As a basic study on fatigue damage accumulation of structural members (for example; truck frame and plate spring of rolling stocks) subjected to service load, the authors have carried out fatigue tests under superposed stress wave on fifteen metalic materials (about 400 experimental points.) In this paper, a few factors having influence on the predicted fatigue life have been discussed by comparing the experimental life N with the predicted life Nes with regard to these experimental points. For the fatigue life prediction method, three kinds of stress cycle counting methods were used; the zero cross range pair mean method (ZRPM method), the range pair mean method (RNPM method) and the range pair method (RNPR method). The modified Goodman's method was used to evaluate the effect of mean stress on stress amplitude. The lives were predicted by using the linear cumulative fatigue damage law. The results obtained were as follows. The most suitable cycle counting method was RNPM method, because of the scatter band of N/Nes being narrowest in three cycle counting methods. The mean stress effect on fatigue life amplitude was very complex, and the modified Goodman's method was not always adaptable for the prediction of the above experimental points. In order to make narrow the scatter band of N/Nes, some discussions were made on the fatigue damage accumulation caused by the secondary stress amplitude cycling.
Residual stresses which developed through the molding process of epoxide composite with metallic electrodes embedded were studied. Viscoelastic properties of the epoxy were obtained by bending creep tests under constant temperature conditions and the validity of the time-temperature superposition principle was verified through the tests under slowly varying temperature conditions. The residual strain measurements were conducted using circular epoxy plates in which Cu or Al electrodes were embedded. The comparison of the results with the linear viscoelastic solutions of the residual strains obtained by assuming linear cool-down from Tg to room temperature showed good agreement, and it was suggested that the simple elastic FEM solution was applicable for practical applications. The long-term change in residual stress of the epoxy plate with metallic electrodes embedded placed in a cyclically fluctuating temperature atmosphere was estimated based on the theory of linear viscoelasticity and the related thermal fatigue strength was discussed.
Energy absorbability of foamed rigid polyurethane was studied by static compression and impact tests on knee and head models with the shapes prescribed in MVSS (Motor Vehicle Safety Standard in U.S.A.). The following results were obtained. (1) Foamed rigid polyurethane absorbs much energy in a manner similar to ductile materials. (2) The MVSS for the knee model is satisfied by a foamed rigid polyurethane with the density of 0.075g/cm3 and the thickness of 120mm, and the MVSS for the head model is satisfied by the one with 0.04g/cm3 density and 80mm thickness. (3) No dependence of deformation behavior and energy absorbability of foamed rigid polyurethane on the deformation rate was observed. (4) The compressive load F and absorbing-energy E in knee model or head model tests can be described simply as follows; F=σpS, E=σp∫Sdl where σp is the plateau stress obtained by the flat indenter test, l is an indentation depth of the knee model or the head one, and S is the cross section of the indenter perpendicular to the direction of a compressive load.
The boronized process is famous as one of the chemical methods to achieve case hardening of steels. There have been a number of studies concerning this treatment process and boronized layer. Consequently, it becomes well-known that the boronized layer has above 1400 Vickers hardness and a high resistance especially against wear. Also the present authors have reported the residual stress distribution by using X-ray method, the effects of thickness and compressive residual stress of boronized layer on bending fatigue strength properties and the results of fatigue crack observation. On the other hand, few reports exist on the tensile mechanical properties. One of the characteristics in boronized steels is the formation of a carbon redistributed layer below the boronized layer. It is expected that this carbon redistributed layer contributes to the improvement of the total tensile and yield strength. In the present study, tensile tests were conducted on mechanical and structural steel specimens of different diameters, which contained various amounts of carbon and had different thickness of boronized layers. The Effects of the thickness of boronized layer and the carbon redistributed layer on tensile strength were investigated. The results obtained are summarized as follows; (1) For SCM 420 and S50C, both the tensile and the yield strength were improved with increasing thickness of the carbon redistributed layer. (2) Rupture elongation became less with increasing thickness of the boronized layer. (3) As the carbon content became more, the ratio of boronized area to carbon redistributed area effective in improving tensile and yield strengths became smaller. The ratio was 45% in SCM 420, and 25% in S50C. (4) When the carbon redistributed layer became thicker than the tensile residual stress distributed layer, both the tensile and yield strengths improved.
