|
|
|
A Variety of Sensitivity AnalysisThe analysis of uncertainty involves measuring the degree to which each random variable and its distribution parameters contribute to the uncertainty in the output quantities (limit-state function, failure probability and reliability index). UNIPASS™ provides a variety of important sensitivities results. These include the sensitivities of the reliability index and failure probability with respect to random variables and their distribution parameters. The distribution parameters include the means, standard deviations, and correlation coefficients of probability distributions.
In general, UNIPASS™ provides the following sensitivity information:
For CDF/PDF analysis, sensitivities values are available at different limit state function values as shown in Figure 1.
Figure 1. Example of a Sensitivity Plot in the CDF/PDF Analysis
This 3-D plot can be projected into a two dimensional view as shown in Figure 2. Click the button as Step A in Figure 1, one can plot this projected 2-D sensitivity plot.
Figure 2. The cross sectional view of a sensitivity plot with curve fitting in CDF/PDF analysis By fixing the limit state function level, one can plot a 2-D sensitivity measurement as in the conventional probability analysis. This is shown in Step B in Figure 1 and the corresponding 2-D sensitivity plot is given in Figure 3.
Figure 3. A sensitivity plot for a selected limit state function value in the CDF/PDF analysis
For a selected random variable, a 2-D sensitivity plot with coordinates of g values vs. sensitivity values can be obtained as shown in Figure 4 by following the Step C in Figure 1.
Figure 4. A sensitivity plot for a selected random variable with curve fitting
|
|
Last Updated 11/12/08
Send mail to info@predictionprobe.com with
questions or comments about this web site.
|
