Search

Effects of metal cations on the corrosion behavior of carbon steel SM490Y in 10 mM Cl- model fresh water were investigated by immersion tests, electrochemical tests, and surface analysis. Immersion tests and electrochemical tests results showed that Zn2+ had a significant inhibition effect on corrosion of SM490Y. The results of EDS and XPS showed that Zn2+ was incorporated in the oxide films by making a strong bond and formed zinc hydroxide or zinc oxide which can cover the surface of carbon steel. It is suggested that Zn2+ forms a protective film which can efficiently prevent break down of surface film by Cl- and inhibit the corrosion.

In order to clarify the effect of environmental factors on the atmospheric corrosion of steel, novel model for predicting the reduction of atmospheric corrosion considering relative humidity and rain falls was developed. We conducted a one-year calculation simulation of atmospheric corrosion in Miyakojima City, Choshi City, and Tsukuba City using the developed model. Corrosion weight loss by the simulation could reproduce the measured value well. Corrosion weight loss was greatly affected by the amount of airborne sea salt, relative humidity, and rain falls.

Raman micro-spectroscopy was conducted for in-situ observation of atmospheric corrosion occurring on carbon steel surfaces, on which NaCl nanoparticles with a size of < 1 mu m were deposited, under the controlled relative humidity at room temperature. Increasing relative humidity induced the formation of amorphous-FeOOH at a micro-droplet of NaCl solution where an NaCl nanoparticle was dissolved on the steel surface, and the initiation time of the corrosion generation decreased with increasing relative humidity. Drying treatment at 100 degrees C for 600 s resulted in modification of the surface condition of carbon steels and obstructing the corrosion initiation at high humidity conditions but reducing the initiation time at low humidity conditions.

塩化物を含む硫酸ミスト環境において，NiSO4含有樹脂を被覆した炭素鋼の耐食性を評価した．形成した腐食生成物の構造を解明するため，放射光を用いたXRDおよびXAFS解析を実施した．NiSO4により，ゲーサイトおよびアカガネアイトの形成が促進された．このアカガネアイトは正方晶系のβ-FeOOHではなく，Niを含有した単斜晶系のアカガネアイトであると考えられる．, Corrosion resistance of carbon steel covered with resin coating containing nickel sulfate has been evaluated under chloride and sulfuric acid mist environment. The structure of corrosion products formed on steel surface was investigated by XRD and XAFS analyses using synchrotron radiation. Nickel sulfate promoted the formation of goethite and akaganeite. It was considered that this akaganeite was not tetragonal β-FeOOH but monoclinic akaganeite containing nickel.

Influence of static load on torsional fatigue strength was investigated on smooth hollow specimens of carbon steel S45C and austenitic stainless SUS316L. Fatigue life under completely reversed cyclic torsion can evaluate without depending on a control system. Addition of static load had a little effect on the torsional fatigue life under deformation-controlled test. However, fatigue life decreases under the load-controlled remarkably. The reason is accumulation of ratcheting and variation of stress-strain relation, difference of fracture surface by addition of static load.

資料番号: SA1005123000

Accepted: 2013-04-23, 資料番号: SA1004700000

We have observed the cathodically polarized specimen surface of S25C and S50C carbon steels, SUS304 stainless steel and pure platinum by using scanning electron microscope. The carbon steel specimens were heat-treated at 100～900℃ after cold-rolling. When these carbon steel specimens were polarized cathodically in an aqueous solution of 0.5kmol/m3 H2SO4, needle-shaped deposits were observed on the specimen surface regardless of heat-treatment temperature. In the specimen of SUS304 stainless steel, slight corrosion was observed, but the deposits were not observed. In the specimen of platinum, a change of the specimen surface was not observed. These experimental results indicate that the deposits observed in the carbon steel specimens are corrosion products, and less noble metals corrode even when they are polarized cathodically.

