42CrMoVNb steel with ultrafine prior austenite grain, its size as fine as 2 m, was obtained through rapidly cycle heat treatment. The effect of ultrafine grain size on delayed fracture behavior is studied using sustained load tensile test with notched specimen. It is shown that both strength and delayed fracture resistance of 42CrMoVNb steel increase when its grain size is refined from 20 m to 4 m, but they do not increase when refined to 2 m. When the steel was tempered at high temperature, its fracture characteristic at crack initiation area changes from intergranular to transgranular when its grain size is refined from 20 m to 8---4 m, while the fracture characteristic is intergranular when its grain size is refined to 2 m. Its fracture characteristics are all intergranular fracture for all the grain sizes investigated when tempered at low temperature. The reason for this kind of delayed fracture behavior is discussed mainly from the stress concentration and segregation at grain boundary.

Objective functions on minimizing the atomic volume elastic strain and the atomic size misfit in close--packed directions due to the atomic substitution in sublattices of r’ phase are established, respectively. Constrain conditions on equilibrium partition of alloying elements and on their concentration range in sublattices of r’ phase are correspondingly built up. Using superalloys CMSX--2 and SRR99 as objects investigated and adopting a compound optimization calculation method with the Powell and the Simplex algorithms, the elemental concentration and the site fraction of the two sublattices can be calculated. In addition, a calculation formula on lattice parameter of r’phase closely associated with the elemental concentration and with the site occupancy of the sublattices is derived. All the calculated results by use of the present methods show quite good accordance with the reported values. The method proposed can predict the elemental concentration in sublattices and the lattice parameter of r’phase when the chemical composition of the r’ phase investigated is known.

Surface relief effect and their corresponding microstructure accompanying bainitic transformation in the steel Fe--0.33C--3.0Mn--2.15Si--1.25Ni (mass fraction, %) were studied by atomic force microscope (AFM). The results show that surface relief of bainite is produced by bainitic ferrite solely without the involvement of austenite. Furthermore, surface relief of a single structural unit is tent--shaped rather than invariant--plane--strain--typed and the tilt angles of the two sides are 9--10 and 8--9, respectively close to the magnitude expected from the diffusional ledgewise growth mechanism.

High temperature tensile creep behaviors of cast nickel—based superalloy K44 in the temperature range of 850---900℃ and under the applied stress range of 225---380 MPa have been studied. The results indicate that all of the creep curves of test alloys have similar shape: a short primary creep and a dominant accelerated creep stage, and the long accelerated stage is due to the interaction between rafting r’ particles and dislocations. As the creep strain is a softening stage, the accelerated creep can be approximately described by expression. The creep fracture data follow the Monkman--Grant relationship. Cracks originate from the cavities at grain boundary or interdendritic.

The effect of applied stress intensity factor on the electric field induced delayed fracture of PZT 5 ferroelectric ceramics in silicon oil, has been investigated using single edge notched specimens poled along the longitudinal direction. The results show that the critical electric field for instant fracture in silicon oil decreases linearly with increasing the applied stress intensity factor. The threshold electric field for electric field induced delayed fracture in silicon oil decreases linearly with increasing the applied stress intensity factor. The results indicate that there exists a coupling action between stress, electric field and environment on the delayed fracture of the ferroelectric ceramics.

Interatomic potentials of the embedded atom (EAM) type and molecular dynamics simulation are used to calculate the surface energies of the high--index surfaces containing the [001] or [-110] zone axis in Al, Cu and Ni. Two empirical formulas are developed based on structural unit model for high--index surfaces. The calculation result shows these formulas can be used to give an estimation of the energies of the high--index surfaces. The closest packed surfaces have the lowest surface energy and the surface energies of the closest surface (111) surface and the next closest surfaces (110) and (100) surfaces are the extremum on the curve of surface energy versus orientation angle. Both the theoretical simulation results and the empirical formula calculation results consist with the available experiment data.

The hardness, fracture toughness and the critical stress for crack initiation of single crystalline nanobelts have been measured using nanoindentation method. The result showed that a sudden depth excursion occurred in the load--depth curves which corresponds to initiation of an indentation crack. The critical stress for the crack initiation. which is one order of magnitude less than that of the other bulk brittle materials. The average values of the hardness and modulus are H=6.25 GPa and E=86.7 GPa, respectively.

A model to describe dislocation patterns based on cellular automaton and molecular dynamics is presented, which takes into account the short—range interaction and long--range interaction between edge dislocations with opposite Burger's vectors on the same glide plane or the different glide planes. This model uses molecular dynamics to deal with long--range interaction, and cellular automaton to describe the short--range interaction. Using the model, the dislocation structures of single crystal Cu are simulated under no external stress and applied cyclic stress.

