Within the fields of materials mechanics, the consideration of physical laws in machine learning predictions besides the use of data can enable low prediction errors and robustness as opposed to predictions only based on data. On the one hand, exclusive utilization of fundamental physical relationships might show significant deviations in their predictions compared to reality, due to simplifications and assumptions. On the other hand, using only data and neglecting well-established physical laws can create the need for unreasonably large data sets that are required to exhibit low bias and are usually expensive to collect. However, fundamental but simplified physics in combination with a corrective model that compensates for possible deviations, e.g., to experimental data, can lead to physics-based predictions with low prediction errors, also despite scarce data. In this article, it is demonstrated that a hybrid model approach consisting of a physics-based model that is corrected via an artificial neural network represents an efficient prediction tool as opposed to a purely data-driven model. In particular, a semi-analytical model serves as an efficient low-fidelity model with noticeable prediction errors outside its calibration domain. An artificial neural network is used to correct the semi-analytical solution towards a desired reference solution provided by high-fidelity finite element simulations, while the efficiency of the semi-analytical model is maintained and the applicability range enhanced. We utilize residual stresses that are induced by laser shock peening as a use-case example. In addition, it is shown that non-unique relationships between model inputs and outputs lead to high prediction errors and the identification of salient input features via dimensionality analysis is highly beneficial to achieve low prediction errors. In a generalization task, predictions are also outside the process parameter space of the training region while remaining in the trained range of corrections. The corrective model predictions show substantially smaller errors than purely data-driven model predictions, which illustrates one of the benefits of the hybrid modelling approach. Ultimately, when the amount of samples in the data set is reduced, the generalization of the physics-related corrective model outperforms the purely data-driven model, which also demonstrates efficient applicability of the proposed hybrid modelling approach to problems where data is scarce.
Laser Peening (LP) is a surface modification technology that can induce high residual stresses in a metallic material. The relation between LP process parameters, in particular laser sequences, as well as pulse parameters and the resulting residual stress state was investigated in this study. The residual stress measurements, performed with the hole drilling technique, showed a non-equibiaxial stress profile in laser peened AA2024-T3 samples with a clad layer for certain parameter combinations. Shot overlap and applied energy density were found to be crucial parameters for the characteristic of the observed non-equibiaxial residual stress profile. Furthermore, the investigation showed the importance of the advancing direction, as the advancing direction influences the direction of the higher compressive residual stress component. The direction of higher residual stresses was parallel or orthogonal to the rolling direction of the material. The effect was correlated to the microstructural observation obtained via electron backscattered diffraction. Additionally, for peening with two sequences of different advancing directions, the study showed that the order of applied advancing directions was important for the non-equibiaxiality of the resulting residual stress profile.
1 September 2021 in Journal of Materials Processing Technology
Laser shock peening (LSP) is a surface modification technique to improve the mechanical properties of metals and alloys, where physical phenomena are difficult to investigate, due to short time scales and extreme physical values. In this regard, simulations can significantly contribute to understand the underlying physics. In this paper, a coupled simulation approach for LSP is presented. A global model of laser–matter–plasma interaction is applied to determine the plasma pressure, which is used as surface loading in finite element (FE) simulations in order to predict residual stress (RS) profiles in the target material. The coupled model is applied to the LSP of AA2198-T3 with water confinement, square focus and 20 ns laser pulse duration. This investigation considers the variation in laser pulse energy (3 J and 5 J) and different protective coatings (none, aluminum and steel foil). A sensitivity analysis is conducted to evaluate the impact of parameter inaccuracies of the global model on the resulting RS. Adjustment of the global model to different laser pulse energies and coating materials allows us to compute the temporal pressure distributions to predict RS with FE simulations, which are in good agreement with the measurements.
28 September 2021 in Fatigue & Fracture of Engineering Materials & Structures
This study deals with fatigue crack growth prediction for thin sheet AA2024 specimens with residual stresses introduced through laser shock peening and laser heating treatments. Different weight functions are used to calculate the stress intensity factor due to the presence of residual stresses. A superposition principle is used by calculating the total stress intensity factor considering the applied loads and residual stresses. The fatigue crack growth is predicted using the Paris' law based on the effective stress intensity factor range, in that the effect of residual stresses is considered by changing the total stress intensity factor ratio. It is concluded, that weight functions are a powerful tool to predict fatigue crack growth in AA2024 containing residual stress fields, as long as the gradient in the loading direction, as induced by laser shock peening and laser heating, is moderate.
The refill friction stir spot welding (refill FSSW) process is an innovative solid-state spot-welding method, which has evolved from the concept of friction stir welding. Compared to riveting, the process has the advantage of avoiding stress concentration by eliminating holes. In addition, weight can be saved compared to riveting as no additional material is needed. However, the fatigue strength of refill FSSW joints under cyclic loading is still not satisfactory. To address this challenge, laser shock peening (LSP) is investigated as an innovative residual stress engineering technique to improve the fatigue performance of refill FSSW AA2024-T3 joints. Two application scenarios are investigated, one investigating the LSP technique as a complementary manufacturing process to the refill FSSW technology, and the other investigating the LSP technique as a repair process for damaged joints. The fatigue test results showed that the application of the LSP treatment can significantly improve the fatigue behaviour of the refill FSSW overlap joints. In terms of Basquin fatigue strength, the LSP treatment resulted in an improvement by a factor of 1.51 and 2.82 for the one- and two-sided LSP-treated specimens, respectively. The life of specimens with refill FSSW joints that had been specifically pre-damaged by stopping the fatigue test at approximately 51%, 75% and 83% of the number of cycles to the Basquin fatigue strength, applying LSP treatment and continuing the fatigue test was also significantly extended. The results of this study show that LSP is a very effective technique for significantly extending the fatigue life of refill FSSW joints. Therefore, the combination of these two manufacturing processes, refill FSSW and LSP, represents a promising technology for industrial companies that require high fatigue performance for their structural components.