 lb. It is thought that a greater pull could have been withstood had there been no bending stress induced. It was de- sired to obtain the strength of each weld per lineal inch, but, because of the bending moment, this calculation was not made. In the fourth test, Fig. 4, two cover plates were we = NRG ian aca ly Dcaee nip ebabai si _ oe cn . » , ‘. tf i % rim 7 x Spa ee be eek 7 7 * Rito , 7 . > ee ~ : ; . ‘ r . ‘ + . . ‘ . ; . ii °° . e « > * 1424 THE IRON AGE welded, in shear, to two end plates, as in the second test. By using four plates welded together in this manner, the bending moment was reduced to a mini- mum, and the results obtained represent true values of shear. The breaking strength was 149,500 lb. Four of the eight welds broke almost simultaneously along a line parallel to the line of force and through their ap- proximate centers. Calculations shown elsewhere give values per lineal inch of weld, and for shear per square inch of welded area on a plane parallel to the length of the weld. There was no designed difference between the fifth specimen, Fig. 5, and the one immediately preceding it, except that shorter lengths of weld were deposited in the latter instance. This specimen broke as did the other, to give an ultimate strength of 116,320 lb. The lineal strength, therefore, was equal to 9685 lb. per running inch of weld; that in the fourth test was 9340 lb. In this and in the last test the weld metal was deposited continuously in one solid bead; in test pieces of the second group, which are yet to be described, each welded volume was made up of two beads, a small one running along the line of intersection of the two welded surfaces, and a second one over the first and filling in to the required volume. For cross section of bead thus described, refer to Fig. 5-A. ie eee ee Oe ee ee ¥ aimee snniteeiniaid = x F WY | 1 it | me ~ 3 a ry zs ; In 5 2s. Sx 9x35 “ Pf 2x1/Ox4s sg 8 é } } Above) Two Cover Plate Welded to Two End Plates Center of Halftone Show the Weld of Fig. 4 Pulled Apart lhe right unit shows the weld of Fig. 5, pulled apart but t entirely separated | Below ) Two Cover Plates Welded to Two Knd Plates Shorter Welds Than in Fig. 4 ee } ) f i a et ame | Up Fig. 6 shows that this specimen was designed to be the same as Fig. 5, save that the length of material deposited was as nearly as possible one-third that laid down in Fig. 5. It was thought that this weld would not be one-third as strong as in Fig. 5, because of weak nesses at the beginning and at the ending of the beads, which would be more pronounced in this, the smaller weld, thereby causing it to break sooner. This was found not to be the case On the contrary, the smaller weld was slightly more than one-third as strong as the longer weld, which, perhaps, was accounted for by the fact that in the shorter weld there was deposited a slightly greater pro portionate amount of electrode material than in the other. Further tests might prove that weld strength varies directly with the length of the weld. The ulti- mate strength of this specimen was 45,570 Ib. Frac tures occurred in the welds, parallel to the direction of stress It will have been noticed that, in the tests above described, little was given in regard to the computa- tions for strength of welds. The main reason for making the tests was to determine an approximate value for strength in shear of each type of weld so that, in designing further specimens, there would not be deposited so large an amount of electrode metal that the plates, and not the welds, would fracture. Description of Last Five Tests These five tests are grouped together because of the longitudinal dimensions common to all. In Fig. 7 the dimensions shown are constant; those left vary with the type of specimen and may be further along. Welding is indicated by the shade tions. In each of the five groups of tests about to scribed there were broken four specimens, a! being designed alike, making a total of twenty ji; ual fractures. From the data obtained calculatioy been made and conclusions drawn. The twent pieces were welded under the following conditi nearly as possible alike: Welded length, each specimen: Eight 3-in. we 2 ft Voltage: Open circuit, 60; are, 20. Steel: Structural, taken from stock Electrode: Standard make, of composition diameter, 5/32 in Size rivets: %& in, used in calculations of relatiy: of joint, welded or riveted Rivet All test pieces were of the same longitudi: mensions but, to secure the desired cross sectiona of beads, the thicknesses and widths of the plates employed necessarily varied, informat { which is given in Table I. Table J Specifications All dimensions in inches.) Ie Bead Plates No ‘A” — No. Size 5 li, 2 % x 2 x 4 i, x 6! § Z x9 &exb x 9 4 2 5 x 9 x ‘ 2 8 x 61% x 10 A % x9 x % x64 x Throughout the course of the tests a carefu surement was taken of the amount of electrode terial deposited and of the time required to weld ea specimen. A glance at Table II will show that, as would have expected, the amount of filling mate! increased as the cross sectional area of the bead was !"