Original Date: 05/08/1995
Revision Date: 01/18/2007

McDonnell Douglas Aerospace (MDA)-St. Louis uses Variation Simulation Analysis (VSA) to accurately predict and minimize variation in its products. This capability allows evaluation of alternative aircraft design and process concepts to facilitate the selection of optimum designs based on function, assembly processes, and cost constraints.

The variation simulation process includes four basic steps, beginning with inputting part geometry using an appropriate translator such as IGES. Part assembly tolerances and process capabilities are then identified, and the assembly sequence is defined as a tree structure. Finally, critical measurements associated with key product characteristics are identified. A Monte Carlo simulation is then run in which feature dimensions are randomly varied based on the tolerance, process capability, and assembly sequence data. A number of reports can be generated to provide information on the number of parts expected to be out of tolerance, and to identify the level of contribution of different part features and assembly processes to those failed parts (Figure 2-7).

VSA provides a substantial benefit in verifying design quality using software instead of the more costly fabrication process. Other benefits include improved ease of assembly; more rational assignment of tolerances based on assembly process constraints; and an ability to consider cost tradeoffs associated with lowering tolerances, improving processes, or reworking parts. For example, MDA-St. Louis staff conducted a transmission mounting analysis for a major aircraft design and determined that in-line and parallelism-of-holes tolerance, set at .001-inch, could be increased to .003-inch. Another application of VSA resulted in the assembly of the F/A-18 fuselage extensions without the use of shims.

For more information see the Point of Contact for this survey.