We have observed the cathodically polarized specimen surface of S25C and S50C carbon steels, SUS304 stainless steel and pure platinum by using scanning electron microscope. The carbon steel specimens were heat-treated at 100～900℃ after cold-rolling. When these carbon steel specimens were polarized cathodically in an aqueous solution of 0.5kmol/m3 H2SO4, needle-shaped deposits were observed on the specimen surface regardless of heat-treatment temperature. In the specimen of SUS304 stainless steel, slight corrosion was observed, but the deposits were not observed. In the specimen of platinum, a change of the specimen surface was not observed. These experimental results indicate that the deposits observed in the carbon steel specimens are corrosion products, and less noble metals corrode even when they are polarized cathodically.

著者らが提案した，レーザを用いた被削材の部分的熱処理によるバリの抑制方法について，より効率的なバリの抑制を実現するために，フライスカッタの出口角度，送り速度，切削速度およびレーザ照射強度が切削後の端部形状におよぼす影響を調査した．その結果，切削後の被削材端部の形状は，カッタ出口角度と送り速度によって定まる切取り厚さと，カッタ出口角度と切削速度によって定まる熱処理部の切削時間に依存することが明らかとなった．また，レーザ照射条件を制御して深いマルテンサイト層を生成することによって，より確実にバリの発生を抑制することができることが明らかとなった．

Effects of titanium addition on as-cast austenite grain size in 0.2 mass% carbon steel cooled at 0.03°C/s has been studied. The average austenite grain size was reduced from 4.3 to 1.9 mm by increasing the titanium concentration from 0 to 0.13 mass%. Two different shapes of Ti(C, N) particles were observed in the solidified samples; one was a facet shape which were dispersed over the whole observation area and the other was a filmy shape, most of which were located on the austenite grain boundary. Thermodynamic calculations and EPMA analyses showed that the facet shaped Ti(C, N) particles crystallize at relatively high temperatures, while the filmy shaped ones form from the last solidifying liquid in the inter-dendritic regions at lower temperatures. The latter Ti(C, N) is considered to provide the grain refinement effect of austenite grain structure. The grain refinement effects of titanium are further discussed by comparing the results of 0.2 mass% carbon steel with those of 0.45 mass % one, which supports the substantial effect of the filmy shaped Ti(C, N) on the as-cast austenite grain structure.

本研究は，炭素鋼S45Cの旋削加工においてレーザ熱処理が切りくず処理性や仕上げ面粗さへ及ぼす影響について検討している．実用的な見地から小型の半導体レーザをCNC旋盤の刃物台に設置し，機上にて工作物長手方向にレーザを照射して選択的に熱処理を行っている．工具は超硬P30種とアルミナ-炭化チタン系セラミック工具を用い，すべての実験でブレーカピースを併用している．レーザ熱処理を行わない通常の旋削では低切削速度と高送りの場合にのみ切りくずが折断されるが，レーザ熱処理した炭素鋼を旋削すると広範囲の切削条件において切りくずが折断され，被削性が向上する．レーザ熱処理された炭素鋼の切りくず観察より，連続型切りくず中に変形形態が異なる部分が確認され，この部分を起点として切りくずが折断されている．このとき，超硬P30種およびセラミック工具いずれの場合においても仕上げ面粗さはレーザ照射なしの場合と比べてほとんど同じか若干傾向にあり，硬化したレーザ熱処理部を周期的に切削することによる仕上げ面の劣化は引き起こされない．

本研究では，炭素鋼の正面フライス加工において，加工能率を維持したままバリの抑制を図る手法として，レーザを用いた被削材の部分的な改質による方法について提案した．実験では炭酸ガスレーザを用いて選択的に熱処理を行った炭素鋼JIS S45Cの正面フライス加工において，切削後の被削材端部の形状についてバリの高さの測定と断面の組織観察を行い，被削材へのレーザ熱処理によるバリの抑制効果について検討を行った．切削後における被削材の端部の形状はレーザ照射条件に依存し，表層にマルテンサイト組織が生成されていると微小なコバ欠けが生じる．またマルテンサイト組織が生成されていなくても，レーザ熱処理により硬度が母材よりも高くなっていると発生するバリの高さが小さくなる．最終切取り予定線へレーザ熱処理を行えば, それ以前の切削でバリが発生していても最終切削においてバリの発生を抑制することができる．仕上げ面粗さは，レーザ熱処理の影響をほとんど受けない．