A model responsible for the magnetic--induced rotation of single anisotropic grain in free medium has been proposed firstly based on the classical dynamics of rigid body, in which the grain is treated as a prolate spheroid. A theoretical expression was derived for the alignment time of grain as a function of viscidity of liquid matrix, aspect ratio of grain, difference of anisotropic magnetic susceptibility of grain and applied magnetic intensity. It was applied to predict the alignment time of paramagnetic superconductor material Bi--2212 and ferromagnetic material Bi--3%Mn in melt processing. An experiment was conducted for the latter melted fully at 300℃ then rapidly quenched under different applied magnetic field, and the MnBi texture is obtained. It is found that the alignment time is less than one second, in good agreement with the theoretical prediction.

The quasi--steady state grain size distribution in the 2D grain growth process obtained by phase field method can be described well either by Weibull function or Louat distribution function, since the plots of two functions are exactly the same or very similar to each other under certain conditions. The grain side number distribution also achieves self--similarity or the quasi--steady state finally, however, it approaches a lognormal distribution instead. The relationship between the average number of grain sides and grain size appears to be nonlinear. These results agree well not only with those of high purity Al foil but also those obtained from our Monte Carlo simulation.

Phosphorus is hardly dissolved in the solidified Ni3Al solids,and tends to be segregated in the residual liquids of Ni3Al. The phosphorus--bearing phase is precipitated at the end of solidification. The segregation of phosphorus in the residual liquids inhibits the nucleation of Ni3Al grains and the formation of the equiaxed grains at the central part of the sample, and favors the growth of the columnar grains towards the central part of the sample. The reason for phosphorus to decrease the nucleation rate of Ni3Al is that it lowers the difference of free energy between solids and liquids, and increases the activation energy for the atoms to transit through the liquid/solid interface. Phosphorus lowers the final solidification temperature of Ni3Al, increases the length of mushy zone, inhibits the feeding of casting, and hence increases the tendency of micropores formation.

The dissolution kinetics ordered precipitations in a disordered matrix during retrogression heat treatment was investigated on atomic scale using computer simulations based on microscopic diffusion equations (Langevin equation) with the discrete format. The evolutions of atomic pictures and order parameter profiles with time were analyzed, and the retrogression mechanisms of Al--Li alloys were further studied. In the course of retrogression, it is firstly proved that the evolutions phase in a disordered matrix is in the order; and then the retrogression of alloys at metastable field tends to the anti--process of the phase precipitation of alloys at unstable field, while retrogression of alloys at unstable field tend to the anti--process of the phase precipitation of alloys at metastable field.

Combing Euler framework for flow fluid and Larangian framework for particle motion, a statistic model coupling the motion, coalescence and removal of inclusion in molten melts has been developed to interpret the basic behavior of inclusion in continuous casting tundish. Numerical calculation was conducted for 3D turbulent flow field using turbulence model, then the removal efficiencies and growth rate of inclusion were statistically computed based on the random--trajectory model. The results indicate that the total removal efficiencies of 10, 20 and 30um inclusion are approximately 20%, 36% and 75% respectively, of which the attribution due to adhesion to the refractory of inclusion occupies 1/6--1/4. It is found that the growth of inclusion due to coalescence is not marked, restricted by the realistic condition in tundish.

A mass transfer model has been built up for metal ion implantation into Al target at elevated temperature, based on the transport of ions in matter and the radiation enhanced diffusion theory, which is applicable to calculate the concentration--depth profiles of the implanted species. With the model, the ion implantation process at elevated temperature was simulated by the Monte Carlo method and the local saturation behavior in the crystal target simulated using a maximum allowed atomic fraction. Moreover, the diffusion process was described with the radiation enhanced diffusion theory. Thus the concentration--depth profiles of the implanted species were determined from the diffusion equations for the implanted species and nonequilibrium vacancies, and the radiation enhanced diffusion coefficient was obtained by taking into account the linear annealing defects. The nonequilibrium vacancy source function and the surface sputtering effects were introduced in the diffusion equations. The calculated concentration-depth profiles of Cr ions implanted into Al are consistent with the experimental ones at the target temperatures of 250, 400 and 510℃, with an ion energy of 140 keV.

