Concluding this series on thermodynamic study of the Fe-C system,the presentpaper deals with an analysis of the metastable iron-cementite equilibrium on thebasis of the activity data derived previously by the author and the free energy ofcementite recalculated by the author in this paper.It has not been possible before to calculate the cementite eutectic temperaturedue to lack of data on activities in the liquid iron-carbon system.With the aid of the results of calculation in a previous paper,this temperature has been found tobe 1148°,or 5° below the graphite eutectic temperature.The cementite eutectoidtemperature 723°,determined by Mehl and Wells,is found to be consistent withthermodynamic requirements.Darken and Gurry chose 1146° as the cementite eutectic temperature andwith this value and certain other data calculated the free energy of cementite as afunction of temperature.On the basis of the more rigorously calculated cementiteeutectic temperature 1148° and the carefully determined cementite eutectoid tempe-rature 723° and certain other data,the free energy of cementite has been recal-culated by the author and found to be considerably different from Darken andGurry's results.The solubilities of cementite in austenite and in liquid iron respectively arethen calculated as functions of temperature.It should be pointed out,however,thatthe solubility of cementite in liquid iron thus calculated is only a first approximationand will certainly be revised in the future when more data on the free energy ofcementite at higher temperatures become available.A complete Fe-C equilibrium diagram including both stable and metastableequilibria has been constructed in the light of the results of this and a previouspaper.It is easily seen that certain important differences exist between the conven-tional and the revised diagrams.

In this paper,the activity coefficient of MnO in liquid blast-furnace-type slagshas been evaluated as a function of the basicity ratio,calculated according to a newformula previously proposed by the author,with reference to pure solid MnO as thestandard state.The data of Stukel and Cocubinsky,and Filer and Darken havebeen used as the basis of this calculation.The results of this paper together withthe α_(CaO) values evaluated in a previous paper are discussed from the standpoint ofionic nature of slags.

Bath reactions during the oxidizing period of electric-furnace steelmaking withinjection were studied. It is found that,under favourable conditions,a rate of carbon removal ofabout 0.015—0.020%C per minute(that of the normal smelting with ore beingaround 0.005% C per minute only)may be obtained;(FeO)and[O]decreasesignificantly and approach to equilibrium with[C]during the oxidizing period,andthe decrement of[O]increases proportionately with the increase of the rate ofcarbon removal.Dephosphorization during the oxidizing period is found to be rather ineffective,the average[P]removed amounts to about 0.004% only.With a supplementaryoperation of dephosphorization during the melting period,[P]can be reduced toless than 0.015% at the end of oxidizing period.However,the dephosphorizationmechanism in the case of the oxygen smelting is still not yet clear.

It was found that the addition of a minute amount of boron(0.0005-0.005%)as an alloying element can greatly increase the hardenability of steel.So it wasrecognised that boron can be used as a substitute for the more expensive alloyingelements such as nickel,chromium and molybdenum,and greatly lowering the costof steel production.The aim of this investigation was to find a substitute for steel 18 ХΓТ,in orderto save the expenditure of the alloying element chromium.By comparison of hardenability properties and core properties of the carburizedboron steel and the carburized 18 ХΓТ,a carburizing grade boron steel was finallychosen,which contained less than 0.30% chromium,0.30% nickel and 0.15%molybdenum.The boron hardenability factor decreases with increase of carbon content;itwas found in the present investigation that a linear relationship exists betweencarbon content and boron hardenability factor,expressed asF_B=1+1.88(0.97-C%).In order to obtain the best properties of the carburized boron steel,the maxi-mum carbon content of the carburized surface should be maintained below 1%.

Laboratory melting of low carbon boron-molybdenum steels was successivelycarried out in high frequency furnace.It was found that the steel after forging andnormalizing treatment,a maximum stress of more than 60 kg/mm~2 and a yield stressof more than 48 kg/mm~2 can be obtained.The value of yield stress is more thandouble the yield stress obtained from unalloyed low carbon construction steels andthe steel at the same time retaining good ductility and toughness.The welding pro-perry of the boron-molybdenum steel was found to be good and this steel is recom-mended to be used as a high strength low alloy construction steel.The effect of boron in increasing the mechanical strength of 0.50% molybde-num steel was found to be due to the alteration of the carbide and ferrite dispersionin the microstructure after the normalizing treatment.High maximum and yieldstresses can be obtained from a microstructure consisting of a mixture of fine disper-sed pearlite and acicular ferrite,and the microstructure of the boron-molybdenumsteel is markedly influenced by the normalizing temperatures and cooling rates.By studying the effect of boron on the decomposition of austenite,it was found thatboron greatly retarded the formation of proeutectoid ferrite and pearlite,thus in-creasing the hardenability of the steel.Thus even at relatively slow cooling rates,a high mechanical strength can be still obtained from the boron-molybdenum steel.Calculation according to empirical formulae from the isothermal transformationdiagram of the boron-molybdenum steel shows that at a normal air cooling rate,the maximum diameter of the steel bar which can be used to obtain high me-chanical strength must not exceed 75.2 mm.

The effect of hot working on the graphite formation in 1.1—1.2% carbon toolsteels during subsequent annealing was investigated.Specimens were heated in air to1050℃ for 10 minutes,cooled in a second furnace to the forging temperature,forged rapidly to give reductions in height between 0—40%,air-cooled,and annealed for16 hours at 700℃.Graphite was determined by chemical analysis.It was foundthat:1.Austenitizing in air between 870°and 1050℃ followed by air cooling causedthe formation of graphite during subsequent annealing.The amount of graphiteincreased with increasing austenitizing temperature.Precipitation of graphite occurredalong the austenite grain boundary.2.Hot working at various temperatures increased the tendency of graphitizationduring subsequent annealing.The amount of graphite formed was smallest whenthe specimens were forged at 870℃ and increased with increasing or decreasingtemperature.3.Slow cooling after the specimens had been heated to or hot worked at hightemperatures reduced the degree of graphitization,provided that free cementite wasabsent during hot working.4.Forging during continuous cooling removed the effect of previous heatingor hot working until the temperature was sufficiently low to produce free cementiteduring forging.Forging during cooling gave minimum amount of graphite whenthe forging was interrupted between 820°and 870℃.A possible explanation to account for the phenomena is suggested.

This paper demonstrates a method of obtaining the surface distribution offriction components for plate rolling from the“rule of gradient”and experimentalpressure surface.An example was worked out.It appears that the surface dis-tribution of friction components is continuous except at the no-slip point,which isthe singular point for the case of slipping friction.The surface thus obtained serves as the boundary condition for the threedimentional problem of plate rolling.This result appears to be the first twodimentional distribution of friction components for plate rolling,and this methodis applicable to other cases such as forging.