Effects of Ti addition on as-cast austenite grain size in S45C carbon steel have been investigated in range of Ti addition between 0 and 0.5 mol%. The cooling rate was set to be 0.03°C/s. It is shown that the average austenite grain diameter decreases from 1900 to 250 μm as the Ti addition increases from 0 to 0.5 mol%. The microstructural observation revealed that the motion of austenite grain boundary is inhibited at inter-dendritic position, which is ascribable to pinning effect of Ti(C,N) particle. The CALPHAD analysis showed that this particle crystallizes mainly in L+γ−Fe+Ti(C,N) phase field. As increasing Ti addition, the size of Ti(C,N) particle does not substantially change, however, the number of the Ti(C,N) particles increases, leading to further refinement of the austenite grains.

We performed casting experiment of S45C steel, focusing on the effect of addition of Ti and B on as-cast austenite grain structure. Without the addition of these elements, the austenite structure consists of columnar grains over the whole sample in the present casting conditions. The addition of Ti and B leads to the formation of equiaxed austenite grains and, importantly, fully equiaxed austenite structure was observed in the sample with 0.2 mol% Ti and 0.4 mol % B. The microstructural observations indicated that this behavior of austenite structure stems from the columnar-to-equiaxed transition of ferrite dendrite structure. The addition of these elements, furthermore, leads to refinement of the columnar austenite grains

The R-curve method for predicting the fatigue thresholds of notched components was applied to the fatigue thresholds of notched components under combined loading of cyclic torsion and tension-compression. The prediction was compared with the experimental data obtained from thin-walled tubular specimen with a hole under the combination of cyclic torsion and axial loading. The effect of the hole size on the fatigue thresholds was calculated from the R-curve method for three cases: cyclic torsional stress, cyclic torsional stress with superposition of a in-phase cyclic tension-compression stress with the same amplitude, cyclic axial tension-compression stress. The stress ratio was R=-1. The experimental data agreed well with the prediction both for crack initiation and fracture. The measured length of nonpropagating cracks had some scatter and the maximum length agreed fairly well with the predicted line. The nonpropagating crack length normalized with the hole radius at the threshold stress for fracture was predicted fairly constant without respect to the hole size, while it varied slightly with the loading condition. The effect of the in-phase combination of axial and torsional stress loadings on the fatigue threshold was predicted by assuming the crack direction perpendicular to the maximum principal stress. A good agreement between predicted and experimental results was obtained.

application/pdf, 論文(Article), Machines and components in operation are usually subjected to varying loads. Despite this, wear testing has usually been conducted under constant loads. In this study, in order to investigate the ef­feels of changes in loading on the transition from mild to severe wear, pin-on-disk type wear tests of carbon steel in contact with itself were conducted in moist air under dry sliding. Dming wear testing, the contact load was changed in aヽtep-wisemanner just one time. from low to high levels. A longer sliding distance under mild wear conditions at low load in the first stage produces a more flattened and oxidized sliding surface with good wear resistance. In other words, a "quasi-mild wear''regime with low rate occurred, sinular to the 1ruld wear. even at high loads in the second stage. The critical sliding distance Ls,crtical in the first stage necessary for the transition to quasi-mild wear in the second stage is given by an empirical formula. The proposed formula indicates that Ls,crtical decreases with an increase in the first-stage load and a decrease in the second-stage load. In conclusion the quasi-nuld wear can be attained under sh01tcr sliding distance at higher loads in the first stage.

In general, surface should be cleaned when measuring surface hardness. The relationship between hardness and surface uncleanness, however, has scarcely been researched. In the present study, the influence of uncleanness on the hardness of carbon steel is investigated; the Vickers hardness of a cleaned surface is compared with that of a surface lubricated with liquid paraffin and/or liquid paraffin containing 3 % oleic acid, and then discussed. An unclean surface was found to have a detectable influence on hardness, for example, the hardness of semi-hardened steel, JIS-S45C of 205 Hv for a clean surface decreased to 187 Hv when lubricated with liquid paraffin.

