Department of Defense MIL-STD-31000A — MIL-STD-31000A Technical Data Packages Standard — DoD Manual PDF (EN)

Downloaded from https://www.everyspec.com APPENDIX C MIL-STD-31000A There are multiple reasons that feature re-mastering can’t be fully automated. First, CAD systems employ subtly different feature-creation algorithms to generate the same shape. For example, systems may represent the intersection of simple round fillets in different ways. Figure 8. Example of two ways of representing the intersection of two round fillets. In some cases features defined in one system don’t exist in the target system. There may be other ways to create a similar shape, but computer programs aren’t always capable of determining which of these many possibilities might work.

Because of the labor that re-mastering requires, it should be undertaken only when absolutely necessary. Do not re-master 3D models just in case you might need them. Re-master only if it is known that changes requiring parametric features in that particular model will be required. An example of a model that might be worth re-mastering would be a model of a tool, such as a mold base or fixture, used to make similar products whose shapes vary. It is usually more economical to translate 3D master models as featureless solids, and re-master them later, if necessary. Be aware that modern CAD software is capable of making some changes to featureless models.

For example holes can be relocated, resized, or plugged. Draft can be applied to surfaces. Ribs, bosses, and slots can be moved. These capabilities further reduce the need to fully re-master models. Re-mastering can be accomplished by one of two processes. The first is for skilled CAD operators to manually recopy the original models into the target CAD systems. This process can provide the best quality models, but only if the technician doing the re-mastering understands the design intent and follows it. However, manual re-mastering is costly and often done poorly.

Semi-automated re-mastering is done with software capable of recognizing features in the source model and replicating them in the new master. After the new features are generated, the geometry of the new model is compared with the original. If the dimensions vary beyond allowable tolerances, the differences are noted and a human operator makes a decision about whether to translate the feature or translate only featureless geometry in the vicinity of the deviation. When all of the features that can be translated are complete, the model geometry is compared for accuracy and checked for quality defects. 81