hexagon logo
Register Sign in
  • Replies 11 replies
  • Subscribers 221 subscribers
  • Views 204 views
  • Users 0 members are here
Options
Related

Slot Width/Length Bonus?

blind829
I've attached a print and code of my program on v3.5MR2.

First question is why is PC-Dmis adding both bonus tolerances, width & length, together to calculate the true position bonus? I'm looking at a old CMM report of v3.7MR3 and it is only using bonus tolerance of the width to calculate the true position bonus...Is this a known problem in v3.5MR2?

Second, I don't fully understand boundary and if it's possible to dimension correctly in v3.5MR2...

Thanks for your help.
Attached Files
  • I've attached a print and code of my program on v3.5MR2.

    First question is why is PC-Dmis adding both bonus tolerances, width & length, together to calculate the true position bonus? I'm looking at a old CMM report of v3.7MR3 and it is only using bonus tolerance of the width to calculate the true position bonus...Is this a known problem in v3.5MR2?

    Second, I don't fully understand boundary and if it's possible to dimension correctly in v3.5MR2...

    Thanks for your help.
    Well, TP and it's associated bonus has been a bug-a-boo in Pcdmis for EVER. Way back in V3.5, if I recall, I had to report the TP with ONLY the diameter and report the length as it's own dimension. Is it possible to make the length RFS and the width MMC in the same dimension?

    Also, technically (and I am sure there will be a lot of input on this) a slot is not a singular feature of size, there are TWO sizes involved, length of slot can only give bonus in the direction of the length of slot, width can only give bonus in the direction of width and so on. If the slot is square to alignment (length is along X axis for example), then dimension it as 2 TP's, one for the width and axis, one for the length and axis. Don't know how the width bonus can be applied to the length direction and vise-verse.

    Boundary, that really requires a scan to be done of the profile and then dimensioned as a profile with the correct tolerance (form & location) applied to it.
  • Also, technically (and I am sure there will be a lot of input on this) a slot is not a singular feature of size, there are TWO sizes involved, length of slot can only give bonus in the direction of the length of slot, width can only give bonus in the direction of width and so on.

    What if one end of the "round" slot is high side of tol. and the other end is at the low end; would PC DMIS use the average diameter fro bonus?
  • Also, technically (and I am sure there will be a lot of input on this) a slot is not a singular feature of size, there are TWO sizes involved, length of slot can only give bonus in the direction of the length of slot, width can only give bonus in the direction of width and so on.

    What if one end of the "round" slot is high side of tol. and the other end is at the low end; would PC DMIS use the average diameter fro bonus?
    Good question. Next?

    I don't run into that much, I deal with pierced sheet metal, "perfect" punches punching "perfect" holes. Only time I could see any difference from end to end is on a developed hole (pierced before forming) and those generally have a "pass-a-pin" type callout on the print.

    Exactly how Pcdmis calculates the width of a round slot (and I hate the way Pcdmis measures them) has not been revealed by the coders, at least, NOT that I have seen on the site.
  • This is from Y14.5M 1994 Pocket Guide;
    "The position boundary specification locates the feature to the datum reference frame.
    No portion of the surface may be permited to lie within the boundary of MMC contour
    minus the position tolerance when position with respect to the DRF established by datum
    features A,B at MMC, and C at MMC."
  • Well, TP and it's associated bonus has been a bug-a-boo in Pcdmis for EVER. Way back in V3.5, if I recall, I had to report the TP with ONLY the diameter and report the length as it's own dimension. Is it possible to make the length RFS and the width MMC in the same dimension?

    Also, technically (and I am sure there will be a lot of input on this) a slot is not a singular feature of size, there are TWO sizes involved, length of slot can only give bonus in the direction of the length of slot, width can only give bonus in the direction of width and so on. If the slot is square to alignment (length is along X axis for example), then dimension it as 2 TP's, one for the width and axis, one for the length and axis. Don't know how the width bonus can be applied to the length direction and vise-verse.

    Boundary, that really requires a scan to be done of the profile and then dimensioned as a profile with the correct tolerance (form & location) applied to it.

    I can't make the length RFS and the width MMC all in one feature. I ended up doing what you suggested and simply dimensioned the length out separately. It must be something they fixed after V3.5.

    Matt, can you got into further detail on how you would go about measuring this as a boundary, what auto features you would use and such? Thanks
  • I can't make the length RFS and the width MMC all in one feature. I ended up doing what you suggested and simply dimensioned the length out separately. It must be something they fixed after V3.5.