Load-controlled fatigue tests were conducted on smooth tubular specimens of an annealed medium carbon steel under fully-reversed torsion with superimposed static torque or tension. Effect of mean stress on initiation and growth of short crack was studied in the intermediate life regime. Microcracks initiated around the maximum shear stress amplitude planes rather than the planes of maximum shear stress, irrespective of mean stress conditions. The dominant crack grew accompanying coalescence of the shear cracks initiated in the early stage of life and the preferential growth direction was different depending on the type of mean stress. The increase of the crack growth rate and the reduction of the fatigue life with application of mean stress were slight in the present material. The relatively smaller mean stress effect appeared to be resulted from the decrease of the plastic shear strain amplitude with static loading. Therefore, two-stage short crack growth model proposed by Hobson and Brown was modified incorporating the effects of mean stress acting on the crack planes and the hardening/softening of material. The model showed a satisfactory estimation of fatigue lives in torsion under various mean stress conditions.

金沢大学高度モビリティ研究所 / 金沢大学理工研究域機械工学系, A practical model of a temperature and strain-rate dependent dynamic constitutive equation for carbon steels is proposed. Unknown material constants involved in the proposed constitutive equation can be determined by the non-linear least squares method, which approximates ordinary-impact-test results with calculated values of elasto-plastic waves propagation analyses. In addition static stress-strain curves, longitudinal elastic modulus and yield stress, which are needed for elasto-plastic waves propagation analyses with the proposed constitutive equation, are expressed as a function of temperature. These results enable elasto-plastic waves propagation analyses under wide strain-rate and temperature ranges easily. The present paper has actually determined material constants involved in the proposed constitutive equation and compared the calculated results with measured ones. As a result, it has been confirmed that the proposed equation is applicable at a wide strain rate range and at a room temperature to less than one in which blue brittleness occurs.

The phases and microstructure of the alloy layer formed during hot dipping of carbon steels in molten Zn-Al bath were investigated. Pure iron, 0.2%C and 0.82%C steels and 2.45%C cast iron were used. A plate shape specimen of 300μm in thickness was firstly hot dipped in a molten Zn bath for 3s and then hot dipped in a molten Zn-6mass%Al alloy bath for certain times, and quenched with water. The microstructure of the quenched specimen was observed by using a SEM, and change in thickness of the specimen was measured by using a laser microscope. The pure iron specimen was entirely changed to an intermetallic compound of Fe2Al5(Zn) phase after 1.2ks hot dipping. The thickness of 0.2%C steel specimen was decreased linearly with the dipping time after an early non-reaction period, and 2/3 of the specimen was changed to the Fe2Al5(Zn) phase after 3.6ks hot dipping. It was observed that primarily precipitated ferrite phase preferentially reacted and pearlite structure scarcely reacted. The reacted thickness of the 0.82%C steel and cast iron specimens were only several μm after 86.4Ks hot dipping. In the cast iron specimen, cementite reacted with Al and an Al3C4 phase was formed. The growth of the alloy layer was restricted by the Al3C4 phase. A heat-treated 0.2%C steel specimen with sorbite structure also exhibited the same low reaction behavior as the 0.82% C steel due to homogenized ferrite + cementite structure.

Hardening charactertic of carbon steels have been discussed by comparing the flow curves with the shape and distribution of cementite. The complicated microstructure state induding lamellar or granular cementite has been quantitatively estimated with one parameter, Sv, in stereology. Magnitude of intemal structure variable, A∞,capable to express hardening ability has been eva1uated from each experimental flow curve in accordance with the random barriぽ'stheory. A good correlation between Sv and A∞has been found. It is clarfied that hardening ability of lamellar cementite is larger than that of granular cementite. These results suggest that the random barriers theory can be applicable for describing any

Fretting fatigue tests of carbon steel coated with titanium nitride (TiN) by PVD method were carried out to discuss the effect of ceramic coating on fretting fatigue behavior. Fretting fatigue strength of the present carbon steel significantly increased by TiN coating onto the specimen. The fretting fatigue fracture process in TiN coated steel was as follows. Fretting cracks initiate at a certain stage of fatigue life from the flaws or worn region in TiN film initiated by fretting action, and propagate until final unstable fracture occurs. The improvement of fretting fatigue strength by TiN coating resulted mainly from the retardation of fretting fatigue crack initiation due to existence of hard TiN film on the contact surface of the specimen.

