Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
Tangent Line as datum
- Thread starter Bux1986
- Start date
Burunduk
Mechanical
- May 2, 2019
- 2,574
Regarding using a tangent plane (not line) to establish a datum, if you really insist you could define a datum target plane similarly to something that I think is shown in the digital product definition standard by ASME. I'm not really directing you there because in your case the idea doesn't look right. The part looks symmetrical, why one side only as datum feature C? And will the user be able to distinguish this side from the other if the part is flipped over? And more importantly, how does this mate to other components?
- Thread starter
- #3
Hi Burunduk,
Thank you very much for your feedback.
The part looks symmetrical, why one side only as datum feature C? --> I was hesitant to use the symmetry line as the 0.00 line since it is not a hard line that the user can see or easily measure since the corresponding symmetrical surfaces are curves, hence one side was chosen as shown in the picture.
This is what I want to understand " Can I use the line/plane of symmetry (as in the picture) as a datum? If not then what should I use? My initial thought is to use a tangent to the curved surfaces as shown in the picture, But I am not confident.
The functionality of the part: It is a metal plate that sits on a connector. It should work when flipped upside down. The hole locations are most critical.
Thank you very much for your feedback.
The part looks symmetrical, why one side only as datum feature C? --> I was hesitant to use the symmetry line as the 0.00 line since it is not a hard line that the user can see or easily measure since the corresponding symmetrical surfaces are curves, hence one side was chosen as shown in the picture.
This is what I want to understand " Can I use the line/plane of symmetry (as in the picture) as a datum? If not then what should I use? My initial thought is to use a tangent to the curved surfaces as shown in the picture, But I am not confident.
The functionality of the part: It is a metal plate that sits on a connector. It should work when flipped upside down. The hole locations are most critical.
Using the width of the connector as the datum feature would eliminate the worries about flipping it.
The problem with one side only is that the part can pass in one orientation and fail in the other, meaning the inspector might have a 50/50% chance of failing the part initially and then accepting it, but that chance will depend on the actual distribution of the variation among the parts.
The better method is to call out all the features to the mating face as a simultaneous requirement and not even use datums B and C, if the features used for B and C aren't being used to force the related features into position.
Alternatively, use the holes as datum feature B and a profile tolerance that references A and B.
The problem with one side only is that the part can pass in one orientation and fail in the other, meaning the inspector might have a 50/50% chance of failing the part initially and then accepting it, but that chance will depend on the actual distribution of the variation among the parts.
The better method is to call out all the features to the mating face as a simultaneous requirement and not even use datums B and C, if the features used for B and C aren't being used to force the related features into position.
Alternatively, use the holes as datum feature B and a profile tolerance that references A and B.
Burunduk
Mechanical
- May 2, 2019
- 2,574
If the part is designed to be used symmetrically, the datum feature selection and the tolerancing of part features should reflect that. Selecting a single side to be used as a reference does not achieve that.
If there is no way to tell datum feature A from the face opposed to it, you could designate the thickness of the plate as datum feature A (center plane datum) and use the two holes as a pattern as datum feature B, at MMB if you are working per ASME Y14.5 and can tolerate the datum shift. This is especially preferable if the holes are what locates the part.
I would be cautious to use only one face as a datum reference for everything and rely on a simultaneous requirement. Unfortunately, I'm not sure that all inspection software packages are able to apply a simultaneous requirement to different controls in the condition when not all the relevant degrees of freedom are constrained; When all 6 DOF are constrained identically for different controls, the simultaneous requirement kind of happens by itself. But when the datum features and simulators only constrain the part partially, something needs to correlate the measurements when the measurements are ASME compliant. I'm not sure that something is taken care of by the programmers. I wish I'm wrong.
If there is no way to tell datum feature A from the face opposed to it, you could designate the thickness of the plate as datum feature A (center plane datum) and use the two holes as a pattern as datum feature B, at MMB if you are working per ASME Y14.5 and can tolerate the datum shift. This is especially preferable if the holes are what locates the part.
I would be cautious to use only one face as a datum reference for everything and rely on a simultaneous requirement. Unfortunately, I'm not sure that all inspection software packages are able to apply a simultaneous requirement to different controls in the condition when not all the relevant degrees of freedom are constrained; When all 6 DOF are constrained identically for different controls, the simultaneous requirement kind of happens by itself. But when the datum features and simulators only constrain the part partially, something needs to correlate the measurements when the measurements are ASME compliant. I'm not sure that something is taken care of by the programmers. I wish I'm wrong.
Good news, an optical comparator can do this, no software required. Datum shift is the same problem to solve as the simultaneous requirement is - the software has to reposition the selected points to find a location and rotation to meet all the requirements at the same time.
pmarc
Mechanical
- Sep 2, 2008
- 3,247
Bux1986,
Show two tangent lines in the top view and label each of them with a datum target symbol C1 and C2 (don't use the datum feature C symbol at all). You may use movable datum target symbols for that but they are not required - the fact that C is referenced at RMB (no (M) or (L) modifier specified after C) in the position feature control frame means that the two datum target planes need to be adjustable as far as the distance between them is considered.
I assume you work to ASME Y14.5.
Show two tangent lines in the top view and label each of them with a datum target symbol C1 and C2 (don't use the datum feature C symbol at all). You may use movable datum target symbols for that but they are not required - the fact that C is referenced at RMB (no (M) or (L) modifier specified after C) in the position feature control frame means that the two datum target planes need to be adjustable as far as the distance between them is considered.
