This node uses an adaptive response surface method (ARSM) to compute the probability of failure.
The idea is to find a suitable approximation of the limit state function and perform the reliability analysis on this metamodel, rather than by direct solver calls. The method first samples a set of parameters (design of experiments) by Latin Hypercube Sampling. The samples are centered around the mean value of the respective random variables and may be assigned a larger scatter than that of the original variables. The scatter may be enlarged by the algorithm if no unsafe designs are found. For each design, the responses are evaluated by the external solver.
Internally, the responses are used as supports to compute the approximate limit state function. The probability of failure is computed by Directional Sampling. ARSM then refines the approximation within the following steps, using points on the failure surface found within the Directional Sampling procedure. These additional supports are added to the previous set.
ARSM is a general purpose method which is fast and accurate as long as the responses used to calculate the limit state function can be approximated, for example have no severe non-linearities. The number of designs to be evaluated by the actual solver depends mainly on the number of random variables. Limit states with convex and non-convex shape, even with non-linearities or islands, are possible to be treated to some extent.
Further information about methods of reliability analysis used in optiSLang can be found here.
Initialization Options
To access the options shown in the following table, double-click the ARSM-DS system on the Scenery pane and switch to the ARSM tab.
| Option | Description | ||
|---|---|---|---|
| Metamodel | |||
| Number of supports in 1st step | Sets the number of supports in the initial steps. | ||
| Scaling factor in 1st step | Sets the initial scaling factor of the standard deviations (according to standard normal space). Use a value larger than one to ensure that at least a few designs will be located in the failure domain. The recommendation is similar to the expected reliability index. Nevertheless, ARSM may still treat the problem if no design of the initial step is in the failure domain. In the subsequent adaptation steps, the scaling factor is reduced to one. | ||
| Number of supports per step | Sets the number of supports in all consecutive adaptation steps. | ||
| Number of steps | Sets the number of adaptation steps (including the initial step). | ||
| Directional Sampling | |||
| Number of directions | Sets the total number of simulated directions per step. | ||
Additional Options
To access the options shown in the following table, in any tab, click .
| Option | Description |
|---|---|
| Limit maximum in parallel |
Controls the resource usage of nodes in the system. When the check box is cleared (default), a value is chosen to ensure the best possible utilization of the child nodes. When the check box is selected, set the value manually to specify how many designs are sent to child nodes, limiting the maximum degree of parallelism for all children. Ansys recommends keeping the check box clear. |
| Auto-save behavior |
Select one of the following options:
The project, including the database, is auto-saved (depending on defined interval) after calculating this node/system (either when the calculation succeeds or fails). By default, all parametric and algorithm systems have selected, all other nodes have selected. |