A new precipitate phase (X phase) has been studied by means of TEMin 2024--2324 Al alloys in different aging conditions. The X phase is determined tohave a c-centered orthorhombic crysta1 structure with a=0.492nm, b=0.852nm andc=0.701nm. It is found that the X phase forms on the helical dislocation in theinitial stage of aging. At lower aging temperature, the S series precipitate phases do-minate the precipitation; while at higher aging temperature, the X phase dominates theprecipitation.

The internal friction was measured as a function of temperature inMn-7.2wt-%Cu and Mn-10wt-%Cu alloys at -100-400℃ by inverted torsionpendulum. The well known peaks at -40℃ and 160℃ are caused by twin boundaryrelaxation and martensitic transformation respectively. The new peak at 40℃is associated with the martensitic phase transformation in the poor Mn region, whichwas frmed by spinodal decomposition. The relaxation peak at 200℃ is attributed toα-Mn precipitate. The activation energy of the peak is 1.1 eV. The relaxation peakat 340℃ occurred only after aging under vacuum at 430℃. The peak height increas-ed through aging. The activation energy is 1.47 eV. The peak is supposed to relateto spinodal decomposition in which the Cu-rich region formed.

The nucleation of boride at boronizing initiation and the microstruc-tural difference in the boride crystals at different stages of growth have been observedunder TEM by using the single-side thinning technique. It seems that the controllingfactors proposed in growth of boronizing layer are the density of nuclei of botidecrystals on the surface of specimen. The process of boronizing may be described bytwo stages. First, the layer will thicken as the time prolonged according to a parabo-lic growth. Until the compressive stress in the front of layer is increased over a cer-tain critic value, the stacking faults will be induced in the boride crystals, so as toacceralate the diffusion of boron atoms and the growth of layer. This is called thesecond stage. In this stage, the deviation from parabolic growth and the approach toa straigh line relationship would be revealed.

The kinetics of bainitic transformation of GCr15 steel below M_s andthe influence of the pre-transformed martensite formed at different temperatures onit at 240℃ isothermally have been investigated by means of magnetic measurementand dilatometry. The supercooled austenite transforms firstly into martensite and thenlower bainite during isothermal process below M_s. With the decrease of isothermaltemperature and the increase of pre-transformed martensite, the bainitic transforma-tion is accelerated by stress-induction effect, and the incubation period is the shor-test at about 200℃. However, the bainitic transformation becomes rather difficult,if the temperature is continued to decrease and the pre-transformed martensite exce-eds a certain limit. Moreover, it seems in a similar manner of the pre-transformedmartensite increased at different temperatures below M_s to be influenced on the su-bsequent bainitic transformation at 240℃ isothermally.

The effects of boron on the hardenability of austenite in dual-phasesteel and on the tensile properties have been studied. With the increase of B con-tent from 0.0005 to 0.0029 wt-%, the volume fraction of the austenite decreases atconstant intercritical annealing temperature, the hardenability of the austenite inc-reases remarkably, as well as the martensite content and the sensitivity of tensile pro-perties to cooling rate decrease. It was observed by using isotopic tracing autora-diography that after intercritical annealing, the B atoms segragate at the grainboundaries of austenite produced during normalization, no evidence of the migrationof B atoms to α/γ interfaces was found.

The effect of Cr, Mn, Co, Ni, Si or Mo on the diffusivity of hydro-gen in α-Fe has been studied near room temperature by means of sensitive electro-chemical method. With the of increases concentration of alloying elements, the hydro-gen diffusivity in binary Fe alloys decreases and the activation energy increases. Onthe basis of Oriani's approach, the trapping parameters of the alloying elements inthe Fe alloys are calculated and the trap binding energy between the above elementsand hydrogen are obtained. The increasing sequence of trap binding energy is asfollows: Si, Co, Ni, Mn, Cr, Mo. Finally, the nature of the trap binding energy isanalyzed according to both the interaction between lattice strain field induced byalloying elements and hydrogen elasticity as well as the interaction between alloy-ing elements and hydrogen electricity.

An investigation was made of the crack initiation by cyclic deforma-tion on the two special surfaces of pure Al single crystal. One is the tOp-surfaeccontaining the active Burger's vector, and the other is the top-surface formingthe smallest angle with the active Burger's vector. The fatigue test were performedat the constant resolved shear stress amplitude (5 Pa), zero mean stress. Expe-rimentel results show that many microcracks are initiated on the side-surface onwhich no PSB surface topographies are formed during cyclic deformation. Cracks onthe side-surface are short with mean length of about 10 μm and distributed appro-ximately in a similar direction, perpendicular to the active Burger's vector, whereason the top-surface cracks are not as obvious as those on the side-surface. Observa-tions by scanning acoustic microscopy show that the fatigue damage on the side-surface is much deeper than that on the top-surface. A model of crack nucleationby voiding, and of PSB formation is suggested.

