[Ni...] Posted December 9, 2020 Share Posted December 9, 2020 Good Morning, I need help figuring out whether or not to use constraints on datums measured as planes to calculate profiles. When I calculate the profile with only the base alignment, the profile is pretty close what the surface points show. But that is not what the blueprint is calling. When I calculate the profile using the datums which the blueprint calls out, C | A | B, I get way out of spec. It shows the points are inside the part even though the surface points show otherwise. When I add constraints, "normal vectors " only to be precise, to the C | A | B setup, the results comes a lot closer. Which would be the correct way to calculate this? My Datums A and B are scanned as Planes on one side of the part, Datum C is scanned as a Plane but its on the opposite Face of the part. Part is 9 inches long. Here is my base alignment if that matters or not. Thank you, Link to comment Share on other sites More sharing options...
[Ja...] Posted December 9, 2020 Share Posted December 9, 2020 If your profile has a datum reference frame (DRF), you must report the profile to that DRF and not the base alignment unless the base alignment is formed in exactly the same way as the DRF. The surface points are showing relative to your base alignment, so they may be good in that coordinate system, but they are not good to the DRF. Adding the constraint to your primary datum feature C is simply forcing it to be square to the base alignment, which is essentially the same as measuring the profile to the base alignment again (which is wrong). Open your datum feature C. What is the form error? What are A1 and A2 relative to your base alignment? Link to comment Share on other sites More sharing options...
[Ri...] Posted December 9, 2020 Share Posted December 9, 2020 What exactly is Datum C? If it is a tiny feature relative to A or B then most likely you are getting measurement error due to that. Link to comment Share on other sites More sharing options...
[Ni...] Posted December 9, 2020 Author Share Posted December 9, 2020 Please sign in to view this quote. C Datum is the same face that's highlighted in yellow but on the other side of the part. Measured as a plane.C Datum.png Link to comment Share on other sites More sharing options...
[Ni...] Posted December 9, 2020 Author Share Posted December 9, 2020 Please sign in to view this quote. My form is .00229. I am using the program in inches. My A1 is 0.13855. A2 is 0.03599. I feel like this is a big reason why its so much out of spec since it projects over 9 inches to the other side where I measured for the profile. When I turn on Normal Vectors in the evaluation window, A1 and A2 become 0. It brings my profile from .060 down to .028. Link to comment Share on other sites More sharing options...
[Ja...] Posted December 9, 2020 Share Posted December 9, 2020 The projection over 9" is probably the exact reason why you have such large error. Link to comment Share on other sites More sharing options...
[Ni...] Posted December 9, 2020 Author Share Posted December 9, 2020 Please sign in to view this quote. That's what I figured. Is there any way to address this? any workarounds that's gonna give me the correct results? Or is it impossible for Zeiss CMMs to do that? That's where I was playing with the "Normal Vectors" evaluation since it forces those planes to be straight but I don't want know if its forcing it too much where it will report good even though its bad. Link to comment Share on other sites More sharing options...
[Ri...] Posted December 9, 2020 Share Posted December 9, 2020 Please sign in to view this quote. It's not a "Zeiss CMM" issue, it's a mathematical issue that is apparent in any measurement software. It's called projection. If allowed, you could flip some of the DRF ordering around like make it A B C, or B A C, and test the results that you get. Link to comment Share on other sites More sharing options...
[Ni...] Posted December 9, 2020 Author Share Posted December 9, 2020 Please sign in to view this quote. Oh ok. I can do that. What if changing the order gets me within specification? What if only 1 of those 3 orders shows within spec and the other 2 show out of spec? Which one do I pick? How would I know which one is correct and which one is incorrect? Would I need to start measuring every single plane and cylinder involved to see if they are close to their nominals? Link to comment Share on other sites More sharing options...
[Ja...] Posted December 9, 2020 Share Posted December 9, 2020 Please sign in to view this quote. That's why Richard said "if allowed". You have to report per the DRF on the print unless you receive authorization (from your supervisor, customer, etc...) to use another DRF. The authorization that you receive should also include which DRF you can report the profile within. Link to comment Share on other sites More sharing options...
[Ni...] Posted December 9, 2020 Author Share Posted December 9, 2020 Please sign in to view this quote. Alright. So I should get the profile using the 3 different possible DRF combinations. Talk to the engineers and managers to let them know the DRF(Feature Control Frame right?) on the print is giving .060 out of spec due to the part being 9 inches long and the projection being skewed by the time it reaches the end of the part. Give them the results. Then have them decide with the customer if its possible to change the order and which order to use. Does that sound about right? Just trying to be clear on the steps needed. Link to comment Share on other sites More sharing options...
