On the basis of describing the classical nucleation theory of martensitic transformation and considering in detail the original derivation of nucleation probability through statistical physics by Fisher $et~al$, the present work points out the mistakes appearing in the recent works of some authors who negated the calculation of the nucleation probability of martensitic transformation with Cohen and Kaufman. The nucleation barrier ΔG* in calculating the nucleation probability should be the free energy change of total atoms in an embryo, rather than that of an atom. The experiments cited by these scholars, supporting their viewpoint of homogeneous nucleation of martensitic transformation, are re--explained and the nucleation model suggested by them is critically reviewed by us with the basic concept of statistical physics. Finally, several research directions are briefly suggested for study of the nucleation of martensitic transformation.

Hydrogen--induced delayed fracture during dynamic charging of TiNi shape memory alloy, and the role of atomic hydrogen, hydrogen--induced martensite and hydride in hydrogen--induced delayed fracture have been investigated. The results show that hydrogen--induced delayed fracture in TiNi alloy could occur, and the normalized threshold stress intensity factor decreased linearly with increasing the total hydrogen concentration, i.e., KIH/KIC=2.01-0.25lnCT. The content of hydride increased continuously during dynamic charging, and the fracture toughness of the TiNi alloy decreased gradually, which is the main cause of hydrogen--induced delayed fracture. The role of atomic hydrogen and hydrogen--induced martensite in hydrogen--induced delayed fracture was very small.

As [Si] time series of No.1 BF at Laiwu Iron and Steel Group Co. and No.6 BF at Linfen Iron and Steel Group Co. to be sample space, based on phase space reconstruction technology and G--P algorithm, the saturated correlative dimensions of chaotic attractor of Laiwu Blast Furnace Ironmaking Process and Linfen Blast Furnace Ironmaking Process are about 3.36 and 3.12, respectively. The existence of saturated correlative dimension with fraction value means there is chaos in both blast furnace ironmaking process.

Based on the theory of X--ray stress measurement, a method of measurement and representation for the inter--phase stress of double--phase material was established. The systematic error of the instrument has been effectively reduced by the difference between diffraction angles of two phases. The inter--phase stress of an actual material (SiCp/6061Al composites) was measured, and it was verified that this method has a high measuring precision.

Using rapid solidification (RS) and powder metallurgy (PM) technique,　 Al90Mn8Ce2 bulk alloy with ultra high strength and　20 mm in diameter and 10 mm in length has been prepared. The density of the bulk alloy is up to 99%. Their ultimate pressure strength　and fracture strain are 895 MPa and above 6%, respectively. It has　been found that the high strength is attributed to the metastable icosahedral quasicrystal phase embedded in the α--Al matrix, which has been confirmed to be still stable at 803 K under the　pressure of 1.2 GPa.

The effects of sputtering parameters on the microstructure of Ti--6Al--4V coating on SiC(f) prepared by magnetron sputtering were studied. The results show that the deposition rate is related to the working pressure at a given power density. The morphology of the deposited layer and its binding with the fibre are dependent on temperature. The composition of the coat is close to Ti--6Al--4V, and is uniform throughout the thickness of the coat. The deposited layer consists of crystallites with nonequilibrium hcp structure with grain size of 20---50 nm. The interface between SiC fibre and Ti--6Al--4V coating was investigated by scanning electron microcopy and energy dispersive X--ray spectroscopy. The examination reveals no chemical reaction at the fibre--coat interface which is in nature a sharp interface formed by bonding of the two substances.

The effect of constraint transformation heating temperature Th on the free reverse transformation behavior of 10\% strained Ti44Ni47Nb9 alloy was investigated, and the experimental results showed that after the first constraint transformation the strained Ti44Ni47Nb9 alloy exhibits two separate recovery strains εr1 and εr2 on subsequent free reverse transformation, which increases and decreases with increasing Th, respectively. However, the total recovery strain εr(εr=εr1+εr2)of Ti44Ni47Nb9 alloy decreases with increasing Th. The plastic strain of Ti44Ni47Nb9 alloy is increased by the deformation of martensite under the recovery stress during the constraint transformation, which decreases εr strain of Ti44Ni47Nb9 alloy. Since the martensite of strained Ti44Ni47Nb9 alloy is consisted of M1 and M2 oriented martensites with different strain levels after the first constraint transformation, strains εr1 and εr2 are related to M1 and M2 oriented martensite reverse transformations respectively.

The fatigue properties of high strength steel 42CrMoVNb with three fine grain sizes produced by different heat treatment procedures were studied. The experimental results illustrate that at 10 6---10 7 cycles regime, there is no horizontal asymptote in fatigue S--N curves for smooth specimens, so the conventional fatigue limit disappears. SEM shows the fatigue crack initiation sites in smooth specimens are related closely to fatigue life, and most of the sites are at inclusions. From analyzing the experimental data, the critical size of inclusion and the most appropriate size of prior austenite, which should be controlled in practical production, are obtained.