The tensile properties and fracture behaviour of an Al-Li-Zn-Zr alloyhave been investigated in the temperature range of 295 to 77 K. The ultimatestrength and elongation increase with decreasing temperature, whereas the yieldstrength remains merely unchanged. Its fracture is transgranular obviously at roomtemperature, but it appears mainly as intergranular mode at 77 K. The deformationmode near the fracture surface is similar at both ambient and cryogenic tempera-tures. However, at room temperature, the ftacture of the alloy is mainly concernedwith the planar slipping, which may be weakened markedly at 77 K, and at thesame time, the fracture behaviour of the alloy is dependent on the interaction betweenslip bands and grain boundaries.

The deformation process and the strain field ahead of crack tip in Nisingle crystal have been studied under SEM, and by grid method as well as compu-ter image processing. The presence of hydrogen extends the localized plastic bandinhomogeneous further and significantly increases the plasticity in high deep strainarea.

The prestrain on 16MnR steel was found experimentally to influenceupon its fracture toughness remarkably. During fracturing, the microstructural chan-ge was taken place in the crack tip zone, The cracks will propagate on the second-ary pearlite structure. It is believed that the fracture toughness of 16MnR steel isdependent on the diffusion of carbon atoms and the formability of secondary pearl-ite structure.

The effect of size and distribution of σ-Ni_3Nb phase on the fatiguefracture behaviour of Inconel 718 under constant controlled cyclic strain at 650℃has been studied. The serrated distribution or σ-plates along grain boundaries, form-ed by high temperature solid solution and intermediate aging, may improve the re-sistance to crack propagation along boundary, so as to change the low cycle fatiguefracture from intergranular, undergone standard heat treatment, into transgranular.However, the precipitation of coarse plate σ-phase along twin boundaries and throughgrains, after treated on the schedule for forming the serrated boundary, mayreduce the adoptability to cyclic strain and may often form too early the microvo-ids at interface between σ-phase and matrix and even disintegrate to shorten the fa-tigue life of steel.

The 12Ni3CrMoVA steel was selected to investigate the effect of tem-perature, strain rate and microstructure on yield strength of bcc metals. The therma-lly activated parameters, such as activation energy and volume, were measured andcalculated. According to the P-N mechanism, the plastic deformation of bcc metalswas analysed by the core structure of screw dislocation, the theoretical predictionwas in good agreement with experimental results. It is explained that the "turningpoint" in curve of effective stress against temperature was caused by a double cam-el-hnmp--shape of Peierls potential energy, which resulted in a new pair of doublekinks forming in the front of original one. The microstructure can only affect on athermal stress, but not affect on thermally activated movement of dislocation.

A simple model based on the Landau-Devonshire theory is capable ofsimulating the pseudoelastic hysteresis behaviour, i. e., the hysteresis becomes smal-ler with increasing temperature and disappears at a critical temperature. The exper-imental results obtained from CuZnAl memory alloy singleicrystalline specimensshow that the width of the hysteresis loop does not change significantly with tem-perature, while the size of the interior loops of hysteresis strongly depends on themartensitic phase fraction induced from austenite.

The microstructure of laser melted layer on Cr12MoV steel was re-vealed as austenitic dendrites under observation by TEM and X-ray diffractometer.The interdendritic eutectic carbides together with a little secondary carbides, but nomartensitic transformation were presented. Thus, the hardness of the steel is reducedafter laser melting treatment.

Experimental study has been carried out on the effects of the com-plex strengthening of the cold expansion of hole (CEH) and overloading on the fa-tigue crack initiation life (FCIL) of LY12CZ alloy The overloading has noappreciable effect on FCIL of LY12CZ alloy after CEH. Moreover, an expressionfor FCIL of LY12CZ alloy subjected to the complex strengthening is given. Basedon the above-mentioned results, a new method is put forward to predict theFCIL of LY12CZ alloy after CEH under variable amplitude loading and has beensubstantiated by the fatigue tests under a flight-to-flight spectrum.