Notch characteristics in static strength of sintered steel was investigated experimentally and theoretically. Circumferentially notched bars and edge notched plate specimens with various root radii were examined at room and liquid nitrogen temperatures. The fracture criterion was discussed based on the stress distributions calculated by the finite element method considering the non-elasticity of the material. According to the experiments, sintered steel showed very low notch sensitivity at room temperature. Strength decrease by a notch was within 20% when the form factor was up to 3. At liquid nitrogen temperature it showed higher sensitivity, the strength decrease reached to about 50 for the same notch shape. These notch characteristics are similar to those of cast iron. From the calculated stress distribution, it was revealed that the low notch sensitivity was caused by the following two factors, same as in cast iron: the non-elasticity of the material that reduces the maximum working stress to a much smaller value than the elastic estimation, and the existence of an over stressed region at the notch root where the stress is higher than the tensile strength of the material. The over stressed depth δ at fracture was 1-3mm at room temperature and 0.2-0.5mm at liquid nitrogen temperature, taking a smaller value in small notch radius. The smallest value of δ was several times the particle size of raw powder, that is the minimum dimension below which the porous sintered steel can not be treated as a continuum solid. The larger value of δ at room temperature was inferred to be the propagating condition of an initiated cracks.
The effect of side-grooves on elastic-plastic fracture-toughness (JIc) determination of high- and medium-strength structural steels, HT80 and SM50A, was investigated by using CT-type specimens of three different thicknesses; the ratio of net thickness to gross thickness was kept at 0.75 and the net thickness was ranged from 6 to 18mm. The JIc tests were conducted in accordance with the standard JIc test method proposed by the Japan Society of Mechanical Engineers (JSME). The use of side-grooved specimens generally made it much easier to determine the onset of ductile tearing by the R-curve method and the electric potential method, and improved accuracy and scatter of the toughness values thus determined, provided all the size-requirements for the specimen prescribed in the JSME method were satisfied. The critical stretched zone width (SZWc) of the side-grooved specimens was found to be smaller than that previously determined for the standard CT specimens without side-grooves. This was attributed to higher triaxiality produced by the side-grooves. The SZWc values of the side-grooved specimens were rather insensitive to the specimen thickness in the range of thickness investigated. Good agreement was observed between the JIc values at physical onset of ductile tearing and the JIc values determined by the R-curve method and the electric potential method. The stretched zone width method gave slightly larger JIc values than those by the R-curve method for SM50A, as has been observed for the standard specimens without side-grooves.
The Sheet specimens of mild steel were fatigued at room temperature by plane bending stressing. The dislocation structures during the process of fatigue crack initiation and near the tips of fatigue cracks were observed using a transmission electron microscope and the dislocation density and the size of cell structure were obtained quantitatively. The relationship between the dislocation structure and the plastic deformation zone near the crack tips was examined. The main results obtained are as follows: (1) During the process of crack initiation, the total dislocation density ρt increased to a saturation with progress of fatigue. The dominant structure at high stress amplitude was cell structure, and that at low stress amplitude was loop patch structure. The cell width lw (or the bundle spacing ls) and the free dislocation density ρf in the cell structure decreased gradually during this process. (2) From the observation of dislocation structure near the crack tip, the mean size of cell structure lm and the free dislocation density ρf in the cell structure were found to decrease with approaching the tip. They showed almost the same distribution in all directions from the crack tip, and this influence of decreasing lm and ρf extended to the distance of about 1000μm from the crack tip. (3) The plastic deformation region near the crack tip determined by Vickers hardness tests was larger than the area where the mean cell size lm was found to be below about 4μm.
In order to discuss the influence of lamellar structure and MnS inclusion on fatigue fracture mechanism, the fatigue crack propagation tests were carried out along different orientations of a hot rolled SB49 steel plate with or without homogenized heat treatment. The as-received material showed directionality in fatigue crack propagation due to the effect of lamellar structure, while the homogenizing treatment made crack growth isotropic both macroscopically and microscopically. Fractographic examination revealed that the ordinary striation mechanism was predominant when the crack plane was normal to the lamellar structure as well as for any orientation of the homogenized structure. For the case where crack plane was parallel to the lamellar structure, however, sub-cracks initiated at the edge of MnS inclusions locating ahead of the main crack and each propagated independently by striation mechanism to join with the main crack. The increase in fatigue crack propagation rate was explained by the interactive effect of MnS inclusions and lamellar structure.
In order to elucidate the influences of the variation of eutectic carbide content and the distribution of matrix on the rolling contact fatigue strength of alloy white cast iron, an experimental study was conducted by using an Amslar-type wear testing machine under sliding conditions with light oil as a lubricant. The results obtained are summarized as follows: (1) The rolling contact fatigue endurance limit of alloy white cast iron was lower than that of bearing steel (SUJ2) having the same microstructure of matrix of the alloy white cast iron. The tendency became more pronounced with increasing amount of eutectic carbide. (2) The rolling contact fatigue endurance limit of alloy white cast iron was improved by forging. (3) The rolling contact fatigue endurance limit of the alloy white cast iron increased with matrix hardness. (4) A crack propagated along the boundary of eutectic carbide and matrix structure, and after that it penetrated into the vicinity of the region at which the stress value calculated according to the shearing stress amplitude maximum theory reached the maximum, and then the crack propagation went in the parallel direction with the contact surface.
In order to elucidate the influence of the work hardening phenomenon caused by repeated contact stress on the rolling contact fatigue strength of austenitic stainless stell (SUS 304), an experimental study was conducted by using a Nishihara-type wear testing machine under lubricated state. The results obtained are summarized as follows: (1) The rolling contact fatigue endurance limit of SUS 304 stainless steel increased remarkably with work hardning generated by repeated contact stress. (2) The rolling contact fatigue endurance limit of SUS 304 stainless steel decreased with increasing slip ratio. (3) The depth of crack propagation in SUS 304 stainless steel was deeper than the region at which the stress value calculated according to the maximum shearing stress theory reached the maximum, regardless slip ratio.
X-ray fractography is a useful technique for analyzing the cause and mechanism of fracture from the information obtained by X-ray irradiation on the fractured surface. In this report, the relations between the residual stress on the fatigue fractured surface and some parameters of fracture mechanics were investigated. Special attention was paid to the contact of fractured surfaces due to crack closure, as well as to the effect of surface roughness on the X-ray stress measurement. The materials used were a Ni-Cr-Mo steel (SNCM 439) and a mild steel (SS 41). The test pieces were CT specimens with W=51mm and B=12mm. On the fractured surface the residual stress along the crack propagation direction was measured by a parallel beam X-ray stress measure Cent apparatus. The main results obtained were as follows. (1) In the ductile material the residual stress on the fatigue fractured surface increased at first and then gradually decreased with increasing crack length. In the brittle material the residual stress increased monotonically and kept almost a constant value do the final fracture surface. In both cases the residual stress could be correlated to the Kmax values. (2) The residual stress on the fatigue fractured surface was not affected by the contact of fractured surfaces due to crack closure. (3) The roughness of the fatigue fractured surface increased with increasing K value, and it was found that the stress measured by X-ray diffraction tended to give a lower value than that of the appplied stress.
