[To...] Posted March 7, 2022 Share Posted March 7, 2022 Hello all I'm checking a Surface Profile to ABC using Unequal Distribution and everything appears ok. However, when I select No Datum Reference Frame to check the form, the Unequal Distribution option goes away and leaves only Bilateral. Any reason why Unequal Distribution shouldn't work? ver 7.0.12 Link to comment Share on other sites More sharing options...
[Ro...] Posted March 7, 2022 Share Posted March 7, 2022 Hi Tom, Because selecting no datums means it going to do a Best Fit, which would be the middle of the entire tolerance zone, which is bilateral. Link to comment Share on other sites More sharing options...
[Ja...] Posted March 7, 2022 Share Posted March 7, 2022 With no datum reference then unequal distribution would be from what nominal? Link to comment Share on other sites More sharing options...
[Ja...] Posted March 7, 2022 Share Posted March 7, 2022 I still think software should allow for unequal distribution even if it does the best fit to its nominal surface. This will be helpful in cases where the print callout has unequal distribution to its nominal CAD surface. I get prints with dimension callouts .xxx +.005/-.002. So there must be cases where design intent is similar for a profile callout. In any case Tom, I know its not the question and you might have already thought of this alternate solution: create a geometry best fit alignment where the surface controls all degrees of freedom and use that as a datum in your profile callout. Link to comment Share on other sites More sharing options...
[To...] Posted March 7, 2022 Author Share Posted March 7, 2022 Please sign in to view this quote. I would assume from the nominal surface. Link to comment Share on other sites More sharing options...
[To...] Posted March 7, 2022 Author Share Posted March 7, 2022 Please sign in to view this quote. I would think the best fit is driven by the deviations from the nominal surface, regardless of the tolerance zone. Link to comment Share on other sites More sharing options...
[Jo...] Posted March 8, 2022 Share Posted March 8, 2022 Please sign in to view this quote. Does it still have the option for a "outwards or inwards" to infinity option ? If so, you could break it into two characteristics and satisfy the unequal band. Link to comment Share on other sites More sharing options...
[To...] Posted March 8, 2022 Author Share Posted March 8, 2022 Please sign in to view this quote. No, just Bilateral - One Result or Bilateral - Two Results Link to comment Share on other sites More sharing options...
[Ri...] Posted March 8, 2022 Share Posted March 8, 2022 Just create a Best Fit Alignment inside of the Free-Form Surface, and then use that Alignment inside of the Surface Profile. To me an Unequal Distribution on a Best Fit would seem nonsensical, but that's just my personal opinion. It should split the error evenly (as best as possible) across the Profile, so wouldn't almost all parts show OOS? What is strange is that it works the way you'd like for Line Profile, just not Surface Profile. Link to comment Share on other sites More sharing options...
[To...] Posted March 9, 2022 Author Share Posted March 9, 2022 Please sign in to view this quote. How is that different from using No Datum Reference Frame? Oh, I see I have the option for unequal distribution. Please sign in to view this quote. To me it would make sense to see if the form is good, i.e. would fit freely into a mating feature with the same shape, maybe also with an unequal distribution. Link to comment Share on other sites More sharing options...
[Ri...] Posted March 15, 2022 Share Posted March 15, 2022 Why would form require unequal distribution? When Profile evaluates a feature with which ever distribution method chosen it is evaluating form as well. When it is assigned No feature reference frame then the form is evaluated without regard to location, orientation, etc and the result is of form deviation. Link to comment Share on other sites More sharing options...
[Ri...] Posted March 15, 2022 Share Posted March 15, 2022 Please sign in to view this quote. The issue is that form does not look at the nominal geometry. You can of course evaluate an unequal distribution for a form characteristic - I know people who do this, so they can visualize the high/low points points better, but nevertheless it will always show the part out of specification. Link to comment Share on other sites More sharing options...
[To...] Posted March 15, 2022 Author Share Posted March 15, 2022 Please sign in to view this quote. It might not if unequal distribution was an option. I'm curious if this is a difference between ISO and ASME. I recall Shabu posting a video link where someone explained that the nominal data was shifted in one system but not the other. Link to comment Share on other sites More sharing options...
[To...] Posted March 15, 2022 Author Share Posted March 15, 2022 Please sign in to view this quote. In my case, I have a part with .002 Profile with U.000 to A B C. This means I can't be any larger than the nominal surface but can go under by .002. Calypso handles this just fine. But when I use No Datum Reference Frame, the option for Unequal does not exist. Only the standard bilateral exist. If I increase the tolerance to .004, at least I can see where the points are on the low side and any points on the high side are technically out even though the profile would show good. The goal is to determine of the cutting path is good but just needs an adjustment in location. If the cutting path is bad, we will likely never be able to hit the tolerance to ABC Link to comment Share on other sites More sharing options...