From the thermodynamical analysis, the criterion of optimization inheating process of materials with minimum energy consumption, so called optimalheating process criterion, has been established. Based upon the criterion togetherwith the linear quadratic control theory, a new general method for optimizationin heating process was developed. This method is believed to be quite available tosolve the heat supply problems not only in distribution along furnace length butalso in time-variation.

Taking into account the ionic strength and temperature, the ionicequilibrium equations of Fe(Ⅲ)-H_2SO_4-Na__2SO_4 aqueous solution are suggested andcan be used for the calculation of various ionic species concentration in the aque-ous solution. The calculations of concentration profiles at different conditions in-dicate that the equilibrium concentration of various ionic species in the solutionscan significantly be changed by the ion interactions. And, it suggests that the quan-titative calculation is necessary for the analysis of experimental results.

Radial distribution function of molten solution of LiF-KCl reci-procal system has been obtained by means of X-ray scattering and Monte Carlocomputerized simulation, respectively. It is shown that a more compact cluster tendsto form among Li~(+) and F~(-) ions with small size while the bonding of K~(+) andCl~(-) ions with large-size would be relatively relaxed in LiF-KCl melt, i. e. "larger-larger and smaller-smaller" effect.

The region containing O-50at. -%Nd in Nd-Pd phase diagram has beenstudied with X-ray diffraction, differential thermal analysis and micrography. Ithas been found that there are six intermetallic compounds, viz: NdPd_5, NdPd_3, NdPd_2,Nd_2Pd_3, Nd_4Pd_5 and NdPd in the region. NdPd_3, Nd_4Pd_5 and NdPd are solid-liquidcongruent compoundls with melting points of 1217±4℃, 970±4℃ and 952±4℃respectively. It has been determined also that NdPd_5 has a SmPt_5-type structure withα=0.5265nm, b=0.9219nm and c=2.570nm, NdPd_3 a AuCu_3-type structure withα=0.4137nm, and NdPd below 708±4℃ a CrB-type structure with α=0.3827nm.b=1.079nm and c=0.4547nm. There exist three eutectic reactions L?(Pd)+NdPd_5,L?NdPd_2+Nd_2Pd_3 and L?Nd_4Pd_5+NdPd, and three peritectic reactions L+NdPd_3?-NdPd_5, L+NdPd+3?NdPd_2 and L+Nd_4Pd_5?Nd_2Pd_3 in the region.

The corrosive wear rates, friction coefficient and reformation periodof passive film of stainless steel specimens were determined in dilute H_2SO_4 solu-tion and then the morphology of the wear tracks was observed. The experimentshows that the corrosive wear rate of high alloy stainless steel is higher than thatof 304 stainless steel. It is suggested that the way to develop an alloy to be appli-ed to corrosive wear is not the same as those used in corrosion environment.

Studies were made of the influences of the nature of Gd_2O_3 and itscontent on the oxidation behaviour in air at 1100℃ of the NiAl-type compositecoatings modified with dispersed Gd_2O_3, which were prepared by co-electrodeposi-tion and diffusion aluminizing techniques. In comparison with Gd_2O_3-free coatings,the oxidation of the modified coating may be remarkably retarded and their scalesadherence to substrate may be much improved. It was revealed that some Gd_2O_3 par-ticles are still dispersed in the coatings, but other sub-microscopical Gd_2O_3 particl-es are accumulated at the coating/substrate interface just after aluminizing. Theaccumulated layer is found as the diffusion barrier. A continuous Gd_2O_3 particlelayer would be more effective to reduce the degradation and to prolong the life ofthe coatings.

A technique for in situ heating TEM observation of the dynamiccrystallization process of amorphous Ni-P alloy foils is described. During isothermally annealing at medium temperatures, the fine ordered clusters were firstly ap-peared in the amorphous matrix, and the microcrystalline particles formed withrandom orientations, then crystals grew. In the process of crystal growth, thesemicrocrystalline particles deposited directly onto the crystals front by shearing (i. e.changing their orientations) with their configuratures unalterated. By means ofvideo camera, the "step-shaped" growth, correspnding to the ordered clusters shear-ing deposition, was found after measurement of crystal growing. With the shear-ing deposition under different temperatures, the rate change of crystal growthseemed to be different.

Based on both the authors' experimental results of the in situ TEMobservation of dynamic crystallization in amorphous Ni-P alloy and the microstruc-tural analysis on the amotyhous alloys, a new micromechanism of their crystalli-zation has been approached. It is believed that the crystallization is composed oftwo processes: the diffusion of single atoms from amorphous state to crystallinesurface, and growth of ordered clusters and their shearing combination or shearingdeposition. Its availabilty has been satisfactorily tried to explain the experimentalresults of the structural configurature of the crystals formed in very early stage.

