This paper presents the results obtained from the in situ observations of the crack nucleation and its propagation under tensile loading in HVEM. It was found that some slip bands appeared and slip steps formed at both ends of a slip band through the movement of a great deal dislocations. The crack is initiated at the slip step owing to the stress concentration. Once a crack has nucleated, it will propagate forward under the applied stress. The cracks propagate essentially in two ways: (1) during tensile loading, a large plastic thinning zone is formed near the crack tip. The crack propagates along the thinning zone with a discontinuous manner; (2) the main crack propagates by a mechanism of coalescence of microcracks in the vicinity of the crack tip. Additionally, it was noticed that the grain boundary is not probably the preferential sites of the crack nucleation and propagation in α-titanium.

The CoO inoculant composed of 20% CoO and 80% Al_2O_3 (fused alumina) is an effective nucleator to the cast nickel-base superalloy blades. After calcining CoO and Al_2O_3 powders under 1300℃ for 2h, CoO.Al_2O_3 spinel was formed on the surface of alumina particles. It has been detected by EMPA that metallic Co of fcc crystal structure deposites on the surface due to that the CoO.Al_2O_3 spinel reduced by certain active elements contained in the cast nickel-base superalloys at elevated temperature. This is checked with the relative stabilities of the oxides of Ti, Cr, Al contained in the alloy with CoO. This was also verified by the pure Ni blades for which there was no effect at all. Because of the similarity of the crystal structure and the lattice constants of the metallic Co and the superalloy, the Co particles can act as nucleators refining the grains. In production practice, if SiO_2 is precipitated in local area of inner surface of mould and so the inoculant particles were covered with it, the nucleation effect is markedly reduced. In order to prevent the metallic Co remelting back, lower pouring temperature is preferred; and to enhance the reduction of CoO.Al_2O_3, higher heating temperature of mould is desirable.

The phosphorus segregation in certain steels has been observed by means of metallography and EPMA. It was found that such a segregation may cause the intergranular fracture in steel which is confirmed by an Auger spectroscopic analysis. The coefficient of phosphorus segregation α seems to have a quantitative relationship with the deformation ratio θ of steels lgα=A(1/θ~(1/2)), where A is a material constant of working temperature and test condition. In addition, some influences of the local segregation of phosphorus upon the fracture toughness of steels were determined. The average J_(Ⅰc) and K_(Ⅰc) of normal specimens and ones with cracks through segregation region are 9.87 and 7.68 kgf/mm, and 477 and 419 kgf/mm~(3/2) respectively.

A systematic study has been made of the modification of Al-Si alloys with eleven individual rare-earth elements (La, Ce, Pr, Nd, Sin, Eu, Gd, Tb, Ho, Er and Y) and also mischmetal at a cooling rate of 70—80℃/min. The modifier was added into the alloys by the methods of both reciprocal reaction and electrolysis of corresponding binary molten salts of rare-earth chloride and KCl. From the metallographic observation, it was shown that Eu possesses the greatest modification ability, and La the next, and both of them are permanent modifiers. The modification ability of Ce, Pr, Nd and mischmetal are even weaker than that of La. It was also found that modification ability of rare-earth elements decreases rapidly with the decreasing of their atomic radii. The greatest modification ability of Eu may be referred to the leap of its atomic radius. Owing to the similar reason, Yb (not investigated) would be expected to be the another element possessing strong modification ability in the lanthanide. In the consequence of lanthanide, from Sm to Ho the modification ability is negligible. Finally, under the conditions mentioned above, Er and Y did not show any modification effect on Al-Si eutectic alloy at all. According to the information that the atomic radius of Ac (2.03(?)) is somewhat greater than that of La (2.02A) and the modification ability may become greater in the order Y-La-Ac of the periodic table, it is deduced that the radioactive element Ac may have even greater modification ability than that of Eu. Thus the strongest modifiers would be elements La, Sr and Ac set on an oblique straight line in the periodic table. Therefore, it is assumed that the modification ability of a modifier may be the result of a certain combination of the valence electron charge number, the atomic number and the atomic radius of the modifier.