- creased. This was not the case when the length © time taken to weld was observed; time did not vary ¢ rectly, or even approximately, with the area of the bead May 25, 1922 fo be of commercial value, the strength of any one of a number of welded joints may not vary from their mean value by more than a small and many definitely Table ll—Welding Operations ——Electrode——, ——Time- yee Amount Rate Arc Size, (Each Spec.) Total, (Lineal Its In. Lb. Amperes Ft. In. per Min.) 20 a 1.39 175 33.00 0.73 20 3 0.966 175 19.87 1.22 20 eo 0.77 150 22.75 1.06 20 5 0.551 150 14.25 1.686 20 is 0.328 150 9.00 2.67 percentage. Our experiments demonstrate that, s type of joint, the variations in strength are smal] mpare favorably with joints riveted in the usual Solid Continuous ‘Bead \ At Left Is Shown the Solid, Continuous Welding | Used in Test 5. At right are shown the two successive ed in tests 6, 8, 9, 10 and 11 The solid black in the epresents the first bead, while the stippled outer sec- tion is the second and concluding bead The widest deviations from the mean occurred st No. 8, where the plates were heaviest and where, onsequence, the beads were of the greatest cross tional area. In four of the five groups of tests no ne variation exceeded 2.2 per cent of the mean value; the fifth (Test No. 8), the largest variation was 79 per cent. From this we conclude that the welded nion is not the changeable quantity which some might expect it to be, but is fully as dependable and as strong as any other common type of joint. Table III lists the variations as they were determined from tests made. Table I1I—Strength Variations Ultimate rest Strength of Weld Ne (Average, Lb.) Largest Percentage Variation from Mean 142,500 5.79 131,110 1.22 120,000 2.20 10 96.400 1.73 11 89,650 1.88 is interesting to compute the riveted strengths the joints tested for weld strength and to compare fe ® - C3 -- -->»> “iN kK 6 - B/ ke --6 -- “ ~F ii ¥ | . X — ¢ ‘ . -P Pls. 2% al ya “ing OPIS. & x CS X55 Pl g x7 43 Short Weld of Two Cover Plates to Two End Plates e figures with those obtained from the breaking It will be observed that these particular types nions could not be riveted to give a strength equal it of the same joints, welded. Only one rivet can be ised in each section, and it cannot exceed a diameter s in. Table IV makes the following assumptions: Ulti- ite tensile strength of mild steel ordinarily used in ructural work, 60,000 lb. per sq. in.; rivet strengths, single shear, 45,000 Ib. per sq. in.; bearing, 90,000 per sq. in Our welds gave shearing values which varied, in nsity, inversely as the cross sectional area of the id. In comparing the calculated shearing value of ictural steel with the observed values of the welds, shearing value of 45,000 Ib. per sq. in. was as- sumed as the ultimate shearing strength of the steel. ble V gives the shear for the five groups: it varies om 33,650 lb. per sq. in. in the weld of largest sections THE IRON AGE 1425 to 42,900 lb. per sq. in. in the smallest, indicating that, in some respects, a small weld is more economical of et than a a weld. Based on the value of ounecantanrentsss caneenns wana eae nstaneTiaee eens eeennnne — oni Strengths Table IV- c———— Rivets, Ultimate Calculated for Plates Double Weld Test Gross, Net, Per Cent Shear Bearing Strength, No. Lb Lb Effic. (Lb. per Sq. In.) Lb. 8 180,000 135,000 75.0 54,000 59.100 142,500 180,000 ; 20,000 66.7 78,800 6 169,000 131,500 77.8 54,000 49,250 131,110 157,700 112,500 71.3 59.100 9 159,500 122,200 76.6 54,000 49,250 120,000 157,700 112,500 71.3 59.100 10 159.500 122,200 76.5 54.000 49,250 96,400 105,000 75,000 71.4 39,400 11 150.000 112.500 75.0 54,000 49,250 89,650 105.000 75.000 71.4 39.400 45,000 lb. given above, the ealculated efficiency of the smallest—and of the best—weld, in shear, is 95.4 per cent. Table V—Shear Ultimate Calculated——, Weld Test Area, Strength, Strength, Per Cent Actual Weld No. Sq. In. Lb Lb Efficiency (Lb. per Sq. In.) Ss 4.23 190,500 142,500 74.8 33,650 6 3.60 162,000 131,110 80.9 36,400 9 3.18 143.200 120,000 83.8 37.730 10 2.50 112,400 96,400 85.8 38,600 11 09 94,000 89.650 95.4 42.900 Below is given an equation to show the length of bead required to develop the full strength of the plates used in the joint: AT L in which, S A is cross sectional area