The twin configuration of β-SiC whiskers prepared by the vapour-liquid-solid process was investigated under TEM. In the whiskets with "bamboo jo-int" morphology, the β-SiC part is jointed with the one-dimensional disorderedpart by the {111}_β twin boundary, microtwins of several atomic layer thick werepresent in β-SiC region. It was noted that the {111}_β twin boundary divides someβ-SiC regions into two parts. The second-order twin joint was also observed at thewhisker tip and the branching position of the bicrystalline whisker. Electron diffra-ction analysis demonstrated that the twin boundary in the bicrystalline whisker isparallel to the matrix {001)_M plane.

A proper amount of Ce added under elevated temperature and prolong-ed time may remarkably reduce the wetting angle, i. e., from non-wettable intowettable, between Al_2O_3 and medium manganese steel. The improvement on wetta-bility by adding Ce is due to the adhesion, accumulation and interaction at theinterface to produce CeAlO_3. The equilibrium wetting angle is determined by thedegree of interfacial reaction.

The composition profiles and microstructures of Fe surface layers im-planted with Ni~(+), Mo~(+) or B~(+), respectively, or their combinations have been inves-tigated by AES and TEM. The implantations were carried out with ion energy50--150 keV for dose 5×10~(15)--4×10~(17) ions/cm~2 at room temperature. The residualpressure in the target chamber is around 1.3×10~(-3) Pa. AES shows that a carburiz-ed layer containing 20--30 at.-% C formed at the surface. Its thickness dependson the ion species and the ranges. It has been found that a relatively thick surfaceamorphous layer is formed after combined implantations. Metastable carbides of hcpstructure and fcc austenites are also observed. The effects of collision cascades andthe radiation-enhanced diffusion on the surface carburization and the formation ofamorphous layers are discussed.

According to both theoretical analysis and practice, the decelerationof scale formation on the steel during direct fire heating was adopted by injectionof the water-in-oil emulsified fuel oil, replacing the conventional heavy oil, withan emulsifying equipment. To put into productive practice, the relative deceleratingrate after data from the Shanghai Pressure Vessel Plant and the Shanghai SteelWorks No. 3 is about 20 and 15% respectively. Mossbauer spectra showed that theFe_3O_4 content is much abundant in the scale, especially in its inside layer duringfiring the emulsified fuel oil, The reason, why the scale formation might berecelerated, seems to be a protective Fe_3O_4 film is formed and an excess of oxygenis reduced.

The oxidation of sintered NdFeB magnets in humid atmosphere at90℃ has been studied by TEM, AES, X-ray and Mossbauer spectroscopy. The blackoxidized structure, enriched with Nd and oxygen and depleted of Fe, was found inthe surface layer by the selective corrosion of the Nd-rich boundary phase and thenpenetrated along grain boundary and spreaded into Nd_2Fe_(14)B matrix phase. Thebrown rust powder, consisting of residual Nd_2Fe_(14)B particles and oxides, e. g., Fe_3O_4,Nd_2O_3, α-Fe_2O_3 and cubic Fe_2O_3, was found after 500 h oxidation. The changes indistribution of Fe atoms at six crystalline sites, especially the loss of Fe atoms atc-site, in oxidized Nd_2Fe_(14)B lattice was measured. The decreaslng in magnetizationof oxidized NdFeB magnets was reported.

The kinetics of interaction between PbS and PbO.SiO_2 at 900--1100℃has been studied. A function F(α)=(0.5-α)/(1-α)~(5/3), α being the ratio of SO_2 evolvedin time τ to the total SO_2 to be escaped, is revealed as straight line on a plotagainst τ. The reaction seems to be under chemical or diffusion control during theprior or later stage, of which the activation energy calculated is △E=140 or △E_D=105.8 kJ/mol, respectively.

Mossbauer spectroscopy and magnetic measurement were used to inve-stigate the site and properties of SmFe_(10-x)Co_xTi alloys. When Fe was substitutedby Co in a small amount, Co preferred 8f site; the substitution of Co for Fe re-sulted in a great increase in Curie temperature T_c of the compounds The satura-tion magnetization and the hyperfine field of Fe presented similar curves withincreasing Co content χ, each reached maximum at χ=2. The Co substitutions hadeffects on exchange interaction between alloy atoms, which caused increasing Curietemperature.