The bainite transformation in a 30CrMnSi steel was studied by thin foil transmission electron microscopy. It was found that: (1) The lower bainite of twin rail mode or of side precipitation mode and the upper bainite of branch mode are present in the steel. The midribs occur in the lower bainite plates and the step-like fronts in the upper ones. (2) The determination of orientations not only between the cementite and bainite-ferrite matrix but also between the matrix and austenite has been carried out as:(010)_θ∥(1(?))_B, (1(?))_B∥(011)_A,(103)_θ∥(101)_B, (101)_B∥((?))_A thus the bainite carbide precipitation from austenite may be proposed. (3) The formation of lower bainite seems to be analogous to the second stage of tempering, i. e. decomposition of retained austenite. During this stage, the correlation among four phases: austenite, e-carbide, cementite and bainite ferrite may be possibly expressed as a whole by the following parallelisms:((?)1)_A∥(000(?))_e∥(103)_θ∥(101)_B,(011)_A∥(1(?)10)_e∥(010)_θ∥(1(?))_B,((?)1(?))_A∥((?)010)_e∥((?)03)_θ∥(12(?))_B Specific close packed planes and directions are parallel in all these structures.

The morphology of upper bainite in a 55SiMnMo steel under normalized condition as well as its transformation in tempering has been observed by optical microscopy, TEM, SEM and X-ray diffraction. After austenitising at 900℃ for 1/2h and A. C., the upper bainite resulted was composed of ferrite and 30—35% carbon enriched austenite platelet, however, no carbide appeared until tempering treatment. The tempering temperature for carbide precipitation was found to be 400℃, and the tempering products, so called "tempered bainite", consist of ferrite, carbide and 3—5% austenite. The rate of crack propagation da/dN of the steel increases with the increase of tempering temperature and reaches its maximum at about 400℃, accompanied with relevant change of fracture type from the quasicleavage to the cleavage.

The alteration of microstructure of a NiFeNbAl aUoy has been observed under various heat treatments. It was shown that the magnetic properties and other physical parameters are immediately dependent on its microstructure. The alloy of 79%Ni-Fe-8%Nb-1.2%Al already has made such properties as: μ_0=15—20 kGs/Oe, μ_m=100—150kGs/Oe, H_c=0.015—0.03Oe, B_s=5200Gs, ρ=82—85 μΩ/cm, HV=290, after treated by heating to 1150℃ for 2 h and 970℃ aging 2h. An application of the alloy on the magnetic recording may be available.

The behaviour of vaporization of S, Mn, Cu, Si and P in the carbonsaturated iron melt has been investigated with respect to the effects of smelting time and temperature, as well as pressure in vacuum chamber and specific surface area of melt in a vacuum induction furnace. It was shown that an arrangement in vaporizability order of these elements may be properly put and a maximum rate of vaporization for S, Mn and Cu occurs at 1330℃.

A thermodynamical equation of n-component (n≥3) systems with constant partial molar quantity (?)_0 was derived for interval between (?)_0~o and (?)_0, i. e.where and For such systems that obey Zdanovskii-type rule, (sum from n=k (x_k/x_(k(0))))_((?)_0)=1[1≤k≤(n-1)], one of the simple solutions of the equation may be:For certain systems that obey Turkdogan-type rule, (x_1/x_(1(0))+sum from n=j (x_j/K_jx_(j(0))))_((?)_0)=1 the solution may be:For example, if ψ represents volume, heat capacity or enthalpy: Δψ=0 when (n—1) binary systems of the same (?)_0 value were mixed up to compose a n-component system obeying Zdanovskii-type iso-(?)_0 rule; and Δψ=sum from n=j (K_(ψj)Y_j) when (n—1) binary systems of the 0 component were mixed up to compose a ncomponent system obeying Turkdogan-type iso-(?)_0 rule.If ψ represents Gibbs free energy: the activities of the k component in a n-component system obeying Zdanovskii-type iso-μ_0 rule may be written as Raoulttype equation: α_k/α_(k(o))=X_k; and the activities of component 1 and each component j in the n-component system obeying Turkdogan-type iso-μ_0 rule may be written as Raoult-and Henry-type equations, α_1/α_(1(0))=X_1 and α_j=K_j~*X_j respectively.Satisfactory verifications of the aforementioned results have been made for aqueous solutions, liquid alloys or molten salts.

According to ν=c_0(τ/τ_0)~m and the condition of dislocation source operation, the glide motion of a dislocation group emitted from a source under a constant applied stress, is simulated on a computer by using experimental data of isolated dislocation velocity in silicon crystals. The following conclusion is obtained:1. The velocity of a leading dislocation in a group emitted from a source is approximately 1.36 times the velocity of an isolated dislocation. This result agrees with the experimental ones obtained in silicon crystals.2. The average velocity of all dislocations in a group emitted from a source, is equal to the velocity of an isolated dislocation.3. In a dislocation group emitted from a source, the average density of all dislocations, is proportional to τ_a, and the rate of generation of dislocations from a source is found to be:dN/dt∞τ_a~(m+1) e~(-Q/kT).

