In-depth design analysis and simulation are typically used as a means to assess the performance of a part or a system for one given set of CAD dimensions, loads and material properties.

However, simulation can also be used as a way to answer "what if" questions: What happens if my load changes by 10%? Which parameters are really influencing the behavior of my system? These types of questions can be answered by an appropriate parametric study of the model.

A parametric analysis can be either deterministic or probabilistic. In the former case, all parameters are supposed to vary continuously within a given range (defining the design space) and the expected result is the continuous response of the various performances. The deterministic analysis is a first step to take to understand the product and find a feasible design within the design space. Once a feasible design has been found, the next question is: how robust is it?

The probabilistic parametric analysis will help answer that question. Loads, dimensions and material properties are not deterministic parameters; a dimension has a certain tolerance and material properties depend on the manufacturing process. How does the scattering of these parameters influence the performance of the product? How much is the product likely to fail?

Design Exploration for All Physics

ANSYS offers an unparalleled breadth of solutions across a broad range of physics that can accurately address the fluid, structural, electromagnetic and thermal modeling of any product. Through the combined use of ANSYS DesignXplorer technology and the comprehensive multiphysics solutions from ANSYS, parametric analyses are available for virtually every simulation. ANSYS DesignXplorer software supports all physics available from the ANSYS Workbench schematics: structural (both implicit and explicit), fluid flow and multiphysics. Combined analyses in which multiple physics are analyzed independently or in a coupled manner are also supported.

Parametric analysis of coupled physics simulation

Design of Experiments and Response Surfaces

Achieving a good design point often means making trade-offs between various objectives, and the exploration of a given design cannot be performed exclusively by using direct optimization algorithms that lead to a single design point. It is important to gather enough information about the current design to be able to answer what-if questions and to quantify the influence of design variables on the performance of the product in an exhaustive manner. In doing so, the right decisions can be made based on accurate information, even in the event of an unexpected change in the design constraints.

ANSYS DesignXplorer software provides a description of the relationship between the design variables and the performance of the product by using Design of Experiments (DOE) combined with Response Surfaces. DOE and Response Surfaces provide all the information required to take advantage of Simulation Driven Product Development. When performance variations due to design variables are known, it is easy to understand and identify all changes required to meet the product requirements. Once the Response Surfaces are created, the information can be shared in terms that are easy to understand: curves, surfaces, sensitivities, etc. They can be used at any time during the development of the product without requiring additional simulations to test a new configuration.

Response surface

Sensitivity plot

Optimization

Good design starts with identifying the relationship between performance and design variables. In most cases, there will be a number of "optimal" designs to choose from, depending on the parameters and objectives of the product. ANSYS DesignXplorer software offers several ways to investigate response surfaces, including interactive tools and optimization algorithms such as multi-objective search. This technology analyzes input constraints, such manufacturing cost, and provides several design candidates from which you can choose. With optimization, there are usually trade-off solutions.

Trade-off plots: visualization of product performance and design candidates

Parallel plots: visualization and interactive manipulation of design variables and product performance on a single graph

Multi-objective optimization providing several design candidates