[Ja...] Posted December 9, 2020 Share Posted December 9, 2020 That sounds like what I would do. You should be able to look at the print and show that this can happen depending on how datum C is defined. It should be assigned some type of form tolerance that should be significantly less than your profile tolerance. For example, if C is assigned a .100 flatness tolerance, there is no way you would ever hold a .005 profile tolerance using C as a primary datum for a feature 9" away. Link to comment Share on other sites More sharing options...
[Ni...] Posted December 9, 2020 Author Share Posted December 9, 2020 Please sign in to view this quote. There is no flatness callout or anything for Datum C. I did just check the flatness and it is showing .0020. Datum A and B do have a flatness callout of .012 but both measure around .0005. Link to comment Share on other sites More sharing options...
[Ja...] Posted December 9, 2020 Share Posted December 9, 2020 Please sign in to view this quote. There is nothing controlling the form of your primary datum? Not even through a size/width dimension? Link to comment Share on other sites More sharing options...
[Ni...] Posted December 9, 2020 Author Share Posted December 9, 2020 Please sign in to view this quote. There are dimensions on how big some of the tabs should be but that's only on one side. The side where datum C gets called out there's nothing for this extrusion. After it gets machined it gives a perpendicularity of .006. Right now its showing .009. Link to comment Share on other sites More sharing options...
[Ja...] Posted December 9, 2020 Share Posted December 9, 2020 It sounds like a poor design. Ideally, your primary datum will have a form tolerance, your secondary datum will have an orientation tolerance to the primary, and the tertiary will have a position tolerance with respect to [primary|secondary], or something along these lines. There is usually an order of precedence and the more significant the datum feature, the tighter the geometric tolerance. I'd bring up that there is nothing controlling the form of the primary datum. You will have nothing but issues down the road with this. Link to comment Share on other sites More sharing options...
[Ni...] Posted December 9, 2020 Author Share Posted December 9, 2020 Please sign in to view this quote. Thank you for the info. It is tiring trying to figure out what is wrong with the program when it turns out the design was the issue. I will suggest changing the orders of the datums like someone else suggested. But even when I try that, it still shows out of spec. I am guessing they will need to make those Datums tighter if they want to use them.datums.png Link to comment Share on other sites More sharing options...
[Ja...] Posted December 9, 2020 Share Posted December 9, 2020 Now that I can see the print, the 9.000 width is controlled with a +/- .020 tolerance. Imagine that the other face (not C) is perfectly flat. That means that datum feature C could measure a little less than .040 flatness and the width would still be satisfied. Having a limit on flatness of .040 is absurd considering that is the primary datum to your profile tolerance. Link to comment Share on other sites More sharing options...
[Ni...] Posted December 9, 2020 Author Share Posted December 9, 2020 Please sign in to view this quote. So getting a tighter tolerance of .005 would be the best to measure the profiles? This is an extrusion, it does get machined later on and that could be the reason why it has a big tolerance. But that should mean they should just remove a profile check while its still not machined? Link to comment Share on other sites More sharing options...
[Ja...] Posted December 9, 2020 Share Posted December 9, 2020 I would assume they’d want to check the profile once the datum feature is fully established (after machining). Link to comment Share on other sites More sharing options...
[Da...] Posted December 9, 2020 Share Posted December 9, 2020 Please sign in to view this quote. So you've tried |profile|.012|A|B| and |profile|.012|B|A| ? Link to comment Share on other sites More sharing options...
[Ni...] Posted December 9, 2020 Author Share Posted December 9, 2020 Please sign in to view this quote. I tried Profile | .012 | A | B | C and Profile | .012 | B | A | C and they both get me around .015. Doing it like the print calls for which is Profile | .012 | C | A | B gets me .060 I haven't checked without Datum C. Should I be doing that too? Link to comment Share on other sites More sharing options...
[Ni...] Posted December 9, 2020 Author Share Posted December 9, 2020 Please sign in to view this quote. Their blue print has 2 pages. 1 with the drawing as an extrusion where the profile is called out. and the second page with the part machined, there is no profile on the machined side. Link to comment Share on other sites More sharing options...
[Ro...] Posted December 9, 2020 Share Posted December 9, 2020 Datum -C- is the midplane of each end, not one end. Try that. It' will reduce your projection error. Link to comment Share on other sites More sharing options...
[Ja...] Posted December 9, 2020 Share Posted December 9, 2020 Please sign in to view this quote. That doesn’t seem right. Perhaps it’s not an important dimension? I’d speak with Engineering and get their take on it. 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