|
<< Click to Display Table of Contents >> Navigation: Analysis Output > Advanced Analyzer and Optimizer > Simulation-based Sensitivity > SBS Limitations |
SBS can only predict Mean and 'n'-STD. Predicting the Curve Type and Est. Range would require much longer simulation time and many more calculations in SBS. These are part of our long-term vision for SBS but since computers are not really powerful enough for SBS now, these improvements are postponed for a later time.
SBS calculation can be slow for large models or when using a higher number of Interactions. To speed up the calculation time Shared Memory (on a single machine with Multi-Core Processing) or Distributed Computing (on the Cloud) can be used.
|
SBS vs Geo Factor
In Geo Factor, we want to emphasize the effect of the design so users can easier spot cantilever effects. The effects of distributions are only shown with hidden multipliers in the 6-Sigma column.
oIn SBS, every setting in the contributor is included in the coefficient. So if a Constant or Uniform distribution is used, if a scale is not equal to 1, etc., then the coefficient will be affected.
oGeoFactor treats points/features in out-of-grouped tolerances as separate contributors. SBS the tolerance is one contributor.
oWhen we say the SBS coefficients are similar to the G Factors this is what we mean - that under the right prerequisites the number is the slope of the line relating the input and output (linearity). SBS is better because it meets the prerequisites more often. It doesn't, however, break down the coefficient into components such as G Factor, bonus tolerance, etc., the way Geo Factor does.