Changes in fractions of δ-ferrite, austenite and liquid phases during peritectic solidification in iron-carbon system are calculated. The peritectic solidification, δ+L→γ, is divided into δ→γ transformation at δ/γ interface and L→γ solidification at γ/L interface. Both the transformation and solidification proceed by (1) carbon diffusion from liquid through austenite into δ-ferrite and (2) precipitation of austenite from δ-ferrite and crystallization of it from liquid due to cooling. Approximately 80% of the austenite formed during the peritectic solidification is the product of δ→γ transformation, which may lead to the generation of tensile stress in the solidification shell because of the difference in density between the two solids. The amount of the δ→γ transformation is largest, when the initial carbon content is 0.17 mass%. However, when the transformation in a well-developed dendrite network is focused on, the carbon content for the largest amount of the transformation decreases to 0.14-0.16 mass%, which corresponds to the carbon content at which surface cracking of continuously cast slabs is reported to be most frequent.

Chromium nitride (CrN) has good corrosion resistance in addition to its good electrical and mechanical properties. A decrease in the number of defects such as pinholes or voids formed during deposition is required for improvement of corrosion resistance and corrosion fatigue strength of coating materials. A multistage physical vapor deposition (PVD) method was proposed in order to fabricate dense thin films. In an anodic polarization test in aerated 2%H_2SO_4 solution, CrN thin film deposited on 0.37 wt.%C steell using the multistage PVD method showed better corrosion resistance than that of the specimen coated using a conventional method. This is due to a decrease in the number of small defects in coating film by the multistage deposition. Evaluation of corrosion fatigue strength in the multistage coating steel was performed by means of a cantilever-type rotating-bending fatigue test in 3%NaCl solution. Corrosion fatigue strength of the multistage coating steel was improved as compared with that of the uncoated steel, but the increase in fatigue strength was less in the multistage coating method than the conventional method, contrary to the experimental results obtained from the anodic polarization test. It was pointed out that corrosion fatigue resistance is dependent on not only the number of defects but also the shape of large defects in the coating film.

A high temperature stress-strain testing machine capable of operating with gradient temperature applied to test-pieces has been developed to simulate thermomechanical properties of solidifying shell formed at the initial stage of solidification. Optimized design of induction heating coil has been elaborated for materializing desired temperature gradient within a tensile test-pieces, resulting in a sophisticated coil shape which has been found satisfactory. The results obtained from this study are compared with mechanical properties obtained from previous hot tensile tests and "in-situ" measurements of tensile strength of solidifying shells.

Rotating bending fatigue tests were carried out on the circumferential notched specimen of three kinds of carbon steel with nearly equal ferrite grain sizes of 20μm but different volume fractions of pearlite. Fatigue tests were also carried out for 0.15% carbon steel with ferrite grain size of 50μm and 0.45% carbon steel with ferrite grain size of 5μm. The effect of microstructures on the fatigue strengths σ_ (the limit stress for slip band), σ_ (the limit stress for macrocrack initiation), and σ_ (the limit stress for fracture) is discussed by comparing these values between three kinds of carbon steel. The main results are as follows : (1) Since σ_ is localized to a highly specified region, variation of K_tσ_ is small irrespective of notch root radius. (2) If the grain size of ferrite is nearly equal, variation of σ_ is small irrespective of pearlite volume fraction under the certain amount of pearlite. (3) In medium carbon steel, σ_ increases with increasing grain size of ferrite. (4) Notch root radius at the branch point ρ_o varies with the grain size of ferrite.

