Search Results

Various prediction methods have been proposed for evaluating the fatigue life of welded joints by combining finite element analysis (FEA) with an experimental database. However, to obtain more universal and accurate fatigue life predictions, it is necessary to have criteria for making integrated evaluations of the fatigue strength of welded joints. This paper presents a study that focuses on the local cyclic plastic zone size (ω*) as the criterion of fatigue strength and investigates its validity. The definition of ω* was given by the relationship between the stress state at the notch tip and the elastic strain which was defined along the strain-life fatigue curve (ε - N diagram) of a base metal. As a result of using ω*, it was found that an integrated fatigue life prediction was possible to a certain extent for notch and arc-welded joint specimens.

The study by the “Committee on Fatigue Strength & Structural Reliability “of the Society of Automotive Engineering of Japan (JSAE) reveals that the method of fatigue life prediction based on cyclic plastic zone size (ω*) which proposed by Koibuchi at el.(1) (2) (4) (5) have accurate prediction ability for the fatigue life of arc-welded joint. But in this method solid elements with sides smaller than 0.001mm long for calculation of ω* is required and it is not practical to apply for all the arc-welded parts in an entire vehicle. Then, The method of simply calculating ω* etc were examined, and fatigue life prediction method based on ω* for arc-weld structures which used virtual prototype vehicles was developed. Moreover, the validity of the method was verified from the comparison with the fatigue test result.

Two working groups in the JSAE Committee of Fatigue–Reliability Section1 are currently researching the issue of fatigue life by both experimental and the CAE approach. Information regarding frequent critical problems on arc–welded structures were sought from auto–manufacturers, vehicle component suppliers, and material suppliers. The method for anti–fatigue design on arc–welded structures was established not only by a database created by physical test results in accordance with the collected information but also with design procedure taking Fracture–Mechanics into consideration. This method will be applied to vehicle development as one of the virtual laboratories in the digital prototype phase. In this paper, both the database from bench–test results on arc welded structures and FEA algorithm unique to JSAE are proposed some of the analysis results associated with the latter proposal are also reported.

A method of fatigue life estimation for the spot-welds of vehicle body structures by means of Finite Element Analysis (FEA) was studied. 6 general forces applied to a nugget of spot-weld under multiaxial loads were determined and the Nominal Structural Stress (σns) was calculated from them. It was confirmed that fatigue strength of the spot-welds under various multiaxial loads could be estimated universally by using σns. Based on the theory of elasticity of plates, stress of spot-weld nugget was analyzed. The theoretical equations for determining the principal stress at the nugget edge from6 general forces acting on a nugget were derived. And the principal stress was defined as the σns. The value of σns was determined by FEM that used a solid model and compared with the theoretical calculation value. They agreed quite well. Fatigue tests of DC specimens under various multiaxial loads (shear plus cross tension and tensile shear plus torsion) were conducted.

Effects of various scatter factors on fatigue strength for arc-welds were studied by fatigue test results and sensitivity analysis using the method of cyclic plastic zone size (ω*). The followings were clarified. First, effects of flank angle could be decided by the sensitivity calculated based on the relation between fatigue life for finite life and flank angle. Second, effects of material could be explained by the sensitivity for which the difference of fatigue strength coefficient for each material was analysed. The results, it was verified that there was no effect in notch specimens and there was effect in smooth specimens. Third, effect of thickness was defined by function of the ratio of thickness.