The effect of stress ratio R on the threshold stress intensity range ΔKth for fatigue crack growth has been studied in several steels. ΔKth decreased with increasing R, and the relation was generally expressed as ΔKth=(1-α·R)·ΔKth0 for R≤Rc =(1-α·Rc)·ΔKth0=const. for R>Rc where Rc depends only on the material, and α andΔKth0 are functions of the specimen geometry as well as of materials. The effective value of the threshold stress intensity range ΔKeffth was 2 to 4 MPa√m for all materials and stress ratios examined. The constant value of ΔKth for R>Rc in the above equation was also 2 to 4MPa√m, which is nearly equal to ΔKeffth. The crack closure is concluded to be mainly responsible for the effects of stress ratio and material on ΔKth in steels.
The effect of the variation of stress frequency on the fatigue crack propagation rate in 99.5% pure titanium, which has remarkable strain rate dependence in the plastic region, was studied. Fatigue crack propagation tests were carried out under four stress frequencies: two constant stress frequencies (20Hz, 0.02Hz) and two stress frequencies changed step-wise (20→0.02Hz, 0.02→20Hz). An elasto/visco-plastic analysis of fatigue crack propagation was performed by the finite element method (FEM), and comparison between the change of crack propagation rate due to variation of stress frequency and the visco-plastic strain behavior at the crack tip calculated by the analysis was made. The results obtained in this study are summarized as follows: (1) It was found from the experiments that the crack propagation rate changed characteristically depending on the variation of stress frequency. (2) A parameter closely related to the fatigue crack growth rate is the visco-plastic strain range at the crack tip. (3) The effect of the variation of stress frequency on the fatigue crack propagation rate may be explained by the variation of visco-plastic strain range at the crack tip resulting from the strain rate dependence of the material.
The influences of the grain size down to 0.8μm and the ageing at 475°C on the crack growth behavior were studied on the 32%Cr-8%Ni steel with the ferrite matrix and the stable austenite phase. The main results obtained are as follows; (1) Both 0.2% proof stress and the crack growth resistance increased with ageing at 475°C for the coarse grained specimen, whereas the crack growth resistance decreased for the ultrafine grained one. (2) The high crack growth resistance of the coarse grained specimen with ageing may be associated with the change of slip mode in the ferrite matrix from wavy type to planar one. The crack growth mode was transgranular for both the aged and quenched specimens. The low crack growth resistance of the ultrafine grained specimen, on the other hand, may be caused by the change in crack growth mode from transgranular to interfacial. (3) The influence of grain size on the crack growth resistance is more pronounced for the aged specimen than for the quenched one.
The crack propagation behavior in fretting-fatigue was investigated with type SUS 304L stainless steel. To induce fretting-fatigue the specimen which was pressed by the fretting-pads made of SUS 304 was subjected to the cyclic direct stress. The conditions of the fretting-fatigue tests were the stress amplitude of σa=120MPa, the contact pressures of p=40, 80, and 120MPa, and the stress ratios of R=-0.33, 0, and 0.33. The unfretting-fatigue tests of p=0MPa were also carried out for the same values of σa and R so as to compare with the results of the fretting-fatigue tests. The fretting-fatigue crack initiated at the site on the specimen which was in contact with the edge of the fretting-pad, and propagated at the early stage to the direction inclined to the normal of the cyclic stress axis. The effect of the contact pressure on the crack propagation rate da/dN in the fretting-fatigue under R=0 appeared in the short crack region to about 1mm in crack length. Although the da/dN in the short crack region showed the tendency to increase with increasing p, the da/dN for p=120MPa was the same as that for p=40MPa in the cause of the contact of the opposite surfaces of the oblique fretting crack. The effect of the stress ratio on the da/dN in the fretting-fatigue was mainly owing to the existence or inexistence of the crack surface contacting. Under R=-0.33, the heavy contacting occured at the oblique crack surfaces so that the crack propagation was retarded to the long crack region compared with the case of unfretting-fatigue. Under R=0.33, however, since the contacting did not occur, the da/dN in the fretting-fatigue always surpassed that in the unfretting-fatigue.