To clarify the effect of flaws in coating film on fatigue strength, cantilever-type rotating bending fatigue tests were conducted in air and in saline solution (3.0%NaCl) using the specimens of 0.37% C steel with flaws in titanium nitride (TiN) thin film coated by PVD and CVD methods. Flaws in the coating film on specimen surface were introduced by the application of 1.1～1.6% static tensile strain before the test. The obvious decrease in fatigue life of the coated specimen with flaws was observed in both environments, as compared with that of an uncoated specimen and coated one without flaws. This behavior occured markedly at the fatigue in air, that is, the decrease of fatigue life is 90～75% in air to 70～50% in saline solution. Though film thickness is 3～5μm, the flaw in the film acts as a notch effect for crack initiation on substrate. Many cracks were induced in the substrate directly under a flaw and formed a large crack by coalescence at the early stage of fatigue in air. In corrosion fatigue, corrosion pits at which crack initiates occur on the substrate under a flaw at the early stage of fatigue process and the incubation period prior to formation of pits disappears.

Fe-0.51wt%C alloy was cooled from 1000°C at 0.3°C/s and rolled at 710°C, at which austenite was super-cooled by 55°C. The isothermal transformation behavior at 710°C from the super-cooled and strained austenite to ferrite was investigated. It has been found that the nucleation rate of the ferrite particles increases with strain and that the rate constant for the growth of the ferrite particles remains constant regardless of the strain. The time required for the transformation and the average size of the ferrite particles decrease with increasing strain.

A strain-controlled low-cycle fatigue test was conducted in air, using the specimen of 0.37% carbon steel coated with TiN by the physical vapour deposition (PVD) method. From the experimental results, the increase in fatigue strength of the coated specimen was observed under the region of low total strain amplitude, as compared with that of an uncoated specimen, because crack initiation was delayed by the hard coating film on the specimen surface which acted as a barrier to the egress of dislocations. On the other hand, fatigue strength of the coated specimen decreased at the rigion of high strain amplitude. Many cracks were induced at the substrate by flaws in the coating film. Flaw in the coating film on the specimen occurred at the static tensile strain of 0.34～0.40 % and density of flaw increased with tensile strain. The value of strain amplitude at the reversal of fatigue strength corresponds to the tensile strain at flaw initiation of the coating film.

The corrosion fatigue process under cyclic torsion was investigated in annealed carbon steel. The results obtained were as follows. (1) During the fatigue process in laboratory air and salt water at the high stress amplitude region, slip bands appeared on the smooth specimen surface and cracks initiated from some of these slip bands. These cracks propagated in the direction of the specimen axis at mode II and thereafter branched and propagated in the direction inclined 45° to the specimen axis at mode I. (2) During the corrosion fatigue process at the low stress amplitude region, cracks initiated at corrosion pits and directly propagated in the direction inclined 45° to the specimen axis at mode I. (3) In torsional fatigue tests, the decreasing rate of fatigue strength due to the corrosive environment was smaller than that of rotating bending tests. (4) Differences in the initiations and growth behavior of corrosion pits between the torsional corrosion fatigue test and the rotating bending corrosion fatigue tests were not observed.

In order to clarify the effects of the fluid flow rate and stress amplitude on the the initiation and growth behavior of corrosion pits, round, smooth specimens were rotated with various speeds in sodium chloride aqueous solution. The results are summarized as follows : [1] The pit growth behaviors were prescribed by Eq. (1), regardless of fluid flow rates and stress amplitudes. [2] The pit initiation time decreases with increase of the fluid flow rate and stress amplitude. [3] For small pits whose sizes are about 0.03mm, the pit growth rates increase with the fluid flow rate, while the growth rates for pits above 0.08mm are independent of the fluid flow rate and smaller than those of 0.03mm. The pit growth rate in the case of 100MPa is 5～10times lager than those of nonstressing.

Corrosion fatigue behaviors of carbon steel specimens having different diameters were investigated experimentally in a sodium chloride aqueous solution. Based on these experimental observations Monte Carlo simulations of the corrosion fatigue process of different-sized specimens were conducted. The results of the simulations showed a good agreement with the experimental results. At a constant fatigue life ratio N/N_f, there was no specimen size dependence in the distribution of nondimensional crack lengths F(2l/d) and, at constant stress cycles, the distribution of crack lengths F(2l) is nearly constant regardless of the variations of the specimen diameters. Based on the above results, the simple estimation method of corrosion fatigue lives for the different-sized specimens was proposed. The estimated results give a good agreement with the experimental results.

