True Position v Legacy - Explaining to People.

MK1966 over 1 year ago

The above callout is on my drawing and it seems to be how our designers tolerance everything.
I need help explaining to designers about GDT in terms they can understand.
My datums have been generated using 3 generic planes with the correct vectors ie 1,0,0 / 0,1,0/ & 0,0,1 from 0,0,0, origin.
Sample below,
F1 =GENERIC/PLANE,DEPENDENT,CARTESIAN,$
NOM/XYZ,<0,0,0>,$
MEAS/XYZ,<0,0,0>,$
NOM/IJK,<1,0,0>,$
MEAS/IJK,<1,0,0>
I can now calculate my true position for each individual hole.

If I use legacy dimensions without datum ref I have to use ±0.125
Is this the correct way to do this.

Matthew D. Hoedeman over 1 year ago

use legacy position, only difference is you MUST be correctly aligned to the datums and legacy 'use datums' isn't right, so no (M) or (L) for datums
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
mrollings53 over 1 year ago

I don't think their drawing has any datums defined. It sounds like they are aligning the part on the table then constructing the equivalent of alignment planes to define datum planes.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
vpt.se over 1 year ago

True position (evaluated as true position) is always a positive number and never a ± tolerance. Regardless of using datums or not.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
Neil Challinor over 1 year ago

The above callout is on my drawing and it seems to be how our designers tolerance everything.
I need help explaining to designers about GDT in terms they can understand.
My datums have been generated using 3 generic planes with the correct vectors ie 1,0,0 / 0,1,0/ & 0,0,1 from 0,0,0, origin.
Sample below,
F1 =GENERIC/PLANE,DEPENDENT,CARTESIAN,$
NOM/XYZ,<0,0,0>,$
MEAS/XYZ,<0,0,0>,$
NOM/IJK,<1,0,0>,$
MEAS/IJK,<1,0,0>
I can now calculate my true position for each individual hole.
{"data-align":"none","data-size":"full","title":"image.png","data-attachmentid":537342}
If I use legacy dimensions without datum ref I have to use ±0.125
Is this the correct way to do this.

This is just lazy design practice in my opinion but what you've currently done should give them what they want. If it were me, I'd perform a 3D bestfit of all features / pointcloud data and then construct three ALIGNMENT Planes, one for each workplane (+X, +Y & +Z). I'd then use those three alignment planes as my datums (Generic planes constructed at the alignment origin in each of the thee workplanes would be equivalent). If I wanted to really play them at their own game, I'd name my primary datum TO, my secondary datum _C and my tertiary datum AD. That way, the datum reference frame would spell out TO_CAD
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
MK1966 over 1 year ago in reply to Neil Challinor

I agree with your comments Neil, might try your suggestion and see if they notice.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
markoor99 over 1 year ago in reply to Neil Challinor

Hahahaha! That's awesome. And I agree, lazy.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
louisd over 1 year ago in reply to Neil Challinor

I LOVE THIS! addaboy! That's how to stick it to -em! If the engineers can't put the effort the define clear datums, may as well give them exactly that same level of ambiguity! [TO][_C][AD] (derived from bestfit centroid) is fantastic!
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
MK1966 over 1 year ago

Tried your suggestion Neil, and guess what, the designer did not notice....lol
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
Department S over 1 year ago

No surprise there then :-)
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
NinjaBadger over 1 year ago

Whilst I like Neil's idea (sticking it to the designers!) I'd probably take a different approach.

We don't know what the parts are, but it's possible (and although not technically valid anymore as far as I'm aware) there may be implied datums - i.e. if all/most dimensions are struck off one face or bore then we could say this was an implied datum and use that. As for datum order of precedence we could make an educated guess.

As you likely know your parts, and what they do / how they assemble you could probably make an educated guess.

So rather than a somewhat belligerent approach, you could actually do your best to confirm whether these parts will assemble and function as intended (even though that's incorrectly specified).
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel