Sodium sulphate is formed from pyrite and other sulphur compounds in theraw materials, such as bauxite and limestone, and in fuels, such as coal and fueloil employed during sintering, causing higher soda losses of the process and givingtrouble to a number of operations, for instance, the ring-formation in the ro-tary kilns. The addition of anthracite, as a reducing agent to the raw mix for thekiln-feed, will reduce most of the sulphate into ferrous sulphide, which wouldeventually be removed from the process with red mud. This method has been suc-cessfully employed in alumina plants of soda-lime sintering process and Bayersintering combination process in China since 1961. Plant operating results as wellas experimental data, together with brief discussion are presented.

The method of shotcrete for roadway and shaft support has been developedin no less than fifty underground metal mines in China, but so far a theoreticalanalysis is lacking. Since the deformation on the support and that on the contact plane of rockmass are identical and at the same time taking into account the reaction of rockmass on support, the correlation of stress-strain and displacement of rock mass inroadway and shaft of circular cross-section under condition of plane strain vary-ing with time may be viewed on a rheological model of rock mass. Hence, thecorresponding expression for the distribution of stress and strain over the rockmass and the equation for calculating stress-strain and displacement of shotcretesupport are worked out. The formulae of thickness and strength in various desi-gnings required for shotcrete thus obtained are reliable for engineering practice.

Results of the beneficiation of the diaspore-kaolinite-bearing bauxite by flota-tion, from laboratory and small-scale continuous processing to pilot plant testingare described. Using raw materials with initial Al_2O_3/SiO_2 ratio of 4.6-5, theflotation concentrates finally attained a ratio greater than 8 with a recovery of Al_2O_3amounting to 70-80%. Results of digesting test for producing alumina by Bayer process, as well astests for making comprehensive utilization of the flotation tailings are also presen-ted.

Martensitic transformation of a 18-8 austenitic stainless steel cathodically charg-ed of hydrogen has been investigated by means of optical microscopy, scanningelectron microscopy and X-ray analysis. It is shown that hydrogen promotes thetransformation of martensite and the α′-martensite is formed next to the ε-marten-site, accompanied by the formation of surface microcracks. The effect of hydrogenon crystal defects has been regarded to be responsible for the aforementionedphenomena in such steel.

An approach to further investigation of the recent understanding on marten-site transformation has been developed after reviewing the relevant previous works.In particular, the physical meaning of the following formula has been clarified: ΔG~(γ→M)=ΔG~(γ→α)+ΔG~(α→M)The concept of estimating and the means of calculating ΔG~(α→M) have been dealt with,and the method for treatment has been refined. Calculation of ΔG~(γ→α) has been madewith certain newer data. The theoretical M_s value given by direct thermodyna-mical treatment agreed fairly well with that obtained by typical experiments. It isshown that the M_s of pure iron is 800K, and a linear relationship exists between M_son one hand, and the carbon concentration of the alloy or the yield strength ofaustenite at M_s on the other. The driving force of phase transformation increaseswith the increase of carbon concentration and with the decrease of M_s.

The tensile strength, ductility or impact value of a wrought nickel base super-alloy Ni-15Cr-6W-3Mo-2Al-2Ti depend markedly upon the silicon content of thealloy. On plotting these mechanical properties vs the silicon content which rangesfrom 0.1 to 0.89%, a saddle shown by the existence of a minimum at about 0.4-0.6% Si occurs. There is reason to believe that such a saddle behavior may be caus-ed by the variation of mechanical properties with the type, amount and morpho-logy of carbides as well as with the sequence of carbide precipitation in the alloyas influenced by its silicon content. Thus, the amount of M_6C and M_(23)C_6 precipi-tated along the grain boundaries would appear to proceed in four stages as follows: 1. For silicon content up to 0.1%, globular M_(23)C_6 may be the sole type ofcarbide formed; 2. In the range from 0.1 to 0.4% Si, besides M_(23)C_6, M_6C begins to appearand increases with increase of silicon content. The morphology of the carbidesseems to be different, however, being in a discontinuous, blocky form; 3. In the range from 0.4 to 0.6% Si, continuous films of M_6C were present.As might be expected, this type of carbide is very detrimental to the tensile andimpact properties at room temperature. The precipitation of M_(23)C_6 was slowed downor even suppressed; 4. In the range from 0.6 to 0.89% Si, M_6C may be altered to discontinuous,blocky form, whereas M_(23)C_6 may be very much limited. In conclusion, the detrimental effect of carbide is mainly associated with itsfilm-like morphology especially along the grain boundaries. Of course, carbides ingranular form are good strengtheners. According to prevailing view, criterion for the formation of either M_6C orM_(23)C_6 in nickel base superalloys would seem to be the relative Mo and W con-tents, e. g. if Mo+1/2W>6%, M_6C will be formed. Present work indicates that si-licon plays as equally important role in the formation of M_6C besides the Mo andW contents of the wrought nickel base superalloys.