Steels containing 0.01 to 0.49 mass% carbon were melted and then cooled at rates of 0.3 to 40°C/s to investigate the condition favorable for the formation of columnar austenitic grains. The results were summarized as folllows. 1) Large columnar grains were most remarkably developed in 0.1-0.3 mass%C region. 2) The region of the carbon content favorable for the development of columnar grains extended with increasing coiling rate. The computer simulation of the cooling process of the ingots was carried out to investigate the formation mechanism of the coulmnar grains. From the experiment and the simulation, it was concluded that the formation of columnar austenitic grains was attributable to the steep temperature gradient at the δ→γtransformation front.

In order to investigate variations of spatial distributions of cracks observed on the unnotched specimen surface during the fatigue process, plane bending fatigue tests were performed on carbon steel JIS, SS 41 in laboratory air and a sodium chloride aqueous solution. Two methods, developed by Morishita and Ono in the field of mathematical ecology, were used in analyzing the spatial distributions of Cracks. The following results were obtained. The spatial distributions of cracks are influenced by both stress amplitude and stress cycling. In the high stress amplitude region, the distributions are nearly random, while in the low stress amplitude region, the distributions have clumps indicating nonrandom distribution. This result was common to both fatigue in air and corrosion fatigue. With increasing stress cycling, the spatial distributions of cracks approach random distributions in the high stress amplitude region, while in the low stress amplitude region, the distributions show a strong tendency to have two kinds of clumps whose sizes are 1/4 mm^2 and 8 mm^2.

The interaction of crack propagation on multiple ended fatigue cracks in hybrid structures is examined with the coupled compact tension specimens of 0.45%C carbon steel and A2017-T4 aluminum alloy under high cyclic stresses. The crack propagation behavior of coupled specimens fatigued under the stress ratio of 0.05 and 0.7 is compared with that of single cracked specimen. The deviation of crack propagation rate from the stable relation between the rate and ΔK of a single specimen is found to take place with the difference of crack closure and the onset of the ratcheting extension of specimens. The crack growth acceleration is affected by another specimen coupled because of the incomplete reversal of plastic deformation in ratcheting deformation. Based on the theoretical prediction of an interaction curve for crack growth of a coupled specimen, it is concluded that multiple ended fatigue cracks under high cyclic stresses propagate with a mechanism such that the increment of total strain energy release per cycle in a hybrid body including some cracks is distributed to each crack with the ratio of strain constraint factor.

Interaction of crack growth in hybrid bodies subjected to cyclic loading is investigated using the coupled compact tension specimens model. Coupled CT specimens of 0.45% C steel, JIS S45C and aluminium alloy, A2017-T4, individually precracked to their initial lengths are loaded by common pins in the grip of a fatigue machine. The crack propagation behaviour of coupled specimens fatigued under a stress ratio of 0.05 is compared with that of a single specimen. It is found experimentally that crack growth rate of one of the coupled specimens is accelerated by the plasticity induced crack closure of another specimen. This behaviour is affected by the difference in residual plastic stretch on the crack wall in each specimen. An interaction curve for coupled hybrid specimens is theoretically predicted, and the interaction law is discussed. It is concluded that multiple ended fatigue cracks may propagate with a mechanism such that the increment of total energy release rate in a hybrid body is distributed to each crack with the ratio of the strain constraint factor.

The distribution of corrosion fatigue cracks observed on the un-notched round specimen surface differs with specimen diameter, especially in the low stress amplitude region. At a constant fatigue life ratio, many long cracks are initiated on the larger specimen, 12 mm (diameter), in comparison with the smaller specimen, 6 mm (diameter). On the other hand, in the high stress amplitude region of corrosion fatigue and fatigue in laboratory air, the distribution of cracks during the fatigue process does not depend on specimen diameters. In the low stress amplitude region of corrosion fatigue, crack growth rate does not show the usual behavior of the dependence of crack length, but shows only the dependence of stress amplitude. This unusual crack growth behavior explains why the distribution of corrosion fatigue cracks in the low stress amplitude region changes with specimen diameters.