argon said:
how do i compute for the precision/tolerance ratio if i only have a single spec limit existing. for example in doing strength tests, i only have a minimum limit for my process...
I haven't seen this addressed before, but it should be a fairly common experience. Here is my take on it.
The P/T ratio is similar to the Cp Index for capability (the same as the old GM CR Capability Ratio). Therefore, you could develop the P/T equivalent of Cpk.
Since P = 6 StdDev and T = USL - LSL, a unilateral equivalent would be:
P = 3 StdDev and T = Mean Strength - Minimum Limit.