Machining Variation Analysis
Winner of a 1997 R&D 100 Award for one of the 100 most technologically significant new products of the year
The Purpose of Machining Variation Analysis
Machining Variation Analysis (MVA) is a powerful technique for analyzing the accuracy of machine tools. It allows manufacturing engineers to predict the true shape of a manufactured part including the effects of random and systematic variation. Thus, MVA improves design and manufacturing practice by giving feedback on the effects of design decisions on the quality of manufactured parts.
What Machining Variation Analysis Does
Machining Variation Analysis allows the designer or operator of machine tools to simulate the production of specific parts on specific machines. The user provides information on the nominal geometry of the part, the construction of the machine, the shape of the cutting tool, and sources of error in the machine's operation. The sources of error can include machine parametric error, fixturing error, sensor noise, etc.
With these user specified parameters, MVA determines the exact shape of the part including all the consequences of the specified errors in machine operation. Using that shape, MVA based software can compute the expected values of any tolerances held on the part as defined in geometric dimensioning and tolerancing (GD&T) including size, location, and form.
Further, MVA allows the designer to assign statistical characteristics such as mean, standard deviation, and distribution shape, to random errors. This permits calculation of yield rates, quality loss, and process capability.
How Machining Variation Analysis Works
Machining Variation Analysis models a machine tool as two chains of rigid bodies. One chain locates the workpiece, the other locates the cutting tool. MVA then animates this model, stepping through all the motions required to manufacture a specific part. MVA uses fundamentally new algorithms to precisely compute the swept envelope of the cutting tool as it moves with respect to the workpiece. Thus it can determine the exact part shape even for complicated tool paths of five axis machines. The unique swept envelope algorithm works even for complex cutting tool shapes such as those used in form grinding. To handle random errors, MVA creates a data base of "error signatures" for each source of error, then applies the Monte Carlo method to the data base. Thus, MVA can simultaneously handle systematic and random errors of any probability distribution.