    Matt, can you got into further detail on how you would go about measuring this as a boundary, what auto features you would use and such? Thanks
    I would do a CLOSED SCAN with the scan generated from the cad data. Pretty much, that's what is required. None of the scan can violate the boundary boundaries, whichever why they chose to dimension it, either "bigger" or "smaller" or bi-lateral.
  • I would do a CLOSED SCAN with the scan generated from the cad data. Pretty much, that's what is required. None of the scan can violate the boundary boundaries, whichever why they chose to dimension it, either "bigger" or "smaller" or bi-lateral.

    Thanks Matt, so looking at the original print I attached I would do one dimension of the length and width size tolerance. Then perform a linear closed scan to the CAD of the round slot...edge point scan or just a vector scan?

    Then dimension that closed scan directly using the "Profile" dimension tool, form and location, with a tolerance of ±0.1. Does that sound like it'll satisfy the print?
  • Thanks Matt, so looking at the original print I attached I would do one dimension of the length and width size tolerance. Then perform a linear closed scan to the CAD of the round slot...edge point scan or just a vector scan?

    Then dimension that closed scan directly using the "Profile" dimension tool, form and location, with a tolerance of ±0.1. Does that sound like it'll satisfy the print?
    Based on that print, I wouldn't even report TP, only the scan results, but I am sure that would confuses your customer.

    Since it has a different tolerance for the length than it does for the width, it's gonna be a beech no matter how you try to report the scan. This is one of those cases where you are pretty much screwed no matter what you do. Personally, I would ignore the BOUNDARY callout since I seriously doubt your customer "gets it" and it was put on there by an engineer who didn't really know what it meant either. V3.7 does not have the "BOUNDARY" option for a dimension, don't know if they people at Wilcox have added it in the newer versions.

    IF you just have to report BOUNDARY, generate the scan from the cad data and give profile tolerance a -0.9 tolerance (from TP and size) and give it a +0.9 tolerance (from TP and SIZE). That's the best (I think) that you will be able to do given that you do not have a BOUNDARY dimension option. Not exactly correct since you have different size tolerances, but the best you can do with what you have.
  • Based on that print, I wouldn't even report TP, only the scan results, but I am sure that would confuses your customer.

    Since it has a different tolerance for the length than it does for the width, it's gonna be a beech no matter how you try to report the scan. This is one of those cases where you are pretty much screwed no matter what you do. Personally, I would ignore the BOUNDARY callout since I seriously doubt your customer "gets it" and it was put on there by an engineer who didn't really know what it meant either. V3.7 does not have the "BOUNDARY" option for a dimension, don't know if they people at Wilcox have added it in the newer versions.

    IF you just have to report BOUNDARY, generate the scan from the cad data and give profile tolerance a -0.9 tolerance (from TP and size) and give it a +0.9 tolerance (from TP and SIZE). That's the best (I think) that you will be able to do given that you do not have a BOUNDARY dimension option. Not exactly correct since you have different size tolerances, but the best you can do with what you have.

    You're good Matt, last couple questions.

    What "Hit Type" would the linear closed scan be? Edge, Vector, Surface, ect..

    Correct me if I'm wrong but you calculated your ±0.9 tolerance by splitting the 1.0mm positional tolerance in half, makes sense, and adding the tightest hole size tolerance, just to be on the safe side...Is that right?
  • You're good Matt, last couple questions.

    What "Hit Type" would the linear closed scan be? Edge, Vector, Surface, ect..

    Correct me if I'm wrong but you calculated your ±0.9 tolerance by splitting the 1.0mm positional tolerance in half, makes sense, and adding the tightest hole size tolerance, just to be on the safe side...Is that right?
    Yep, that's how I did it, 1/2 the TP and guessing that cad is mean size of slot, half the total size tol to each side.

    As for hit type..... Edge points with a surface sample will probably be the easiest way to go about it.

    IF YOU HAVE both sides of stock AND the edge between them, you could do a local alignment on it (level on surface around it and origin, slot length for rotate, center point for origins), then do vector points on the edge data. While it would take a lot longer to set up doing the local alignment, the scans will go a LOT QUICKER! Of course, if this is a flat piece, do a local alignment (level at least) and do edge points WITHOUT the sample hit. There's about 3 ways to skin this cat.
© 2024 Hexagon AB and/or its subsidiaries. Privacy Policy | Cloud Services Agreement