I assume you work to ASME Y14.5.
Burunduk
Mechanical
- May 2, 2019
- 2,574
A simple optical comparator and an overlay is a classic measurement solution for a part like this, but not everyone uses it. Many prefer devices that are considered more accurate and advanced, and they indeed are in certain aspects, but then the user is at the mercy of the developers of the software that analyses a scan. That's why I think caution should be taken.3DDave said:
Since ignoring datum shift doesn't impose a risk of accepting bad parts, many CMM software packages still do. Regardless, "meet all the requirements at the same time" indicates this is still a problem related to supporting simultaneous requirements - correlation of different controls that was either considered or not.3DDave said:
pmarc said:
pmarc,
Why are you emphasizing to NOT use datum feature symbol? I am just trying to understand the full reasoning beside of the facts some figures from Y14.5 (7-58/2018) clearly shows your idea (of not using datum feature symbol)?
Again, why not?
The datum feature symbol is applied to datum features. Datum target lines aren't a feature and the datum feature symbol doesn't apply.
The width of the part could be a datum feature - that would be two parallel planes - applied as it would to a rounded slot or other irregular feature of size.
The datum line is evaluated at some vertical height between the two faces and that height needs to be specified in this case.
The width of the part could be a datum feature - that would be two parallel planes - applied as it would to a rounded slot or other irregular feature of size.
The datum line is evaluated at some vertical height between the two faces and that height needs to be specified in this case.
3DDave said:
Should I understand that if tangent lines are used as datum features we shall NOT use datum feature symbols, BUT, if we are going to use datum target POINTS (not lines) then is okay to use datum feature symbol?
Could someone clarify my doubt?
Thank you so much
Burunduk
Mechanical
- May 2, 2019
- 2,574
greenimi,
It is more about what type of feature this is.
I think pmarc intended that a center plane datum would be established by two opposed datum target planes (shown as lines on the edge view of the opposed features).
There would be nothing to attach the datum feature symbol to in a valid way.
It is more about what type of feature this is.
I think pmarc intended that a center plane datum would be established by two opposed datum target planes (shown as lines on the edge view of the opposed features).
There would be nothing to attach the datum feature symbol to in a valid way.
Burunduk said:
If one side is datum feature C and the angled side is datum feature D and a 2X profile (datumless) with two leader lines are shown pointing to the applicable surfaces, why you are saying that you have nothing to attach the datum feature symbol to? Just asking for your opinion.
Or instead of the datumless profile as described above, you can use angularity to A primary and B secondary again with two leader lines pointing to the angled surfaces (and the angled surfaces, again will have datum feature symbols attached, C for one and D for the other). And maybe the tertiary datum feature, as requested by the OP would be C-D.
What could be wrong with this approach?
Burunduk
Mechanical
- May 2, 2019
- 2,574
greenimi,
I don't think you are describing the same scenario pmarc addressed.
But generally, yes, if a single profile tolerance FCF controls both surfaces that are also designated by the datum target symbols such as C1 and C2 then I guess attaching datum feature symbol C to the profile FCF is OK.
I don't think you are describing the same scenario pmarc addressed.
But generally, yes, if a single profile tolerance FCF controls both surfaces that are also designated by the datum target symbols such as C1 and C2 then I guess attaching datum feature symbol C to the profile FCF is OK.
greenimi - interesting find. They do annotate the datum feature symbol, which is on an extension of the surface the datum target points are on... I think the committee bends over backwards to allow practices of convenience far too often.
Why the labels A1, 2, 3 and not A1, A2, A3? Why the label A1 at all, since it is datum feature A, the targets are already known to be A1, A2, and A3. Could it be 1,2,3 instead? That would also be clear.
It's also only partially redundant, which is a generally bad thing, and doesn't specify where the targets are.
But it is there so if someone isn't reading the entire drawing then maybe they will see the "A1, 2, 3" and realize the entire surface is not the datum feature? That seems like an invitation to trouble if someone is looking at a computer extraction that says datum feature symbol A is associated with the diameter of the cylindrical feature. It adds another chance for error.
"The datum feature symbol should be attached
only to identifiable datum features. Where datums are
established by targets on complex or irregular surfaces,
the datum may be identified by a note such as DATUM AXIS
A or DATUM PLANE A."
The standard doesn't say "A datum feature is established by datum targets." Sometimes the terms are used interchangeably, other times as distinct ways to establish a datum.
Why the labels A1, 2, 3 and not A1, A2, A3? Why the label A1 at all, since it is datum feature A, the targets are already known to be A1, A2, and A3. Could it be 1,2,3 instead? That would also be clear.
It's also only partially redundant, which is a generally bad thing, and doesn't specify where the targets are.
But it is there so if someone isn't reading the entire drawing then maybe they will see the "A1, 2, 3" and realize the entire surface is not the datum feature? That seems like an invitation to trouble if someone is looking at a computer extraction that says datum feature symbol A is associated with the diameter of the cylindrical feature. It adds another chance for error.
"The datum feature symbol should be attached
only to identifiable datum features. Where datums are
established by targets on complex or irregular surfaces,
the datum may be identified by a note such as DATUM AXIS
A or DATUM PLANE A."
The standard doesn't say "A datum feature is established by datum targets." Sometimes the terms are used interchangeably, other times as distinct ways to establish a datum.
- Status
Similar threads
- Question
- Locked
- Question