The directionally solidified Tb--Dy--Fe magnetostrictive alloys with <110> orientation were prepared in a velocity range from 30 to 720 mm/h, and the solidified morphologies change from planar, cellular to dendritic with increasing growth velocity. An obvious magnetostriction jump effect was observed in the sample with high growth velocity (>120 mm/h). Two kinds of twin growth mode, single--{111} and two--{111}--twin system were observed. The measured magnetostrictions of the samples are basically consistent with the calculated values by the above two twin growth mechanisms. As a comparison, the magnetostriction of <112> oriented crystals prepared with v=900 mm/h is corresponding to the <110> oriented crystals prepared with high growth velocities.

<110> oriented rods of a Tb--Dy--Fe magnetostrictive alloy have been successfully prepared by zone melting unidirectional solidification. A homogenization annealing for 4 or 48 h at 900---1100 C has been carried out in a quartz cylinder under Ar protection after pumping to 2X10 -3 Pa. It was found that, after annealing, <110> preferred orientation is unchanged, the net shape pseudo--eutectic rare earth phase trends discontinuous and spheroidal, magnetostriction slightly decreased without pre--stress but increased obviously with pre--stress. An obvious magnetostriction jump effect was observed in the annealed samples. A satisfactory magnetostrictive property of 1950X10 -6 was obtained under 15 MPa pre--stress after 4 h heat treatment. The compressive stresses of the heat--treated samples decrease monotonously with increasing the soaking time. A satisfactory combination of the magnetostriction and mechanical properties was obtained at annealing for 4 h for the <110> oriented Tb--Dy--Fe alloy.

The effects of alloying element Si on oxidation resistance of several γ--TiAl based alloys containing different Nb contents have been studied. The results reveal that the lower the Nb content of the alloy, the more obvious the beneficial effect of Si. The improvement of oxidation resistance by Si is mainly attributed to promoting the formation of continuous and dense Al2O3 layers in the binary alloy. In addition, more homogeneous and finer oxidation products appear on the surface of higher Si--bearing alloys than on lower Si--bearing alloys, leading to further improvement of oxidation resistance in Nb--containing alloys. As a result, the parabolic rate constant decreases with increasing Si content. Isothermal oxidation experiment, X--ray diffraction (XRD) and scanning electron microscopy (SEM) analysis of the oxidation products are presented.

The corrosion behaviors of pure iron and two commercial steels were examined in presence of ZnCl2--KCl deposits in controlled atmosphere containing HCl at 500℃. All of three materials suffered accelerated corrosion by forming thick oxide scales, and complex chlorides mixed molten salts appeared between the oxide scales and substrate materials. However, materials with higher Cr content have better corrosion resistance due to the formation of Fe--Cr spinel or chromia in the inner part of oxide scales. The mechanism for the corrosion accelerated by molten chlorides deposits is discussed.

By using a specially designed device and gas permeation method, the hydrogen diffusion coefficients for the tube--shape samples of Cr18Ni10Ti austenite steel were measured. The testing temperature covers the range of 360---600℃. The curves of hydrogen permeation flux and the relation of hydrogen diffusion coefficients and specific time were simulated by finite difference method (FDM). According to the simulated curves the experiment data related to the hydrogen diffusion coefficients can be treated. The hydrogen diffusion coefficients for aluminum film inner coated tubes at room temperature is 5.789X10 -20 m2/s, its activation energy of hydrogen diffusion is 8.144X10 -20 J/mol, which are 5---6 orders lower and 1 times higher than those of uncoated tubes respectively.

Ni--Co--Cr--Al--Y--Si--B coatings were deposited on Ni base superalloys DZ125 and DSM11 by arc ion plating (AIP). The corrosion behaviors of DZ125, DSM11 and Ni--Co--Cr--Al--Y--Si--B coating in molten 75%Na2SO4+25%K2SO4 at 900℃ were studied. The results indicate that the coating can protect the DZ125 and DSM11 alloys from hot corrosion destruction due to the formation of a continuous and dense α--Al2O3$ scale. In the last stage of the hot corrosion process, the hot corrosion behaviors of base alloys have important influences on behaviors of the whole systems.

An investigation on the growth of passive film on 304 stainless steel by using alternating voltage (A.V.) passivation was carried out by electrochemical impedance spectroscopy (EIS), coulometry measurement and atomic force microscopy (AFM). EIS was performed on 304 stainless steel electrode, which is covered by A.V. passive film, to identify the formation of pore as a function of higher potential parts ($E_{\rm h}$) in alternating square wave electric field. Coulomb--time curves were measured during the growth of passive film on different conditions. A new concept, probability of pore formation, which describes the behavior of the pores during the growth of film, was presented. Results indicate that the growth mechanism of A.V. passive film is different from that of D.C. passive film, which makes the A.V. passive films have a faster growth rate. Coulomb--time curves and AFM analysis confirmed the growth mechanism of A.V. passive film.