It was previously found that the fretting-fatigue crack propagation rate in type SUS 304L stainless steel increased with crack length, but at a certain length it decreased to a minimum value and then increased again for stress ratios of R=-0.33 and 0. This phenomenon has been interpreted based on the contacting of the crack surfaces. Since SUS 304L is a metastable austenitic stainless steel, the martensitic transformation induced by deformation strengthens the crack tip and may reduce the crack propagation rate. In order to clarify the effect of martensitic transformation on the crack propagation under fretting-fatigue, microbeam X-ray Fractography on the fractured surface was performed. The martensitic transformation occurred both in the fretting- and in the unfretting-fatigue fractured surface. The half value breadth Δ(2θ) of the martensitic phase α'(211) varied with crack length a, stress ratio R, and the presence of fretting. The relation between a and Δ(2θ) was similar to the relation between a and the crack propagation rate da/dN. It was shown, however, that Δ(2θ) did not correspond to da/dN owing to the effect of crack surface contact. Optical metallography showed that the martensite transformation at a=2mm for R=-0.33 had occurred already to the depth of several hundreds micronmeter from the crack surface. Although the crack propagation may be affected to some extent by this strengthened zone at the crack tip, the retardation of crack propagation in fretting-fatigue was attributed to the heavy contacting of the crack surface in the stage of slantwise crack propagation.
The combined creep-fatigue loading tests have been carried out on SUS 316 stainless steel at 650°C. The data obtained were evaluated using the linear life fraction damage rule. The results evaluated were discussed on the basis of the observations of cracking and fracture modes. SUS 316 stainless steel under combined creep-fatigue loadings at 650°C was ruptured by an accumulation of either creep damage or fatigue damage. The better evaluation of accumulated creep damage to rupture could be made using a term of ΣNc1εc/(Σεc)cyclic, where Nc is the number of combined creep-fatigue loading cycles to rupture, εc is the creep strain in creep loading Part of combined creep-fatigue loading test, and(Σεc)cyclic is the creep strain to fracture in cyclic creep test. The specimens of which the rupture life was evaluated by the accumulation of fatigue damage were fractured by the growth of fatigue transgranular cracks. On the other hand, the specimens of which the rupture life was evaluated by the accumulation of creep damage were fractured by the growth and linkage of wedge-type intergranular cracks. It is suggested from discussing on the present data and the previously obtained data for SUS 304 stainless steel at 700°C that further investigation is needed concerning the correlation between creep-fatigue interaction and creep fracture characteristics.
Electric melting has been widely applied to glass melting in glass industries. However, there are few experimental data concerning the reaction at electrode-molten glass interface. In this paper, the electrochemical behavior such as anodic dissolution, passivation and cathodic deposition was investigated for nickel and molybdenum in a molten silicate glass of the composition 16Na2O·12CaO·72SiO2 in wt% at 1300 and 1400°C, respectively. Linear potential sweep voltammetry was carried out at various potential sweep rates (5-0.002V/sec) up to ±5V by using a platinum electrode as reference and auxiliary in Ar gas atmosphere. Moreover, the metal-glass interface after polarization was analysed by EPMA. On cathodic potential sweep of Ni, silicate anions were easily reduced to silicon element and a kind of nickel silicide layer covered the electrode surface. Since the nickel silicide has high electrical resistance, its formation made the measurement of voltammogram difficult at slower sweep rates. The anodic voltammogram of Ni at 3.0V/sec showed a large current peak around 1.2V (relative to Pt reference). The peak potential shifted to more cathodic direction with decreasing sweep rate. The current peak was attributed to the formation of a passive layer on the Ni electrode. The occurrence of NiO layer with 15μm thickness on the Ni electrode has been confirmed by EPMA when Ni was oxidized at 3V for 3min. On cathodic polarization of Mo, Mo formed molybdenum silicide layer with high electrical resistance. The thickness of layer and the ratio of Si/Mo in the layer increased with increasing cathodic potential. The pronounced current peak was observed around 0.4V on anodic voltammogram of Mo, followed by a stationary current up to 4V at 3.0V/sec. The current peak was accompanied by a discernible shoulder at 1.0V/sec and split into three peaks at slower sweep rates than 0.05V/sec due to the formation of molybdenum oxides such as MoO2, Mo2O5 and MoO3. At slower sweep rates a stationary current with many small irregular was found, which can be attributed to the volatilization of MoO3. The occurrence of a stationary current indicates that the layer is made up by a stable oxide like MoO2 and the surface of Mo electrode is covered almost perfectly with the oxide layer.
