The mechanism of the surface grain refinement for superalloy castings has been investigated by examination of some recent allied approaches and by observations on the investment molds of various effectiveness using X-ray diffraction, microscopy and EPMA. It was identified that the grain refinement of the castings is dependent upon the decomposition of cobalt aluminate and the active alloying elements contained in the alloy. After the alloy casted, the isolated particles of high temperature type cobalt were found on the surfaces of the investment mold and the castings, as well as the products of the mixture of cobalt aluminate together with different active elements heated under vacuum. Both the vacuum DTA of cobalt aluminate mixed with Al powder and the thermodynamical estimation of the reaction may confirm that the formation of Co particles did not accompany with an endothermic peak as formerly expected, but, in fact, a small exothermie effect was observed. It seems that the surface grain refinement of the castings is mainly due to the nucleation of the cobalt particles produced during casting. Thus, a rationalized process proposed has already been put into industrial practice with remarkable results.

Experiments were carried out on the hot salt stress corrosion cracking (HSSCC) of titanium alloy Ti-6Al-4V. The steady-state threshold stress for HSSCC in the temperature range of 300—420℃ was found to be lower than their creep strength at the same temperature. The post-exposure ductility of the specimens after HSSCC is susceptible to the strain rate. By ion microproble analysis, results show that hydrogen is markedly concentrated along the fractographic surface. It seems that the hydrogen produced halide salt corrosion and subsequently absorbed by the titanium alloy is responsible for cracking.

Using the technique of acoustic emission energy, the deformation and fracture in low strength and high plastic α-Ti has been studied. The logarithm of total acoustic emission energy is proportional to the logarithm of long range stress, and its turning point corresponds to the point of cracking. Therefore, it seems to be available to determine accurately the state of the cracking tip of these low strength and high plastic alloys by the method.Two kinds of acoustic emission sources were found during the plastic deformation and fracture process of α-Ti: the dislocation movement and the microcracking and growth of a crack.

A direct observation under different temperature and stress is reported of the martensite morphology variation vs the longitudinal change in a Cu-Al-Ni alloy specimen. The results indicate that its pseudo-elasticity and shape memory effect are closely related to the variation of martensite morphology.

A new X-ray diffraction topographic method was developed and used to study the fracture morphology and process in three point bending test of Mo single crystals. It was shown that not only the different regions similar to ones of optical fractograph, namely, unstable and stable extension regions, as well as triangle region formed by tearing, but also the different orientated subgrains, subgrains' rupture and their stress states may be revealed from the X-ray diffraction pattern. By the fracture process investigation, the subgrain rotation, the lattice bending and the spreading of plastic zone were found in front of the precracks.

The "operation effects" resulted from various adhesives were quantitatively evaluated during measuring the internal friction and modulus defect of low damping specimens by Marx-method. It was observed that the background of internal friction for some weak adhesives, e. g., phenyl salicylate, water glass mixture, etc. seems to be neither negligible nor constant value which is dependent on such factors as their temperature, thermal stress, time, dryness and strain amplitude, similar to that of dislocation damping. Two resonant peaks were observed for high temperature thermospalled adhesives with cracks. The internal friction measured at an anomalous resonant peak with slightly higher frequency decreases with increasing of the strain amplitude or the exciting time at high amplitude and increases with aging under lower amplitude. Peaking effects were found on the plot of internal friction vs strain amplitude or time when the resonant frequency of anomalous peak closed to the normal one. The hysteresis loops were observed on the plots of resonant amplitude vs frequencies. In certain combination of non-linear resonant hysteresis loop with aging effect, the almost periodic jumps of the repeatable amplitude of specimens may be presented. The aforementioned phenomena had vanished by using a high temperature and high strength adhesive, but the cracking of quartz dumb bar and plastic flow of specimen undergone by the thermal stress. An approach to its applicability and improvement has also been attempted.

