Process automation and process integration

Robustness evaluation

• Pearson’s linear correlation
• Workflow definition via graphical user interface
• Reliable use with help of wizards
• robust default settings
• Connection of arbitrary complex solver chains
• Parallelization and distribution of design evaluation
• Direct integration of Matlab /Octave, Excel/OOCalc
• Python and SimulationX
• Supported connection of Ansys, Abaqus, LS-DYNA, Adams, MADYMO
• Arbitrary connection of ASCII file interfaced solvers
• Full integration of optiSLang in Ansys workbench
• Python interfaces to optiSLang library

• More then 20 probability distribution functions
• Distribution fits using measurements
• Continuous and discrete random variables
• Correlated input variables using the Nataf model
• Monte Carlo and Advanced Latin Hypercube Sampling
• Statistical assessment of output variation including:
  • histograms with automated distribution fits
  • stochastic moments
  • quantile and sigma level estimation
  • Sensitivity analysis using correlations and MOP/CoP

Sensitivity analysis

Reliability analysis

• Classical Design of Experiments
• Advanced Latin Hypercube Sampling
• Correlation coefficients (linear, quadratic, rank-order)
• Principal Component Analysis
• Polynomial based Coefficient of Determination
• Polynomial based Coefficient of Importance
• Metamodel of Optimal Prognosis (MOP) with Coefficient of Prognosis (CoP)
• MOP/CoP based sensitivity indices for important variables

• Definition of arbitrary limit states
• Monte Carlo and Latin Hypercube Sampling
• First Order Reliability Method (FORM)
• Importance Sampling Using Design Point (ISPUD)
• Directional Sampling
• Asymptotic Sampling
• Adaptive Response Surface Method

Multidisciplinary nonlinear optimization

Robust Design Optimization (RDO)

• Continuous, discrete and binary design variables
• Gradient based optimization (NLPQL)
• Global Response Surface optimization using MOP with best
• design verification
• Adaptive Response Surface Method
• Evolutionary Algorithms (EA)
• Particle Swarm Optimization (PSO)
• Stochastic Design Improvement
• Multiobjective optimization using weighted objectives
• Multiobjective Pareto optimization with EA and PSO
• Start design import from previous samples

• Sequential and fully coupled procedures
• Variance based RDO
• Reliability based RDO
• Flexible definition of robustness measures using e.g. mean values, variances, Taguchi loss functions and probability of failure
• Consideration of robustness measures in optimization
• constraints and objectives functions

Parameter identification

Post processing

• Parametrization of response signals
• Signal function library including FFT, filtering etc.
• Sensitivity analysis using MOP/CoP to check identifiability
• Flexible definition of identification goal functions
• Local and global optimization methods to search for optimal
• parameters

• Statistic post processing including anthill plots, correlation plots and sensitivity indices
• Approximation post processing including 2D and 3D plots of response surfaces and the MOP
• Optimization post processing including Pareto frontier and convergence history of design variables, responses, objectives and constraints
• Parallel coordinates plot
• Traffic light plot
• Full interaction of single plots
• Design classification using coloring, selection/deselection
• High quality outputs in BMP, PNG, SVG, EPS and PDF format