Design for Manufacturability Overview
see also 

The following procedures helps you establish tolerances for each element in an assembly, so that the entire assembly will perform to its specifications with a high probability (Z = 4.5).

The procedure is accomplished in two stages, using two dialog boxes: Calculate Gap Pools and Allocate Gap Pools.

I. Calculate Gap Pools

In the first stage of the procedure, you enter information about the elements that make up the assembly, and how they come together to establish the gap. For comparison you can include your initial guesses at tolerance limits for the individual elements.

Minitab calculates statistics that indicate where the assembly stands, relative to the specification limits on the assembly gap. When you do not enter specification limits on the assembly, Minitab assumes a lower specification limit of 0.

At this stage, Minitab gives the following output:

·    Z-values. If the long-term gap Z.Bench is less than the goal (usually 4.5), you have to change either the means or the variances, or both, to achieve the goal. This is done by allocating the gap pools to the elements in the assembly.

·    Gap mean and variance pools. Tells you how much the mean and variance of the assembly gap must be adjusted, using Allocate Gap Pools, to get a Z.Bench of 4.5. If you had an upper specification on the gap, you will have both a mean and a variance pool. If there is just a lower specification on the gap, there will only be one pool to allocate.

·    Short-term and long-term statistics. Short-term statistics represent what the assembly would be like if each process were performing perfectly. Because this scenario does not represent reality, Minitab recalculates the statistics, building static shifts or dynamic drifts into the model, and gives you long-term statistics.

·    Z.Bench for the design. A measure of how good the design is.

II. Allocate Gap Pools

In the second stage of the procedure, you allocate the mean and variance pools, by designating the proportion of these pools you want made up by each of the elements. This allocation results in a new set of means and/or standard deviations.

Minitab then runs through the procedure again, recalculating new statistics based on the new means and standard deviations. The result should now be a Z.Bench of 4.5 for the design.

The final output includes recommended specification limits for each element in the assembly. If you set the specification limits at these values, you will be able to manufacture each element, as well as the entire assembly, with the PPMs shown on the graphical output.