The structure of amorphous Gd-Co films 400-700Ain thickness has been stu-died by electron diffraction and results of the radial distribution functions (RDF)have been obtained for five samples. In comparing the RDF for two of the samp-les with perpendicular and in-plane anisotropy respectively, it is shown that cor-relation between the nearest neighbour coordination of different kind of atoms inthe amorphous Gd-Co films and its magnetic anisotropy may be established.

The crystallization process of Gd_(17)-Co_(83) sputtered thin films has been exami-ned by electron microscopy and electron diffraction. The samples showed a typicalamorphous feature at room temperature and a crystalline state under 550℃. It wasobserved that the crystallization process of Gd_(17)-Co_(83) films underwent a change fr-om an amorphous state through two steps of metastable phase Ⅰ and Ⅱ, and theninto a stable state finally. The metastable phase Ⅱ, however, was not a single ph-ase, but was a mixture of several complex metastable phases. The influence of tem-perature on the domain structure of amorphous Gd-Co thin films has also beenstudied.

The stability of SmCo_5 permanent magnet prepared by liquid phase sinteringwas found to depend substantially on its intrinsic coercivity and hysteresis loopsquareness. A set of curves concerning initial irreversible loss, irreversible lossduring long-term(up to 10000h) aging and total loss (including both reversibleand irreversible) under certain elevated temperatures (e.g.250℃) against H_(ci) andH_k has been plotted.

The microprocess of failure by ductile fracture under tensile stress in speci-mens of two carbon steels containing 0.15% and 0.45% C respectively and of a25Cr20Ni type stainless-steel has been examined in-situ in SEM. Cracking wasnucleated preferentially in three different ways, e.g., at the inclusion/matrix, atthe twin/matrix boundary and at the σ-phase/matrix interface. In both the firstand second cases, the cracks were formed at the inter-faces and widened in thesteel matrix along the direction of applied stress. In the third case the crackswere nucleated at the σ-phase/matrix interface and grown into σ-phase particle.As the applied stress became greater and greater, other oracks nearby appear tocoalescence along the shear direction of the steel with the result of total failure.

The impact fracture of a medium carbon (0.5%C) steel containing boron(0.0005-0.0050% acid soluble boron) has been examined by SEM after various heat-treatments. It was found that shellshaped fracture occurred only if Fe_(23)(C,B)_6precipitated at austenite grain boundries was in large quantity and with particlesize greater than 2μm, the corresponding impact values of such specimens wereextremely low. Such specimens could not be remedied by the usual quench andtemper treatment, and re-solution treatment at high temperatures (980-1050℃)followed by rapid cooling to room temperature was necessary to modify the sizeand distribution of the boron-containing phase, with the consequence of improvingproperties and eliminating the shell-shaped fracture. Similar fractures were alsoobserved in a 40MnMoB steel.

The extention of the K_Ⅰ-K_Ⅱ composite mode (K_Ⅱ/K_Ⅰ=0-14) cracks has been inves-tigated by means of three-point and four-point bend specimens of high or mediumstrength steels as well as of nodular cast iron. All specimens were broken withbrittle fractures under linear elastic plane strain conditions, and the extending di-rection of the cracks appeared to agree well with the three existing theories, e.g.,the theory of maximum stress, the theory of strain energy and the theory of ener-gy releasing rate. The crack extension resistance, however, was found to increasewith the increase of K_Ⅱ-K_Ⅰ ratio, thus deviating obviously from the theoreticalprediction. This finding has been investigated with regard to the following aspects,e.g., the configuration and the size of the plastic zone, the state of stress at cracktips, etc. and in addition, the assumption that the crack extension resistance isindependent of the deformation characteristics was found to be untenable for speci-mens undergoing plastic deformation. Attempts have also been made to give a ten-tative discussion on a simple and reliable analytical method to deal with composi-te mode cracks for practical purpose.