Material damage at high temperatures is being experienced in thermal power and chemical plants during long term service. This problem occurs in both low and high alloy steels of plants where the service temperature is from 450°C to 650°C, and it is suggested that this damage is caused by carbide precipitation in heat resistance steel. The results of the present study are summarized as follows; (1) Material damage (creep damage) during long term service in heat resisting steel is possible to be measured with the electric resistance method. (2) Non-destructive diagnosis for damaged material can be made by determining the ratio of electric resistivity between the unused and used (damaged) materials.
The changes in material properties caused by creep damage were observed with the test methods of electric resistance and hardness. The tested material was Cr-Mo-V steel for rotor in steam turbine. The test specimens were subjected to long term heating and creep damage before testing. The prepared conditions were in the temperature range of 450∼650°C, and the heating or loading period of 160∼8000hours. The results obtained are summarized as follows, (1) Hardness and the ratio of electric resistivity decreased with increasing heating time or temperature. (2) Hardness and the ratio of electric resistivity remarkably decreased by creep damage. (3) Vickers hardness difference ΔHV (HV in the heated specimen-HV in the creep damaged specimen) increased according to the creep damage calculated by the strain fraction rule. (4) Electric resistivity difference ΔRρ(Rρ in the heated specimen-Rρ in the creep damaged specimen) increased with increasing creep damage φCL up to φCL=0.5. (5) It is possible to estimate the material life and creep damage by measuring hardness and electric resistance.
It is known in structural materials that metallurgical structural changes occur and creep or fatigue damage accumulates during long-term use at high temperatures. In order to develop a non-destructive method to detect the creep damage for materials used at high temperatures, the changes in material properties caused by creep damage were observed by the test methods of electric resistance, hardness and sigma phase areas. The tested material was 17Cr-10Ni-0.4Ti (SUS321) austenitic stainless steel. The results obtained are sumarized as follows; (1) The ratio of electric resistivity decreased with increasing creep damage at first, and it becomes minimum at creep damage of 0.2. Then, the ratio of electric resistivity increased with increasing creep damage. The tertiary creep range started from creep damage of about 0.2. (2) Vickers hardness increased with increasing creep damage. (3) The percentage of sigma phase area increased in proportion to creep damage, and reached 5.6∼7.9% at creep damage of 1.0. These results show that the creep damage in austenitic stainless steel can be detected nondestructively by measuring the ratio of electric resistivity, hardness and percentage of sigma phase area.
The creep crack growth ratio and AE count rate of Cr-Mo-V cast steel for steam turbine casting, were measured, and the applicability of AE method to the monitoring of creep crack growth was examined. The results obtained are summarized as follows; (1) AE event count has no good correspondence to AE energy intensity, so that the relation between the stress intensity factor KI and the creep crack growth rate da/dt or AE count rate dN/dt is not linear in the range of high KI. (2) AE ringdown count has a good correspondence to AE energy intensity, so that the relation between KI and da/dt or dN/dt is linear. Accordingly, it is possible to predict the creep crack growth rate of high temperature materials by measuring dN/dt. (3) Frequency spectra of AE during creep crack growth indicate peaks in the range of 110∼300kHz. Accordingly, it is possible to distinguish AE of creep crack growth from steam noise in a steam turbine by analyzing frequency of output or by adapting a high-pass filter of about 120kHz. (4) According to the facts described above, it is found that AE method is adaptable for the fracture mechanical evaluation of creep cracks and the measurement of creep crack growth in high temperature materials.