HOW IS THE QM CALCULATED ?

Our intent is to develop a methodology which will read the design of the product (in the form of either a CAD file or some other representation) and then compute the QM-Index of the design.

Originally, we had planned for the QM-Index to be a single number. Subsequently, we realized such a representation would not effectively display the results of the QM analysis. We have now concluded that a matrix which gives a Qm estimate for each part in the assembly in terms of each class of defects is the best solution.

Each entry in the QM-Index will be defined on the 0 to 1 scale, with a score of 0 indicating a very low quality manufacturability, and a score of 1 the converse.

The computation scheme is designed around the error catalysts. Let each error catalyst be identified by the notation i,j,k, where k = 1,...,P are defect classes, j = 1,...,M are the specific defects in each class, and i = 1,...,N are the error catalysts effecting each specific defect. For each error catalyst let the function ø i,j,k defines its strength in a given design condition. Then S i,j,k the strength of an error catalyst, is derived as follows:

S i,j,k = ø i,j,k [V] where V is the current values of all factor variables.

ø i,j,k is described by a series of a hierarchically releted functions that is recorded in a functional graph which represents ø i,j,k.

Using ø i,j,k as the basis, the structure for computing the index is:

Maintained by Dr.Das Sanchoy

  Quick jump to main page!   			  Back to The QM Research Project page!!