Optimal Design for Metal Additive Manufacturing: An Integrated Computational Materials Engineering (ICME) Approach

Abstract

We present our latest results on linking the process-structure-propertiesperformance (PSPP) chain for metal additive manufacturing (AM), using a multi-scale and multi-physics integrated computational materials engineering (ICME) approach. The abundance of design parameters and the complex relationship between those and the performance of AM parts have so far impeded the widespread adoption of metal AM technologies for structurally critical load-bearing components. To unfold the full potential of metal AM, establishing a full quantitative PSPP linkage is essential. It will not only help in understanding the underlying physics but will also serve as a powerful and effective tool for optimal computational design. In this work, we illustrate an example of ICME-based PSPP linkage in metal AM, along with a hybrid physics-based data-driven strategy for its application in the optimal design of a component. Finally, we discuss our outlook for the improvement of each part in the computational linking of the PSPP chain.

Post-processing of additively manufactured metallic alloys – A review

“…This fundamental approach can be used to explain the reasons behind the contradictory observations reported in literature regarding the material response during different post-processing operations. Further research on developing physics-based and microstructure-sensitive modelling platformsin line with the Integrated Computational Materials Engineering (ICME) paradigm [255,256] is deemed essential for machinability assessment in view of a large spread in achievable microstructural characteristics and physical properties of AM materials. Fig.…”

Controlling microstructure and mechanical properties of additively manufactured high-strength steels by tailored solidification

“…In order to simulate solute partitioning and thermodynamic driving forces for phase transformations under consideration of the specific temperature profile during LPBF, the phase-field solver was coupled with the Thermo-Calc mobility database MOBFE4, as well as a simulated temperature profile. The temperature profile was extracted from experimentally validated melt pool simulations using a finite-element approach [28,29]. For simplification, a 1D-temperature profile from the middle of a melt pool cross-section was used for the 2Dphase-field simulations.…”

“…This work aims to establish an alternative approach for microstructural control towards grain refinement and texture randomization by modification of the solidification sequence in multi-phase steels. In our previous studies [25][26][27][28][29][30], we identified high-manganese steels (HMnS) as promising alloys for AM. Firstly, processing challenges during conventional processing can be circumvented by LPBF [25].…”

Computational modelling of process–structure–property–performance relationships in metal additive manufacturing: a review

“…Since developing and using common ICME-based models are computationally time-consuming, expensive, and complicated, Motaman et al, [279] have proposed a hybrid ICME-based data-driven model as a performance-oriented design to optimise the metal AM process, as seen in Figure 26. It consists of the following step: (1) Utilising design of experiments (DOE) to understand the effect of processing parameters of part properties, (2) predicting the part performance via process parameters using ICME-based PSPP relationships, (3) evaluating the inaccuracy in the ICME-based PSPP relationships in comparison with experimental PSPP relationships, (4) training a data-driven model using the data obtained from physics-based PSPP relations, (5) using trained data-driven model and creating design-predict-optimise cycle to predict part performance, and (6) model validation with experimental data [279].

Figure 26. Illustration of hybrid ICME-based data-driven model as a performance-oriented design to optimise the metal AM process [279].

…”