The cell volume V~((0)) of fcc/bcc and fee/bet has been calculated for three orientation relationships. The orientation will be acceptable if its V~((0)) is larger than others, and everyone has its own range of lattice parameter ratio. Then the lower energy 0 lattice will appear when γ gets some particular value. It is suggested that the γ value can be changed locally by segregating or depleting solute atoms near boundary so that the density of dislocation on boundary can be decreased. The Zn segregation in α/β brass would be explained with this mode.It is shown that the magnitude and direction of tetragonal distortion affect the V~((0)) considerably. It will be expected to predict the reasonable distortion direction if the orientation and the lattice parameter are given. But the fact that the orientation relationship between austenite and martensite in Fe-C alloy changes from K-S to N-W can be explained satisfyingly if it is suggested that carbon atoms move to transformation interface along <11(?)>_f direction.

The composition of a continuous or a limited solid solution obeyed the Vegard theorem were successfully measured by two diffraction lines with single or double wavelength of X-ray. The relationship between the composition, x, and the diffraction angle difference, δ_x, for cubic crystal system may be deduced as:x≈(δ_x-δ_0)/(δ_1-δ_0) where δ_0, δ_1 and δ_x are the differences between two diffraction angles for x=0, 1 and x respectively. Certain typical examples were used to check the formula. In case of deviating from the linear relation, a correct term, Δ_x, was introduced into it, and Δx vs x plot was found to satisfy a parabolical equation. Thus, the formula may be modified as:x=(δ_x-δ_0)/(δ_1-δ_0)+Δx_(max)4(δ_x-δ_0)(δ_1-δ_x)/(δ_1-δ_0)~2 where Δx_(max) is the deviation at x=0.5, it eliminates the error induced from the linear approximation. Finally, the error of these two methods was discussed briefly. Because these methods depend only on the difference between diffraction angles, the error due to incorrect zero-position and specimen eccentricity can be eliminated, therefore, the methods seem to be simple and rapid, and to have proper applicability as well.

The simplified Phillips' differential equation of tension which derived from Hooke's law on the assumption of uniform strip section is conventionally used up to date for the computation of cold tandem rolling. However, the strip operated through the stands in rolling practice is not uniform in section. An improved empirical differential equation of tension especially for varying strip section has been proposed as:dt_(fi)=E/h_i(l_1/h_i+l_2/H_(i+1))(v_(b(i+1))-v_(fi)dτ under the consideration of influence of step disturbance or dynamic gage change. Based mainly upon this equation, it was made of a mathematical model for dynamic digital simulation to trace the position of the stepped section. Certain examples of the simulation computation on digital computer showed such marked error as it is still calculated by the equation of equal section. While a satisfactory representation of the dynamic process may be resulted by the equation aforementioned.

On the basis of the uniaxis tensile tests of several mitre welding joints, a mathematical treatment for calculating loading capacity of the butt welding joints under complex stress condition was worked out and an empirical formula is given as:σ_(eq)=(σ_w~2+σ_wτ_w~2)~(1/2) The theoretical consideration of this formula is easily approached by the plasticity collapse and the fracture mechanics. It is also clarified that the conventional understanding about the loading capacity in mitre welding joint being always senior to that in perpendicular one, and its related calculation, as well, seem to be not entirely correct.

The process of opposite martensite transformation in Ni32Co4Cu alloy on heating has been studied. The stabilization and super-stabilization of austenite may be sufficiently displayed by the phase change cold work for austenite in the opposite transformation. After the double deep cold cyclic treatment, an alloy of lower linear expansion coefficient and stable microstructure under low temperature was obtained, as well as it was available to improve its mechanical properties and machinability.

The binary A-15 phases formation and their T_c variation have been studied by means of the bond-parametric diagram. It was ascertained that the A-15 phases locate in a definite region on the plot of electron density e/a vs radii ratio of components r_A/r_B. The A-15 phase cannot form any longer for the ratio r_A/r_B over a certain critical value which varies with e/a periodically. On the plot, the T_c distribution of A-15 Phases is shown with regularity. The A-15 phases with T_c higher than 1, 4.2, 10, 15 or 20 K are concentrated in the regions of corresponding isotherms. The ratio r_A/r_B, and e/a as well, may be important parameters in determining T_c. As compared with Matthias' rule, a better result in determining the location of A-15 phases of high T_c has been obtained by the consideration of both factors.