A modified model of double cantilever beam in determining K_(IC) has been pro-posed. In our model the behaviour of the double cantilever beam specimen ischaracterized by a constant stress intensity factor on condition that cracks in thespecimen propagate under constant displacement conditions. Results obtained froma 4340 steel appear to be in reasonable agreement with the established data.

The yield strength of both ends of a hot-rolled 10Ti steel roll was found tobe about 10kg/mm~2 higher than its middle. Various factors possibly contributableto such an inhomogeneity in mechanical properties have been considered andsupported by experimental investigation. It is concluded that size and distributionof TiC particles are the main cause of the difference in yield strength betweenthe ends and the middle of the roll.

The model of rolling force coefficient Q_P has been studied. Three trials oftest showed that macroscopic inhomogeneous draft may occur in the cold rollingof sheets and strips if the parameter l′/h is less than a certain value. The coeffi-cient Q_P given by "slab theory" does not seem proper, and should be modifiedby an amount of ΔQ_P∝(l′/h)~(-1). Three programs of Q_P model are suggested: Ⅰ. for different regions of the parameter l′/ha) l′/h<4.2, Q_P=1.7025-0.0402(R′/H)~(1/2)-0.1972γ(R′/H)~(1/2) l′/h≥4.2, Q_P =0.9599+0.0130(R′/H)~(1/2)+0.0254γ(R′/H)~(1/2)b) Q_P=1.08+1.79μγ(R′/H)~(1/2)-1.02γ l′/h<4.2, μ=0.0917+10.6017/(ε)~2 l′/h≥4.2, μ=0.1059-0.0005ε Ⅱ. taking into account both homogeneous and inhomogeneous deformation Q_P=0.6859-0.4962γ+0.0099(R′/H)~(1/2)+0. 1025γ(R′/H)~(1/2)+1.298h/l′ or Q_P=0.6903-0.5025γ+0.0572l′/h+0.0186γl′/h+1.29h/l′ Ⅲ. taking into account the output elastic recovery contributing to roll-force a) Q_P~*=0.4364-0.1350γ+0.0229(R′/H)~(1/2)+0.0868γ(R′/H)~(1/2)+1. 54h/l′ b) Q_P~*=1.08+1.79μ~*γ(R′/H)~(1/2)-1.02γwhere μ~*=0.0356/(l′/h-2.503)+0.0559 The parameter Q_P~* seems to be more suitable for the mill of most productsbut the relevant computation is more complicated. The physical idea of the equations in program Ⅱ is explicit and the roll-forcethus estimated has been found to agree very closely with the experimental data.It is recommended for the rolling of either cold or hot strips as well as heavyplates. From the results of off-line simulation, however, the accuracy of the modelmay be substantially improved by the combination of the adaptive control to latterpasses of a coil with that to the same pass of the mext coil using the programof the exponential smoothing method, the error can be reduced to within 10%.

The error in the X-ray stress measurement caused by the side declining angle, γ,of the counter scanning plane is shown to be: Δσ_X=-(σ_Ysin~2γ+τ_(XY)csin2ψ_i)in which D_i=∑sin~2ψ_i-nsin~2ψ/(∑sin~2ψ_i)~2-n∑sin~4ψ_i; n is the number of ψ. Such a relation isbeing supported by experimental results obtained in our laboratory. The error in stress measurements along axial direction of a cylindrical surfaceand certain problems encountered in the X-ray stress measurements of anomalous-ly shaped machine parts are discussed based on the above formula.

The activity of RE_2O_3 in both liquid RE_2O_3-CaF_2 and RE_2O_3-CaO-CaF_2 slagsat 1600℃ has been determined by equilibrating the slag with liquid tin in a gra-phite crucible under one atmosphere of carbon monoxide. Pure solid RE_2O_3 andpure liquid RE metals were chosen as the standard states and the equilibrium tem-perature was closely controlled to within ±2℃. The activity of RE_2O_3 in the re-action (RE_2O_3)+3C?2[RE]_(Sn)+3CO↑ was found to increase with the increase ofthe concentration of either RE_2O_3 in the RE_2O_3-CaF_2 system or RE_2O_3 and CaOin the RE_2O_3-CaO-CaF_2 system.