The Synergistic effects between erosion and corrosion of two high velocity oxy--fuel (HVOF) Ni--based coatings (Ni60JH and Delelo50) deposited on a low carbon steel in a liquid--solid two--phase flow were systematically investigated by a weight loss test method and an electrochemical method. And a stainless steel was compared to the two coatings and the substrate. The experiment results indicate: (1) At the erosion rate 5---10 m/s, the erosion--corrosion resistance of Ni60JH and Delelo50 coatings was 8---15 times and 3---5 times higher than that of the substrate, respectively. And the erosion--corrosion resistance of Ni60JH coatings slightly exceeded that of the stainless steel. (2) The weight loss rates of the synergistic effects between erosion and corrosion of the four types of materials changed from 45% to 80% of the total weight loss rates with the erosion rate increasing from 5 m/s to 10 m/s. The weight loss rates of the synergistic effects between erosion and corrosion were primary in the total weight loss rates of erosion and corrosion. (3) In the pure corrosion, the main corrosion characteristic of Ni60JH coating and stainless steel was the uniform corrosion, while that of Delelo50 coating and low carbon steel was the selective corrosion.

The effect of Ti on glass--forming ability of bulk (Zr0.59Cu0.18Ni0.13Al0.10)100-xTix (x=0, 1, 2, ...... 6) alloys has been investigated by differential scanning calorimetry (DSC), differential thermal analysis (DTA), and X--ray diffraction (XRD). It is shown that the liquidus temperature Tl decreases obviously when Ti content (x) increases from 0 to 4, meanwhile the reduced glass temperature (Trg) increases from 0.62 to 0.65. When x exceeds 4, however, Tl increases and Trg decreases respectively with increasing Ti content. The wedge--shaped sample with x=4 has the largest amorphous region, i.e., has the highest relative glass--forming ability, and the relevant supercooled liquid region ΔTx is 77.6℃. The alloy with this composition can be cast into cylinder--like amorphous alloy with 10 mm in diameter and 70 mm in length or sheet--like with 3 mm thickness by arc melting and copper mold casting. More intermetallics phases exist in the amorphous--crystallization zone of the wedge--shape sample with x=4 than those of alloys with x=0 and x=2 as indicated by XRD results. It is suggested that the crystallization would be retarded by the interaction and competition of various phases, thus enabling a higher glass--forming ability.

The glass formation under high--energy ball milling was investigated for a (Ti0.33Zr0.33Hf0.33)50(Ni0.33Cu0.33Ag0.33)40Al10 high--order alloy system. For comparison, the glassy ribbon with the same composition was prepared using melt--spinning (MS) method as well. Structural features of the samples were characterized using X--ray diffraction, ransmission electron microscopy and differential scanning calorimetry. Mechanical alloying (MA) results in a formation of glassy alloy similar to that obtained by MS. But, the glass formation is incomplete and a small amount of unreated crystallites smaller than 30 nm in size still remains in the final product. Like the melt--spun glass, the ball--milled glassy alloy also exhibits a distinct glass transition and a wide supercooled liquid region of about 80 K. Crystallization of the glassy alloy is a two--step p rocess, regardless of the synthesis routes. After the primary crystallization is completed, the remaining amorphous phase shows a detectable glass transition and a large supercooled liquid region of about 100 K.

The concentration field and temperature field of solidifying crystals in the processing of semi--solid metals were analysed under stirring condition. By using solidification velocity and flow speed of the melt as variations of affecting the shape of crystals, a shape determination model for grain growth was developed under the condition of moving liquid. The morphologies of grain growth under different conditions of semi--solid processing were simulated. The simulation results are in accord with the experimental results.

A dissolution model of the plate--like and spherical second phases in the multiphase alloy during continuous heating process with constant rate has been proposed. The effects of heating rate and interface curvature on the concentration profiles in the matrix and the dissolution rate of the second phase were discussed. The model was applied to calculate the dissolved fraction of the second phase before melting during heating process. The calculated results for Al--5.8%Cu alloys, where the eutectic melting fraction of the non--equilibrium eutectic was determined by DSC analysis, agree well with the experimental ones, clearly indicating the validity of the developed model.

Based on the physical process of nucleation and growth of grains and basic transfer equations such as heat and solute transfer equations, a mathematical model for the three--dimensional simulation of microstructure, microsegregation and free dendritic growth of single phase alloys was developed. Many factors such as constitutional undercooling, curvature undercooling and anisotropy, which have vital influence on the evolution of microstructure, were considered in the model. Simulated results show that the nucleation, growth, CET and microstructure evolution of free dendritic crystal could be predicted reasonably and the calculated results are coincident with actual phenomena. Simulations of free dendritic growth indicate that the curvature undercooling has a significant effect on the dendritic growth and final microstructure pattern. The dendritic grain profiles, in which a great number of regular and parallel secondary dendritic arms exist, tend to be formed at high intensity of anisotropy. At the low intensity of anisotropy, however, near octahedral grain profiles with small protuberances of surface tend to be obtained. The existence of small molten pools in interdendritic areas during solidification was confirmed by the simulated results, which is helpful to understand the formation of microstructure related defects such as microsegregation and microporosity.

The calculation of single crystal magnetic induction coefficients from measured texture and magnetic induction data in a grain--oriented silicon steel strip has been discussed. A new method, based on an accurate reliability evaluation of measured magnetic property data and the fitting of texture components, was proposed. It was also shown that, instead of the experimental texture data obtained with X--ray diffraction technique, the fitted texture coefficients according to a Gaussian distribution model should be used in calculation.