On the basis of the vanguard electron model a formula for direct calculation of the maximum ionic probability radius, R_f, has been deduced: where α_0=Boron radius; R=Rydberg constant; Z=effective nuclear charge and t=spectroscopic state corresponding to the configuration of ion.An empirical formula for the contact radius of iso-electronic ion, R_0, may be expressed by: where k and b are empirical values.These formulas are characterized by clear physical definition, simple calculation and high accuracy of the data obtained.

A new high strength Mg sheet alloy, Mg-(3.0—3.8)％Zn-(0.6—1.0)％Y-(0.4—0.7)％Zr, has been developed. The final heat treatment condition is T_8 condition, namely, 515—520℃ solution treated 1h, water quenched+15％ cold work+200℃ 24h. Its room temperature tensile properties are shown as: σ_b=33.5—34.0 or 29.0—30.0 kgf/mm~2; σ_(0.2)=30.0—30.5 or 19.5—20kgf/mm~2; δ=12—14 or 13.5—22.0％ in longitudinal or transverse direction respectively. And its tensile properties below 250℃ are superior to that of both HM21XA-T_8 and AZ31B-H_(24).

A new modified aqueous magnetic colloid was satisfactorily prepared to observe the domain structure of coated Si-Fe sheets. Neither normal field nor other electrical apparatus is necessary for such observation.

Wire specimens of Fe-Ti alloys were loaded with nitrogen at 580℃, homogenized at 650℃ and then quenched at 570℃. When internal friction of these specimens were measured at a frequency of 1 Hz, two peaks were observed. The peak located at 20℃ is a normal nitrogen Snoek peak, while the one near 150℃ is the s-i peak (activation energy=1.13 eV). The addition of Ti to α-iron makes the nitrogen Snoek peak unstable, the later decays rapidly even at room temperature, In the range of atomic ratio N/Ti≤1, neither the Snoek peak nor the s-i peak is observed. Only after N/Ti>1, both peaks grow simultaneously with increasing nitrogen concentration. The fact that the height of the s-i peak changes linearly with Ti content indicates the peak to be contributed by reactions involving only isolated Ti atoms and has nothing to deal with either Ti-Ti atomic pairs' or clustering of solute atoms.It is suggested that two types of defects, the Ti-N pair defects and the N-Ti-N triplets, are the reaction centers giving rise to the s-i peak. The pair defects are formed by nitrogen occupation in the T_1-sites. The binding between the two constituent atoms of a pair defect is so strong that the alloys can be deprived of any free nitrogen as long as there are excess Ti atoms remaining in solution. This is the reason for the absence of any peak in the range N/Ti≤1. After N/Ti>1, excess nitrogen atoms will be distributed between the O_Ⅱ-sites and the T_3-sites at a ratio of about 1:10 until majority of the pair defects being transformed into triplets. Over a critical point of N/Ti≈2, almost all of the excess nitrogen atoms are only alloted to the O_Ⅱ-sites of the triplets, and in the mean time, the s-i peak begins to show a rapid increase in relaxation strength of about 10 times.The supersaturated nitrogen atoms frozed in the α-iron matrix during quenching tend to redistribute at room temperature by diffusing into O_Ⅱ-sites (diffusion length～10(?)), so as to make the Snoek peak transient. The profile of the s-i peak is solely determined by the concentration of excess nitrogen and is therefore independent of quenching temperature as well as cooling rate.

The computer simulation method was used to study the dynamical bchaviour of an array of long, straight, parallel coplanar dislocations containing a source under alternating loading which is sinusoidal function of time. It has been found that under alternating loading the glide motion of such a group is irreversible. After each loading cycle, the dislocation group can not recovered to its initial state, and there are ΔN dislocations in the group and the leading dislocation is still at Δx_1 distance from the source. The relation among ΔN, Δx and the loading amplitude, the loading period, the number of loading cycles was also deduced.