In order to precisely describe the dendrite evolution during solidification process, especially in the micro--scale, a continuous model is adopted to solve the discontinuous physical properties between solid and liquid phases. In this model the physical properties in the interface zone are the average physical properties of solid phase and liquid phase, which can smooth their properties gap and make the properties continuous from liquid to solid phases. In addition, the averaged solute concentration is used to keep the solute conservation in the local interface zone. The simulated results show that this model can deal with the properties gap between the solid and liquid phases in the interface zone, and further simulate the dendritic growth, local interface instability, and appearance of the second dendrite arm, tertiary dendrite arm and the pattern of micro--segregation.

W--20%Cu (mass fraction) nanocomposite powders were synthesized by means of combining spray drying and hydrogen reduction process. The composite powders were characterized by using XRD, SEM, TEM. It has been shown that tungsten and copper are homogenously mixed within each particle and the average particle size is about 240 nm, in which the tungsten particles have an average grain size of 16 nm.

By means of r--ray attenuation method, the density of pure In and its relationship with temperature were measured in both solid and liquid states. The results show that the density data of pure In obtained by this method have a good repeatability within heating run and cooling run each other.

Prealloyed powders of La55Al25Cu10Ni5Co5 and its mixtures with tungsten particles were subjected to a high—energy ball milling. Structural evolution in the milled products was characterized using X--ray diffraction, transmission electron microscopy and differential scanning calorimetry. Mechanical milling of La55Al25Cu10Ni5Co5 alloy consisting of several intermetallics results in a formation of glassy alloy similarto that obtained by melt undercooling. The La--based glassy alloy formed by ball milling exhibits a well--defined glass transition and wide supercooled liquid region.

A new bonding method combining transformation superplastic diffusion bonding with diffusion brazing (T/DB) is proposed. A welding experiment was performed using a low carbon steel as base metal and BNi2 amorphous foil as interlayer at 1200℃. The results are as following: Although the time at elevated temperature is shortened and the number of temperature cycles is reduced to 3, a mechanically sound joint can be obtained easily. The content of Fe at the center of residual interlayer is increased to about 60%--70%, and in the region of base metal adjacent to welded interface, the Ni atoms diffused into the base metal almost mainly exist in the lamellar pearlite of base metal from the EDXS point analysis, which shows the significant selectivity for Ni atoms diffusion path. Furthermore, the phase boundary diffusions are classified as steady phase boundary diffusion and dynamic phase boundary diffusion, and the higher diffusion rate in the T/DB process is attributed to fully utilize phase boundary diffusions and to suitably utilize bulk diffusion.

The corrosion behavior and mechanism of a new nickel base superalloy have been studied at 550℃ and 700℃ in the synthetic coal ash/flue gas environments. The results indicate that low temperature hot corrosion of the alloy occurs at 550℃ and develops as pitting attack resulted from sulfidation mechanism. The frontal attack at 700℃ consists of two successive stages. The corrosion of specimen follows the oxidation and sulfidation mechanism during the initial stage. The protective Cr2O3film forms on the surface of the alloy and the internal sulfidation takes place. The severe low temperature hot corrosion happens due to the presence of molten CoSO4 during the accelerated stage. The porous and loose external scale and the compact internal scale consist of spinels and oxides, respectively. The sulfides of Cr, Ti and Nb precipitate at the front of oxides and in Cr—depletion zone. The rapid degradation of corrosion resistance of the alloy is attributed to the dissolution of both cobalt and cobalt oxide on the surface.

The solidification path and microsegregation during primary solidification of Al--Cu--Zn alloy were investigated comparably by CALPHAD coupled with Scheil model, simple Scheil equation and experiment. It was indicated by the CALPHAD that the solute content at the end point of primary solidification changes synchronistically with the start point, solute content in dendritic arm and the partition coefficient increase with the solid fraction. Simple Scheil equation was shown to be able to calculate the eutectic amount on the base of partition coefficient calculated by thermodynamics, its prediction for the different solutes is self--consistent and agreeable with the CALPHAD. The eutectic amount and SDAS was measured in DC experiment and compared to model value, it was shown that the experimental data are in agreement with the model prediction.

Ti1-xAlxN hard coatings have been synthesized by direct current (dc) plasma--enhanced chemical vapor deposition (PCVD). Dependences of aluminum content and annealing at elevated temperatures on the microstructure and hardness of Ti1-xAlxN coatings were investigated. The results show that plastic hardness measured by means of indentation test increases with x$increasing up to 0.83 and then decreases. XRD measurements indicate that the coatings with x<0.83 are fcc. solid solution with 3--10 nm grain scale. When x=0.83, relatively soft h--AlN phase is precipitated in Ti1-xAlxN coating, while the coating hardness begins a dramatical decrease. Furthermore, the nano--crystalline structure and high hardness of the coating can keep up to 900℃, which indicates the synthesized coating having good thermal stability.