I have some confusion I was hoping someone had come across before or made sense of since I'm having some trouble reconciling.
The end goal of what I'm trying to measure is essentially a width of a box with surfaces that have waviness to them. I'm looking to capture the maximum distance across the width and report that. Because the surfaces are rather irregular, there is the possibility of a peak one side compounding with the peak of another, or for it to cancel it depending on location (if non-symmetrical).
I initially established planes by measuring with various points on each side and measured that distance (plane to plane along the width axis). By enabling the max/min reporting feature, I managed to get some more data. However, after browsing the core documentation, I realized that the distance measurement relied heavily on curve fitting for each plane and was unclear whether the result of this was affecting the max/min values.
I then tried reporting size on a constructed width feature on the two measured planes (using the opposed points option since it's a non circular feature) and I thought that the UAME most accurately represented what I was looking for, which is basically the envelope that contains all the measured points. These values differed from the max/min reported with distance. However they varied rather significantly from physical measurements I took at the same locations so I was suspicious.
I ran tests on reference material with a dial indicator and followed up with distance and size measurements on the same surfaces to see if there was alignment. This included both steel and aluminum to see what the software did as the surfaces became less controlled. On the aluminum, both max values lay within the reference measurement range. However the min values for both were about .003" under any dial reading for that surface. For my steel reference, both mins lay within the reference measurement range (lower bound) but both maximums exceeded the reference max by about .003". The difference in the maximums between both methods was about 0.001" on the aluminum and less than that on the steel.
In my application measurement, I'm measuring across a gap of about 10-20 inches. The differences between reporting/measuring methods on the widths has been in the range of about 0.125". This makes it difficult to take either method at face value and also makes me wonder if either approach is actually reporting what I think it is. I considered that both planes have both their own flatness in addition to the parallelism (or lack of) between them working against them. However, I set up the width feature with opposing points to have surface vectors along the axis that I consider my theoretical width. For the distance measurement, I used a 2D distance along this same axis. Would this not have the same effect between both methods of projecting onto this axis?
I feel like I'm missing something obvious but I've been starting at this too long to see it. Any insight would be greatly appreciated. As a sanity check, I exported the measurement data of both planes and calculated the distance across this axis (what the surface vectors in the width feature would have been set up as) and I came within 0.003" of the reported UAME. If this is the same data that UAME is pulling the maximums from, shouldn't this be exactly zero?
Apologies for my wall of text and thanks to anyone who reads it.
TL;DR: What is the fundamental difference between the max/min reported in distance versus UAME/local on a width reported in size if distance is set up as a 2D measurement along a vector and the width has this same vector as its surface vectors?
Threads I referenced:
https://www.pcdmisforum.com/forum/pc...isolation=true
