ANSYS optiSLang Capabilities
ANSYS optiSLang has an intuitive graphical user interface that enables you to connect computer-aided design tools together in a way that captures both the simulation process automation and workflows, such as sensitivity analysis or robust design optimization.
ANSYS optiSLang supports interfacing with most software tools used in virtual product development. Some tools are fully integrated, others can be connected via text-based interfaces or custom integrations. You can extract result curves using the Extraction Tool Kit (ETK) from tools like ANSYS, Abaqus, Adams and others.
Once a simulation process and/or workflow is setup, the best-practice knowledge is captured and can be shared. The geometry can be replaced and other workflows added or tweaked.
Design and Data Exploration
Understand your design or your product data using fully interactive post-processing and visualization tools backed by powerful DOE and statistical analysis algorithms. Knowing how performance is related to design inputs and operating parameters is key to simulation-driven product development, and is a strong foundation for innovation and competitive advantage.
ANSYS optiSLang provides an automatic process flow that reduces variables and generates the best possible response surface for each response with a given number of solver calls.
ANSYS optiSLang builds metamodels based on simulation or test results. Essentially, a metamodel is a surrogate that learns to respond to inputs in the same way that the simulation or real system responds, but with rapid feedback. Running a simulation to predict a certain design configuration may take hours or days, but running the reduced-order model will give the answer in a fraction of a second. These metamodels can be used for optimization or robust design analysis, or exported for use as a ROM in a system simulation.
ANSYS optiSLang algorithms perform three important tasks. First, they determine the relevant parameter sub-space, potentially reducing the size of the problem and the number of design points required to understand it. Then they develop the metamodel of optimal prognosis (MOP). Finally, optiSlang checks the forecast quality of the metamodel (CoP) using cross validation and other methods. Prediction accuracy is critical to the value of reduced-order models.
Automatic Design Optimization
Starting from the previous simulation steps, optiSLang is already aware of the most relevant parameters and can perform a pre-optimization on the metamodel to locate global optimizations. They can also provide user guidance (a decision tree) to recommend settings for the specific problem.
Available algorithms include Evolutionary and Gradient-based algorithms, as well as Pareto Optimization, Adaptive Response Surface and Global Response Surface technology. All algorithms come with robust default settings.
Models can be calibrated to match experimental data with simulation results. The software can match Scalar values or signals and can also identify the most relevant parameters to calibrate.
Robust Design and Reliability
Evaluate the reliability of a design and perform robust design optimization (RDO) with easy-to-use wizards backed by industry-leading algorithms.
Robustness evaluation is a powerful procedure to check design quality. It starts with optimized Latin Hypercube Sampling and provides proof of reliability with best-in-class algorithms while also checking limits and probabilities of overstepping.
Similar to other optiSLang capabilities, it also identifies the most important scattering variables and provides a decision tree to help you select the most appropriate robustness algorithms for your specific situation.
Easy Access from ANSYS Workbench
From ANSYS Workbench, you can easily access optiSLang with drag-and-drop functionality — you need only to set up the variation space and objectives. Then, backed by powerful algorithms, ANSYS optiSLang automatically identifies the metamodel of optimal prognosis. User-friendly project-setup wizards determine the appropriate methods for optimization.
The technology includes options for parallel computing with ANSYS Remote Solve Manager and supports the use of ANSYS HPC Parametric Pack licensing for affordable, simultaneous solution of many design points.
The ANSYS optiSLang technology inside Workbench is also capable of handling failed designs: for example, non-resolvable parameter combinations or errors that occur in the CAE process chain. With the Continue Crashed Session option, further processing of aborted analyses is secured using all previously computed data. All successful designs are stored in ANSYS optiSLang's database and can be used independently from the Workbench design table. Furthermore, you can add designs or recalculate at any time. The ANSYS optiSLang algorithms are able to handle up to a 50 percent design point failure rate.