506 



METALLl 



part of a nonluminous Bunsen flame on OP 

 the >tei-l having serxed to supjH.rt tin- n 



A el>luu-h l i n t, A n in steady use 



ha 

 to the com I. ined action of n high teinjM-ra- 



for two yeans, and ha exposed during i Imi 



. r com- ..f the flame. 

 As a result the steel had become so brittle that tin- 

 upl-T half crumbled as easily as chalk. The lower 

 leas exposed to the : .th a 



brittle Inyer. while .1 core was still 



\ sevenths of the earln>n seemed to 



have disappeared fmtn tin- brittle part. 



Hani, brittl.- uhu.-ir.-n castings n 

 ami annealed, according to (>. -t-n. 



-tli.- older iiicthiMl. 



used and described by Keaumur in 1 ;,',' ->r i.\ 

 ,t from the combined to the free or 

 graphite state. In the former methcxl. the one 

 most generally followed, the metal i- heated 



n at a bright-red heat for a considerable 



|>enod. the length of which Is ile|ieinlelit "M t he 



thickness of the bar to be annealed. It may al-o 

 be placed in lime, sand, or In.ne a>h in lieu of o\idu 

 of in>n. provided the atiiio>plu-re is used as tin- 

 source of oxycen. The irn must be free from 

 manganese and sulphur, or the elimination of t In- 

 carbon will be delayed. Tin- method by decom- 

 posing the carliide into graphite and free iron is 

 effected by more or less continued heating between 

 850 and 050 C. Iron castings produced l.y this 



1 an- |uite as soft as those obtained l,y the 

 Beaumur method. but <>n account of the graphite 

 present no bend can be obtained. Tin-re is no 

 limit to the size of the castings that may he mad*-. 

 a large piece taking quite as short a time as pieces 

 of smaller size. In another method of softening 

 white iron discovered by the author the ca- 

 are can-fully packed in' some nonoxidi/ing mate- 

 rial and heated t<> the tem|H-rature at which the 

 iron would solidify after fusion or about 36 C. 

 below the (Belting point of the metal. As a result 

 of this treatment the white iron becomes changed 

 into a steel containing 1 ' per cent, of combined 

 carbon, the remainder of tin- carbon being dis- 

 tributed through the metal in the form of finely 

 divided graph; 



From his investigations of the contraction and 

 4 1. formations of iron castings in cooling from tin- 

 fluid to the solid state, Mr. Francis Schumann de- 

 duces th. D> that the deformation of 

 prisms due to unequal contraction can be overcome 

 by providing counter-deformation in the pattern. 



the addition of auxiliary parts that can be 

 n-adily removed from th. In complex ma- 



chinery castings the design should U- so modified 

 sen as to result in the least differences in 

 the rate of cooling. Sudden changes in form cause 

 severe initial stresses, if not frac-t urea, and should 

 be rigidly avoided. Imperfectly proportioned 

 flange*, ribs, or gussets added to the main body of 

 A carting for the pur|K.se of increasing the strength 

 '. ' ' :..-.-.. .:-, , | of weak-in.. Hol- 



low cylindrical columns, although cast of 

 thickness and left in the mold till cold, may be- 



crooked by reason of the unequal rate of 

 wwn the upper and lower halves, due to 

 the currents of air nawing through the column and 

 cling: md.-r side ,,f the upper half after 



the core arbor is removed, which is usually don- 

 mngand while the cast ing'is still 

 red hot. This deformation i avoided by Mopping 

 the ends with sand immediately after' the with- 

 drawal of the core. - 'tention to the laws 

 of cooling and und proj*>rti 

 castings will result in inrrcawd strength and econ- 

 omy, besides the avoidance of annoying crooked 

 castings and mysterious breakdowns. ' 



Th- .dvantages claimed to accrue from 



.Minium and ferro-alumi 



HUH. ill a papt r !i\ .1. A. Mm 



in-n more" fluid, that 



renders hard r :hus in; 



are freed from hard -]>>{> and bi Mat it 



leasens the tendency of the metal t<> chill, ami in- 

 creases tin I the me! Mica! 



so >aid that while good >. 



made more fluid and benefihd to M-HK- d. 

 \et ti i_'. s ,,f tivating \\ith aluminu: 



iiio^i cxident with i \\liitc iron. It i- an 



Unquestionable !a t. .Mr. Mi-inmi-t/ >a\x. that the 

 addition of aluminum \ 



the <|iialit\ of tin- castings for t 1 Tin- 



general co'nrlusion from tests made i> that \\ith 

 white iron small additions of aluminum, si, 

 would be n-M-d in onlinarv foundry pra.-ti.c. in- 

 crease slightly the fluidity; \ of 'i per oent, of 

 aluminum am'l more dMnMHSJ the flun: 

 iron ; i decidedly le>> fluid b\ an\ addi- 



tion of aluminum. Aluminum appear- to : 

 the shrinkage if enough is added. I'.-i i 

 the most striku i- the 



increased tinn- during which aluminum-li 

 iron \\ill remain molten. This jir.'i 

 fluid longer is of direct usefulness in a foundry 

 win-re it is necessary to run a large number of 

 small when tin-re is usually much trouble 



in keeping tin- ordinary metal fluid, unless it was 



!:ot to start with. 'J'he gi 

 that the molten bath of iron with aluminum will 



fluid twice as long as one without it. An- 

 other consideration is that cleaner, more solid, and 

 softer castings are universally obtained, with a 

 large reduction in tin- percentage of defective cast- 

 In a paper n-ad by 1'n-f. (lay before tin- Amer- 

 ican Society of Mechanical Kngiiieers <.n tin 

 point of iron and steel, additional information is 

 given regarding the inciva>c of tin- ultimate 



_rth and of the elastic limit of those metal*. 

 A test piece of iron, which under ordirmn 

 was shown to have a tensile strength of fr.m 



"<KI pounds per square inch, was tested ju-t to 

 tin- yield IK. int. then allowed to rest free of stress 

 for two weeks, and then tested again. The \ield 

 point was found to be considerably higher. Tin- 

 test was then repeated at times from one to 

 weeks apart till ta in all had been made. 



each showing a further increase in the yield point 

 till it had been raised to 55,900 pounds i 

 inch, or considerably Wyoiul the normal ultimate 

 .strength of the iron. The elongation in ea< ' 

 except the last, was not carried beyond the yield 

 point. In the last test it was continued to rupture 

 without any incn-a>e in -tre-- beyond tin- \ield 

 point, the total elongation of the piece being only 

 1 J-.") per cent., or one half of the elongation of tin- 

 same iron wln-u tested in tin; ordinary manner 

 without re-tinir. In a test of a bar of -t.,.| tin- 

 elongation was carried to a point ron-id-rab; 

 yond the yield point, rested seventeen hours, tested 

 again when the yield point was found much h 

 tested twice again with intervals of tw.-nt;. 

 hours, when a still higher yield point was . 

 each time. The total strength of the bar was thus 



ierablv increased over that of tin- bar ' 

 in tin- usual way. but the elongation wa> reduced 

 one half. 



In hi* studies of the various aspects of the I 



_-th in iron and steel by n-a-on of use. .Mr. 

 Thomas Andrews },a* made exhaustive micro-cop- 

 ical. chemical, and physical examination 

 known age and condition of service on main 

 of railway, ami has obtained much valuable infor- 

 mation on the subject. In a lecture upon it lie 



