Multidisciplinary optimization of automotive mega-castings merging classical structural optimization with response-surface-based optimization enhanced by machine learning

Abstract

Large high pressure die castings (HPDC), recently referred to as mega-castings, can replace plenty of steel metal sheets usually employed for body-in-white (BIW) structures. They can save manufacturing expense and unleash additional lightweight potential thanks to additional design freedom and material properties. The BIW plays a major role in automotive design since it must fulfill numerous structural targets ranging from stiffness for vehicle dynamics, dynamic responses for NVH (noise, vibration, harshness), driving comfort standards and several passive safety requirements. The use of mega-casting structures leads to additional requirements with respect to castability and material quality. Achieving a lightweight design considering requirements related to crash or castability is a challenge on its own, due to the high computational cost of related simulation techniques. Considering multiple requirements simultaneously, therefore often leads to non-weight-optimal structures. To exploit the full lightweight potential, we present a generative multidisciplinary optimization pipeline for the structural design of automotive mega-casting parts in this paper. The approach combines established methods in automotive industry such as topology optimization and response-surface-based (RSM) optimization and enhances the latter by machine learning (ML) based clustering and classification. In a first step topology optimization is employed to derive optimal load-paths for multidisciplinary loading conditions. For this purpose, casting manufacturing constraints as well as more than hundred linearized loads are used to incorporate NVH and passive safety requirements. In a next step the optimal thickness distribution and rib orientation of the structure is achieved using RSM optimization algorithms for the computationally expensive nonlinear crash and casting simulations. Performance indicators are treated by unsupervised learning based on clustering. This enables classification constraints based on simulation field results from hundreds of samples to be included into RSM optimization. It resolves a typical risk of pure scalar, regression-type targets, where supposed optimal results fail when domain experts examine the full field result of the corresponding simulation. It is shown how this approach is superior in achieving a weight-optimal design and turnaround time compared to a design workflow classically used for BIW structures.

Figure 1

Classical design approach frequently used for the design of BIW structures.

Figure 2

Large casting component (orange) in BIW structure.

Figure 3

Generative design process highlighting two phases and separate disciplines.

Figure 4

Casting component—design space and base-shell structure.

Figure 5

Example of driving loading conditions and respective measurements of stiffness.

Figure 6

NVH load case: excitation of front and rear suspension towers.

Figure 7

Topology optimization results for static, driving dynamics loads and NVH.

Figure 8

Driving crash load cases: pole-side-impact (left) and rear-impact (right).

Figure 9

Different pole positions for pole-side-impact.

Figure 10

Draw direction constraint to account for castability of ribs.

Figure 11

Resulting topology depending on employed manufacturing constraints.

Figure 12

Individual result of combined free-size and topology optimization for rear-impact load case.

Figure 13

Different rib designs resulting of load balancing process using linearized models.

Figure 14

Best candidate design and corresponding shell realization.

Figure 15

RSM-based optimization process.

Figure 16

RSM-based optimization process including expert emulation.

Figure 17

Shell sizing parameterization using about 20 design variables (blue).

Figure 18

Pareto-optimal designs for the objectives maximum reaction force over safe space. Exemplary illustration of the two clusters integer (green) and collapse (orange).

Figure 19

Clustering result of local horn-section RSM-optimization. Top-right insert sketches the rib thickness design-space in the local horn section. The dendrogram on top shows three clusters based on material failure contour fields of last time steps of 284 samples, respectively. One exemplary contour plot for each cluster is depicted below, visualizing the significantly different failure modes identified by unsupervised clustering: Cluster 1 = local failure only; Cluster 2 = structural shear collapse mode 1; Cluster 3 = structural shear collapse mode 2.

Figure 20

Ingate system simplification for parametrization.

Figure 21

Filling times varying the number of ingates and their position.

Figure 22

Ingate optimization initial and final state. Three material front timesteps t1, t2, t3 are represented by color. On the left the beginning of the filling (t3) is rather circular around the ingate section and the final material front (end of the filling, t1) is only located at the horns. In contrast on the right, the material front at the beginning (t3) is rather imbalanced favoring the lower section leading to a homogeneous material front at the end of the filling (t1). This shifts a potential risk zone of material defects from the structurally critical horn section to the upper crossbeam section.

Figure 23

Comparison of rib topology and orientation in the critical pole side-crash section: on the left a classical design approach with regular pattern. On the right the proposed design optimizing filling together with crash performance.

Figure 24

Final design resulting of two-phase generative design approach.

Figure 25

Performance of design resulting of two-phase generative design approach (final design) compared to design stemming from classical design process (reference design).