On the consideration of series of experimental and theoretical results, the type A and B hybridizations of states of Fe are constructed in the form of 18 discontinuous of hybride levels, belonging to which various valence electrons, atomic magnetic moment and bond length parameters are listed. On the basis of these parameters together with the parameters of N obtained from diatomic molecular spectroscopy, analysis of the valence electron and magnetic moment structures of α-Fe, γ-Fe and Fe_4N=Fe~cNFe_3~f are made. From these structures, interpretation of the space distribution of magnetic moments in the crystalline lattice of α-Fe obtained from neutron diffraction, and interpretation of experimental magnetization curves along [100], [111] and [110] of α-Fe are made and the cause of difference of the atomic magnetic moments of Fe~c and Fef in Fe~cN Fe_3~f are analysed. The phase transformation α-Fe→γ-Fe as due to magnetic origin is contemplated with care. The case hardening by nitriding steel at high temperature and the graphitation in cast iron due to heat effect at high temperature are both traced back to the valence electronic structure origin.

With the aid of TEM observation on the martensite transformation of three steels, the relative position changes between atoms caused by the dislocation slip during fcc→bcc (or bet)lattice variation are mainly carried through shears in combination with different slip systems along<112>_f/{111}_f in the distorted fee lattice. Two types of these changes are: (1) the atom displacement of main shear only along one slip system of (1/m)<112>_f/{111}_f and (2)M=(1/n)[<112>_(f1)/{111}_(f1)+<112>_(f2)/{111}_(f2)] complex displacement obtained by all dislocations operated in different {111}_f conjugate sIip plane (n>6). Thus, an atom displacement mechanism of which the austenite transforms into martensite through conjugate complex slips of partial dislocation during quenching has been proposed. Based upon this mechanism, the crystalline geometry explanation may be undoubtedly made on the orientation correlation between matensite and austenite matrix, as well as the martensite habit planes and some special phenomena in quenched martensite.

An effort was made in the fields of setting up the metallurgical process modules and the process system simulation programs which include: general principles and examples of metallurgical process simulation; the dynamic simulation of chemical metallurgical processes and inorganic thermochemical data base and its applications in metallurgical process simulation.

The oxygen partial pressure and activity of FeTiO_3-MgTiO_3 solid-solution system were studied by means of oxygen concentration cell with the solid electrolyte. Taken Fe/FeO as reference electrode, the oxygen partial pressure was measured, and standard free enthalpy of formation of non-stoichiometric FeO (w/istite) was obtained as ΔG_(FeO)~0=-259030+60.25T (J·mol~(-1)). Furthermore, the oxygen partial pressure and wustite activity of the various compositions of the solid-solution system were measured at the temperature range of 900—1200℃. The results show that both values increase with the increase of temperature. The standard free enthalpy of formation of complex oxide FeTiO_3, ΔG_(FeTiO_3)~0=-40470+19.75T (J·mol~(-1)) and the activity-composition relation of this solid-solution series were obtained.

By determining the polarization cureve, the discharge potentials of Sb (Ⅲ), Sn (Ⅳ), As (Ⅲ) and H~+ ions were obtain edas -0.92, -1.04, -0.81 and -1.12V respectively, which were checked with the thermodynamic calculation. Their dynamic parameters were estimated by the Tafel lines which resulted after modification of their concentration polarization. At 50℃, the rate constant was found to be 1.54×10~(-4) cm/s, the exchange current density 10~(-2) A/cm~2 and the activation energy 9100 cal/mole.The cathodic discharge mechanism of SbS_3~(3-) and SbS_4~(3-) has been investigated with the help of the chronopotentiometric method. It was shown that before charge transfer, the SbS_3~(3-) ions should be subjected to a prior chemical reaction, i. e. SbS_3~(3-)=SbS~++2S~(2-). The SbS~+ ions formed are then reduced to metallic Sb at the cathode. The discharge of SbS_4~(3-) ions at the cathode takes place in two steps: First, SbS_4~(3-) ions are directly discharged into SbS_3~(3-) and then further reduced into metallic Sb.