Engineers routinely seek to improve their designs in order to increase a device’s efficiency, reduce energy losses, and increase performance at a new operating point. This is of course when the right driving forces (business or engineering) exist and design improvements are economically feasible. Oftentimes, the methods we use to improve designs rely on the depth of experience within an organization; the know-how of outside consultants; or knowledge built from scratch. In any case, virtual prototyping with simulation and optimization software, continues to be an extremely valuable set of tools for the modern engineer to simulate, measure, and optimize. This is where continuous adjoint optimization can enable better design.
What can be measured within a simulation, has the potential to be optimized.
Why is simulation so important and such a key component of the design process? Simulation tools (when used properly) can enable the reliable prediction of physical phenomena occurring within a system. Engineers can then quantitatively and qualitatively analyze results to gain a better understanding over the design features that affect performance. What can be measured within a simulation, has the potential to be optimized. With adjoint optimization, engineers can more easily and deeply connect measured objectives with changes in geometry. This deeper connection is where the power of adjoint optimization can enable the new designs that would have taken several design iterations to realize.
Topology Optimization vs. Surface Optimization
HELYX Adjoint enables topology optimization for larger deformations and shape optimization (a.k.a surface optimization) for smaller deformations. Within the topology optimization paradigm, users leverage an immersed optimal surface evolving throughout a packaging space (also called a design envelope). The evolution of the optimal surface is influenced by volume sensitivities and a typical topology optimization is shown in Figure 1.
The details of topology optimization were covered in a previous blog post and webinar titled “Improving Engineering Design with Topology Optimization“. Optimization strategies using smaller geometric deformations, the boundaries of the meshed domain evolve based on the influence of surface sensitivities.
Surface optimization utilizes the surface maps of the sensitivity as a guide for surface morphing. The surface sensitivity, shown in Figure 2, possesses a magnitude and sign to indicate the direction of surface movement and relative magnitude. Within HELYX Adjoint there are two approaches to surface morphing:
- Node-Base: Users are able to specify patches that are deformable or constrained. The deformable patches are then morphed (with smooth transitions to constrained patches), based on a user-specified maximum deformation in conjunction with a surface sensitivity (field G). The G field is implicitly smoothed to remove any noise that may cause sudden changes in surface shape e.g. G peaks at a trailing edge. Figure 3 shows a typical surface optimization using the node-based approach.
- vNurbs: Within the vNurbs, control points are defined by the user in the form of a lattice structure. The sensitivities are then mapped to the control points and the surfaces are morphed in a controlled manner. Figure 4 shows a simple case morphing a shape with vNurbs control points.
Surface morphing can be used on its own or in a hybrid approach using topology optimization as a first step, followed by a surface optimization. HELYX-Adjoint offers the tools to leverage both optimization paradigms easily.
More Advanced Continuous Adjoint Optimization
Within this discussion, only single objective surface optimization was discussed. Though extremely valuable in itself, adjont optimization within HELYX Adjoint can go further with:
- Multipoint Adjoint: Users are able to optimize a given design over several operating points. This approach was disucssed in a very detailed blog post titled “Multipoint Adjoint Optimization“
- Multi-Objective Adjoint: When a design needs to optimize several objectives, the multi-objective framework within HELXY Adjoint can easily be used. This was covered in a recent blog post and webinar titled “Enhanced Topology Optimization with Multi-Objective Continuous Adjoint“.
Overall, users are able to model more complex systems with HELYX Adjoint and gain more insight into how to improve their designs.
Learn More by Watching the Webinar
Virtual prototyping using simulation and optimization tools enables faster, leaner, and more cost effective product development cycles within engineering teams. Click here to watch the webinar and join us while we discuss how surface optimization within HELYX Adjoint enable industry leaders to achieve more. You’ll be introduced to the continuous adjoint method for fluid systems and see how to setup a surface optimization analysis with HELYX Adjoint.
