Filtern
Dokumenttyp
- Zeitschriftenartikel (14) (entfernen)
Sprache
- Englisch (14)
Schlagworte
- Welding simulation (14) (entfernen)
Organisationseinheit der BAM
In various applications, welding-induced residual stresses have a substantial impact on the integrity of welded constructions. Tensile residual stress can promote stress-corrosion cracking, brittle fracture, and reduces the fatigue life in service, as well as influences component design due to critical stress concentrations within the component.
In the present paper, a six bead multi-pass gas metal arc weld of 20 mm thick structural steel S355J2+N is experimentally and numerically investigated. The studies include transient 2D and 3D numerical calculations which consider temperature-dependent material properties, phase transformations, 'thermal' tempering, transformation plasticity, volume change due to phase transformation, an elasticplastic material model, and isotropic strain hardening. The experimentally determined and calculated residual stresses are in a good agreement. Furthermore, the influence of the preheat and interpass temperature on welding-induced residual stresses is shown in the present investigation.
The results of numerical welding simulations strongly depend on its temperature field. In the present paper, the temperature field of a pulsed gas metal arc weld of structural steel S355J2+N (ASTM A572 Gr. 50) with a thickness of 5 mm is experimentally and numerically investigated. In the case of temperature field validation, volumetric Gauss and double-ellipsoid Goldak heat sources are applied. Additionally, different heat source configurations, including adaptations of thermal conductivity, are analyzed regarding their influence on the calculation of welding-induced distortion.
The investigations clarify the influence of heat source configurations on the calculated results, thus, contribute to an improved prediction of welding-induced distortion.
Fusion welding processes are widely used for joining metal structures, such as pipes, ships, and cars. In general, these joining processes offer a very good compromise between reliability, safety, cost and maintenance which are important issues in the current economical context. The negative heat effects of welding, i.e. distortions and residual stresses of the welded parts, are well known and many researches in this field have already been done in the last decades in order to minimize them. On the experimental side, many sophisticated procedures have become state of the art to deal with this problem. On the computational side, the improvement of the simulation algorithms and the computing power enables the simulations of many physical phenomena occurring during the welding process. The implementation of welding simulation techniques is nevertheless not an easy task and often associated with expert knowledge which hinders their global application in an industrial environment. This paper is focused on the industrial requirements of a welding simulation software with special respect to the needs of the automotive industry. The necessary information to run a welding simulation and the expectations of a weld specialist without deep knowledge in numerical methods are investigated. These expectations are tested on an automotive welded assembly with a commercially available welding simulation software designed especially for the needs of the automotive industry. A welding experiment is done and the measured temperature distributions and distortions serve as reference to validate the simulation results. The result quality of the simulations of temperature fields and distortions is in best agreement with experimental data. The workflow is well adapted for the considered industrial requirements and the time-tosolution as well as the computational costs are acceptable, whereas the efficient calibration of the heat input model is still a point which will be further investigated in current and future research works.
A single-layer pulsed gas metal arc weld of structural steel S355J2+N with a thickness of 5 mm is experimentally and numerically investigated. Two tack welds are considered in the numerical simulation into two different ways. First, the tack welds are represented by elements belonging to the initial material. This implies that the 'tack weld material' was not exposed to any thermal load or phase transformation before actual welding was performed. The weld seam is shortened and there is an influence on the stiffness of the whole structure affecting the calculation result. Secondly, the tack welds were simulated as conducted in the experimental welding procedure. The cases considering tack welding are compared to a simulation neglecting tack welding and to the experimental results.
The influence of tack welds on the calculated welding-induced distortion is clarified and a contribution to an improved simulation-based prediction of welding-induced distortion is possible by modeling tack welding according to the realistic fabrication procedure.
The objective of this paper is to demonstrate a new simulation technique which allows fast and automatic generation of temperature fields as input for subsequent thermomechanical welding simulation. The basic idea is to decompose the process model into an empirical part based on neural networks and a phenomenological part that describes the physical phenomena. The strength of this composite modelling approach is the automatic calibration of mathematical models against experimental data without the need for manual interference by an experienced user. As an example for typical applications in laser beam and GMA-laser hybrid welding, it is shown that even 3D heat conduction models of a low complexity can approximate measured temperature fields with a sufficient accuracy. In general, any derivation of model fitting parameters from the real process adds uncertainties to the simulation independent of the complexity of the underlying phenomenological model. The modelling technique presented hybridises empirical and phenomenological models. It reduces the model uncertainties by exploiting additional information which keeps normally hidden in the data measured when the model calibration is performed against few experimental data sets. In contrast, here the optimal model parameter set corresponding to a given process parameter is computed by means of an empirical submodel based on relatively large set of experimental data. The approach allows making a contribution to an efficient compensation of modelling inaccuracies and lack of knowledge about thermophysical material properties or boundary conditions. Two illustrating examples are provided.
