ANSYS 14.0 Help System: 

> Workbench > User's Guide > Systems > Design Exploration

> Workbench > Design Exploration User's Guide  > Overview
> Workbench > Design Exploration User's Guide > Using Design Exploration

• Design of Experiments (DOE) and Response Surfaces

• Goal Driven Optimization

• Robust Design/Six Sigma Analysis

• Design Sensitivities

 > Mechanical APDL > Advanced Analysis Techniques Guide 

> Chapter 2. Variational Technology > ANSYS DesignXplorer

>  Chapter 1. Probabilistic Design

Design Exploration:

1. its main purpose is to identify the relationship between the performance of the product and the design variables
2. it is a CAE design "tool"
   Goal Driven Optimization (user specifies "goal" and limitations)
3. allows evaluation of system response to variations in input
   Response Surface and Six Sigma Analysis (both evaluate "robustness") 
4. is driven by design "parameters" from CAD or DesignModeler^®
   Parameters Correlation

Definitions:

Design Variables are input parameters (independent variables, selected by the user)

Performance Indicators are output parameters (state variables, extracted from the solution results)

Design of Experiments (DOE) is a technique used to determine the location of sampling points in such a way that the space of random input parameters is explored in the most efficient way and required information is obtained with a minimum number of sampling points. The DOE component is available in the Design Explorer: Response Surface, Goal Driven Optimization, and Six Sigma Analysis systems.

Goal Driven Optimization (GDO) allows the user to define the "objectives" (i.e., design goals) and the techniques searches for the "best" possible designs within the limits of the input parameters

The Response Surface is a graphical representation that allows you to see how changes to each input parameter affects a selected output parameter.  There is a response surface for every output parameter in the model.

Six-Sigma Analysis is used to evaluate the effect of uncertainty in the input parameters on the reliability and quality of the product

A Parameters Correlation study allows you to determine which input parameters are significant, i.e.,  having the most impact on your design.

Design Point is a set of input parameters

Response Point is a set of output parameters, estimated from the response surface

"Parametric" models are required

• Parameters are the "variables" of the study
• Parameters can include dimensions, material properties, loads, etc.
  • Excludes simple CAD import as IGES, Parasolid, etc. (i.e., cannot use single-feature CAD models)
  • Limit the number of input parameters to 10 - 15
  • use "Parameter Correlation" study to determine the significance of each parameter and to reduce the number of parameter needed
  • parameters can be continuous or discrete (e.g., no. of holes or ribs)
• Output quantities are also treated as parameters (displacement, stress, force, temperature, etc.)
  • may include "derived parameters" computed from other parameters

ANSYS Classic: Variational Technology (VT), ANSYS DesignXplorer (DX)

VT creates the response surface needed for optimization studies.  It uses mathematical series to approximate the response surface from a set of calculated response points

DesignXplorer applies to structural static analysis with linear material properties and to steady-state linear heat transfer analysis

DX allows a specific list of input parameters: material property, real constant, section property, surface load, body load, temperature load, and discrete variables

ANSYS Classic:  Probabilistic Design (PDS)

A statistical approach to assess the effect of uncertain input parameters and assumptions on your analysis model.  (like Six-Sigma Analysis in WB)

Probabilistic Design assess the reliability or quality of the product by means of a statistical analysis.

Uncertain parameters are described by statistical distribution functions such as the Gaussian or normal distribution, the uniform distribution, etc.

• The scatter of the Young's modulus for many materials can often be described as a Gaussian distribution with standard deviation of ±3 - 5%.
• Likewise, the geometric properties of components can only be reproduced within certain manufacturing tolerances.
• The same variation holds true for the loads that are applied to a finite element model.

The output of ANSYS PDS study is statistics and trend information:  histograms, Cumulative Distribution Function, probabilities, design sensitivities, scatter, and correlation. 
PDS can help to evaluate safety, reliability, and quality of products.