[Ri...] Posted March 15, 2022 Share Posted March 15, 2022 Please sign in to view this quote. My point was a form characteristic can never show a measurement within specification with an unequal distribution because of the evaluation principle used to measure a form characteristic (Minimum Zone) For a Profile, a Best Fit is still doing to attempt to split the error along the Nominal Profile. It does not care if the tolerance is one-sided or not. It is about equally distributing the error. The only case where I could see an unequal distribution working on a Profile is in a close-looped Profile where now the size of the Profile is being evaluated, not just form. Link to comment Share on other sites More sharing options...
[Ri...] Posted March 15, 2022 Share Posted March 15, 2022 I think I did find out how to do what you are looking to do inside of a Curve though. You have to set the tolerance inside of the Curve, and then in the Best Fit, set the parameter to "According to tolerance range". Now it will evenly distribute the error along the nominal zone of tolerance, not the nominal of the profile. Link to comment Share on other sites More sharing options...
[Ja...] Posted March 15, 2022 Share Posted March 15, 2022 Please sign in to view this quote. With no feature reference frame, profile characteristics will also account for size of the feature in addition to its form. Plane feature is the only exception as it is a 2D feature and size does not apply to it. So, surface profile with no datum reference frame is not equivalent to form of the feature except when the feature is a plane. Please sign in to view this quote. a Best Fit is minimizing the error rather than equally distributing it along the nominal. So for profile, unequal distribution should be an option even when no datum reference frame is selected. Please sign in to view this quote. You can use bilateral with two results and add a resultant element which checks that maximum result from two profile results is zero meaning no points lie above the nominal surface. Link to comment Share on other sites More sharing options...
[Ri...] Posted March 15, 2022 Share Posted March 15, 2022 Poor choice of words, but minimizing is what I meant (depending on the shape of the results it will look evenly distributed). It is doing its best to minimize the error along the nominal profile, so that means you are naturally going to have points that are on the negative side and the positive side. It would nearly be impossible to have a case where you would have 0 deviation on one side of the nominal profile if using a best fit. This is the case for nearly all form measurements - take a look at a Flatness or Roundness measurement and see what I mean. Like I said, it is possible with Curve, but all the software is doing is shifting the nominal profile to the nominal of the tolerance band. If a Profile is open, and not closed, it is not a feature of size, so it can float however it needs to "minimize" the error. Link to comment Share on other sites More sharing options...
[Ri...] Posted March 15, 2022 Share Posted March 15, 2022 I just tested it inside of GOM Inspect to see how it would handle it. It shifts the nominal to match whichever direction the tolerance zone is. I can flip the material direction either way, and it will either show I have positive material or negative material. Tom, I understand what you are trying to accomplish by separating the form from the location, but I feel like your form just needs a normal bilateral tolerance zone. Link to comment Share on other sites More sharing options...
[Ja...] Posted March 15, 2022 Share Posted March 15, 2022 Please sign in to view this quote. Oh I did not even consider a freeform surface which can also be an open profile depending on its shape. You are right about that that best fit will minimize the error by trying to fit points on both sides of nominal if the profile is an open one. Link to comment Share on other sites More sharing options...
[To...] Posted March 16, 2022 Author Share Posted March 16, 2022 Please sign in to view this quote. Yes, it is a closed loop Profile. I do understand that "No DRF" is not evaluating using Minimum Feature. My use of the term "form" was incorrect. I did not consider "size" but that is what I should have said. Are you now saying there should be an Unequal Distribution option??? Link to comment Share on other sites More sharing options...
[Ri...] Posted March 16, 2022 Share Posted March 16, 2022 Please sign in to view this quote. It seems that Calypso doesn't allow for a Surface Profile to use an Unequal Distribution if there is no DRF, but yes I would say that it absolutely should allow for it if it is a closed loop Profile. If you are using Free-Form, just create an alignment from the Best-Fit inside of the Free-Form, and use that as your DRF. It will give you the same results, but now since there is an alignment Calypso will allow you to use Unequal Distribution. Link to comment Share on other sites More sharing options...
[To...] Posted March 16, 2022 Author Share Posted March 16, 2022 Please sign in to view this quote. Portion of original post below.... Please sign in to view this quote. Link to comment Share on other sites More sharing options...
[Ri...] Posted March 17, 2022 Share Posted March 17, 2022 Tom, Did you see my note about use Free-Form Best Fit Alignment as the Alignment in your Profile? Link to comment Share on other sites More sharing options...
[To...] Posted March 17, 2022 Author Share Posted March 17, 2022 Please sign in to view this quote. Yes, I did. Thank you. #workaroundsstink 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