Analytical and numerical methods are used to estimate the temperature field due to the heat effects of welding. Numerical techniques are more adapted for industrial complex applications where analytical solutions do not exist yet. However, computational time is much lower with analytical models and a combination of both methods is investigated. Therefore, the two approaches are introduced and confronted in this paper. The finite-element software Ansys has been used for numerical simulations and Scilab for analytical simulations. In order to get a similar result quality, both methods have to be analysed and compared with respect to boundary conditions. These configurations are presented in this paper. Before starting any analysis, the analytical and numerical models have to be comparable. For the numerical simulation, every in- or output is given in discrete form and, for the analytical simulation, in continuous form. Thus, an analysis of the energy input distribution in both models is compulsory to ensure that the same amount of energy is applied. After this first study, a comparison of the analytical and numerical temperature field simulation is done from a fix point source in an infinite volume in steady state to a moving point source in a finite dimension in a transient state. A good agreement between the analytical and the numerical simulation results is found. However, some techniques, like a consideration of an image heat source for the analytical model or the selection of boundary conditions for the numerical model, need to be taken into consideration when the degree of complexity of the study (finite dimension or cooling time) increases. The limit of the comparison is reached when the geometry becomes too complex and when the effect of variable thermal properties with temperature cannot be neglected.
The usage of continuous cooling transformation (CCT) diagrams in numerical welding simulations is state of the art. Nevertheless, specifications provide limits in chemical composition of materials which result in different CCT behavior and CCT diagrams, respectively. Therefore, it is necessary to analyze the influence of variations in CCT diagrams on the developing residual stresses. In the present paper, four CCT diagrams and their effect on numerical calculation of residual stresses are investigated for the widely used structural steel S355J2 + N welded by the gas metal arc welding (GMAW) process. Rather than performing an arbitrary adjustment of CCT behavior, four justifiable data sets were used as input to the numerical calculation: data available in the Sysweld database, experimental data acquired through Gleeble dilatometry tests, and TTT/CCT predictions calculated from the JMatPro and Edison Welding Institute (EWI) Virtual Joining Portal software. The performed numerical analyses resulted in noticeable deviations in residual stresses considering the different CCT diagrams. Furthermore, possibilities to improve the prediction of distortions and residual stress based on CCT behavior are discussed.
A combined experimental and numerical approach is applied for a numerical analysis of the non-uniform martensitic transformation kinetic on welding residual stresses and distortion of a single pass weld. The (γ → α)-transformation kinetic within the weld pool region is governed by a non-uniform distribution of the elements chromium and nickel. The single-pass weld was performed by use of the low-alloyed high-strength steel S960QL with the high-alloyed high-strength filler wire CN 13/4-IG®. A thermo-mechanical FE model of the welding process was experimentally validated against temperature field, solid phase distribution, transformation behaviour, X-Ray stress measurements and transient optical distortion measurements.
The experimentally determined and calculated weld residual stresses and transient distortion are in good agreement. It can be shown that the change on the (γ → α)-transformation kinetic driven by the inhomogeneous distribution of the chemical contents causes a strong influence on the weld residual stresses within the volume of the weld pool, which could promote crack propagation within the solidified weld pool by use of high-alloyed filler materials. Furthermore, a significant influence on the development of the transient welding distortion is visible. This influence should be respected during numerically calculation of welding distortion in case of multi-pass welding using interpass temperatures and high-alloyed filler materials.
Continuous cooling transformation (CCT) behavior affects the transient state of
material properties employed in a numerical welding simulation, having a direct influence
on the developing stress state. Three different CCT diagrams for S355J2 steel
were employed to understand the influence of variations in CCT behavior on the numerical
calculation of welding-induced residual stresses. The CCT diagrams were
constructed from transformation data contained in the Sysweld software database,
measured dilatometric data from Gleeble experiments, and transformation data calculated
from the JMatPro software. The calculated transverse and longitudinal residual
stress distributions provided a qualitative correction only in comparison to experimental
measurements, with the largest deviation occurring near the weld
interface. Overall, the results indicate a weak dependency of the calculated residual
stresses due to anticipated CCT variations. The most significant effect on the calculated
residual stresses was shown to be related to the proportion of formed martensite.
It is suggested that CCT data of approximate accuracy is sufficient for reliable
calculation of welding-induced residual stresses.