[Ri...] Posted March 5, 2020 Share Posted March 5, 2020 I'm looking for information on how or why order of precedence of features changes the result. I have one in particular, the features do not create a Bolt Circle. All of them are a Datum. All of them are cylinders. Cylinder_A Cylinder_B1 Cylinder_C Cylinder_D1 Cylinder_E1 Cylinder_E2 When the features are selected in the order above, only one falls out of tolerance (Cylinder_E2) If I change the order to something like: Cylinder_A Cylinder_C Cylinder_B1 Cylinder_D1 Cylinder_E1 Cylinder_E2 Then D1 and E2 are out of tolerance. My train of thought is I select a order so its easy to determine which true Position result belongs to which feature. If there is no order stated on a B/P, who determines the order? If I use True Position on each individually only E2 is out of tolerance. There doesn't seem to be a way to identify a feature to the "carrot" value (^1, ^2, 63, etc) other than a note in the comments. Calypso help is not much help. Any information or help is appreciated.Untitled.jpg Link to comment Share on other sites More sharing options...
[Da...] Posted March 6, 2020 Share Posted March 6, 2020 The order should be irrelevant. That is interesting what version and what service pack of calypso? Are you sure the setting for translate and rotate are the same? I will do some testing tomorrow and get back to you I am running 5.4, 6.4, 6.6 and 6.8 What is the deviation of the one time that E is out vs when E and D are out? I wondering if the software is finding both results because the formula used to calculate best fit. If Zeiss is using what I think they are using then it is testing and recalculating thousands time refining the result until the variation is negligible. Then there could be two resultant targets that are equal in amplitude. By changing the order it is finding one vs the other. In other words you can make the part fit with all holes lining up but one and it is pretty fare off or you can get all but 2 holes and they are just slightly off in opposite directions. The deviation of the two relative to there location is equal to deviation of the one. Link to comment Share on other sites More sharing options...
[An...] Posted March 9, 2020 Share Posted March 9, 2020 Please sign in to view this quote. I don't think so. See attached.Contribution_09_03_2020_1.pdf Link to comment Share on other sites More sharing options...
[Da...] Posted March 9, 2020 Share Posted March 9, 2020 Please sign in to view this quote. Not Sure what you are trying to say there? The one result is at the Sum of squared deviations is a minimum? It should be what about the other results data? Send the data for the other result that both E and D are out. Thanks Dan Link to comment Share on other sites More sharing options...
[An...] Posted March 9, 2020 Share Posted March 9, 2020 What you can see in my contribution is the simulation of a squared-deviation-calculation when regarding a six-bore-pattern very similar to the pattern in question. (estimated values for all six bores) The calculation has nothing to do with CALYPSO,it's made by an Excel-VBA-programm. I just wanted to show that there isn't any ambiguity in result if using Gauss-Fit. Loop #66460 shows the lowest point in the whole process.There is no other place in the chart where you can find exactly the same value. Link to comment Share on other sites More sharing options...
[Ri...] Posted March 9, 2020 Author Share Posted March 9, 2020 Please sign in to view this quote. A> Varies, from 6.4.24 to 6.4.28 B> I will have to parse the data from measure points, memory is failing me on this. Link to comment Share on other sites More sharing options...
Recommended Posts
Please sign in to comment
You will be able to leave a comment after signing in