Coaxiality Of Opposing Holes In Welded Assembly 1

[donatim24]

A picture of the welded assembly in question is attached. The holes I refer to are circled in blue.

I came across thread1103-398706: opposing face datum holes when investigating a similar problem, and found it very informative. I have a question about my application, as there is an important distinction between my problem and that of the OP in the referenced thread. On the assembly I'm working on, the opposing holes are on two identical pieces that are welded to a plate. The holes are drilled in the identical pieces before they are assembled and therefore the diameters are not defined at the assembly level. Does that mean that the method described in Y14.5-2009, 7.6.2.3 would not apply to my situation? I'm assuming not.

One of the replies in the linked thread mentioned assigning a datum to both holes and tolerancing each hole relative to those two datums through a position tolerance. I thought it would be simpler to set the axis of one hole and tolerance the other hole relative to that datum, but I can see how that would be ambiguous as to which hole the axis belongs to. Can anyone tell if my approach is valid and if there is another way to accomplish this task?

Thank you very much,
Tim

 

[3DDave]

I believe that the diameters and their tolerances can be restated at the assembly level, though most people think that information must be reference. See 2009 Y14.5, 1.4(o). I stand on the the Y14.5 standard that a dimension applies only at the drawing level at which they are applied and therefore any dimension does not apply at the next or higher levels. Because it would not apply then it is not 'double dimensioning' to call it out again at a higher level.

Typical reasons for saying different is that there is a chance for the two callouts to be out of synchronization, that QC refuses to check the same feature twice, that that is just not done, that the feature already exists and therefore cannot be 'redefined.' None of which are in the standard.

I believe one can also 'force' a virtual condition boundary description at the assembly level. I don't have a specific reference for that, but it's been mentioned.

 

[pmarc]

Everything, as usually, depends on functional requirements, so without knowing what really should drive orientation and location of welded pieces in the assembly, I would say that both holes could be positioned to some datums derived from other features in the assembly, and that would indirectly do the orientation and location job for the two components.

If the intent is to control the holes with positon at MMC, the size of the holes can be given as +/- dimensions and then even called out as reference.

 

[chez311]

I would say that both holes could be positioned to some datums derived from other features in the assembly, and that would indirectly do the orientation and location job for the two components

If the requirement was still the holes to a position tolerance at MMC (whether to each other or another feature in the assembly) wouldn't they still need a size tolerance applied? I would say the same in regards to a tolerance held at RFS as well, however if the size tolerance were omitted it might be that the intent could be interpreted more easily, without perhaps being totally compliant to Y14.5

In addition to the above, whereas the assembly/welding process cannot be held accountable for the significant majority of the formation of the hole's size/form I would think that depending on the heat applied/type/length of welds specified, material, section thickness, part/hole relative size and distance to weld, and other factors the act of welding could still account for some non-trivial impact on both the size and form of the holes (ie: making a hole oblong changes its local size), and not just their orientation relative to one another. This could also apply to other features on the welded parts, for example if theres a flatness requirement on either side of the tabs shown as it is not a question whether welding will warp them, its just to what extent. Heat and force can do unexpected things to a part.

 

[chez311]

pmarc,

I should add that I know you addressed adding a size tolerance at MMC, the only reason why I formulated the question the way I did is because I wasn't sure your first statement regarding an alternate datum selection was to somehow get around requiring a size tolerance.

 

[pmarc]

chez311,
The reason I suggested to call out +/- size tolerance as reference was to say that the size requirement is not mandatory at assembly level, yet the size tolerance provides necessary info to calculate allowable bonus tolerance for position at MMC requirement.

 

[donatim24]

Hi all,

Thank you for your replies so far. If it helps, the welded assembly serves as a frame for a roller and axle. The axle fits through two bushings, one bushing inserted in each of the drilled holes. There is significant clearance between the holes and the bushings, though the frame is designed for approximately two-time use (first, to get the product into position at the customer site, and second, during product decommissioning). The design of the frame is such that I'm more concerned with where the holes are positioned relative to each other than where they are in the welded on "ears" of the assembly.

 

[chez311]

pmarc,
I can understand your point about reference dimensions/tolerances. See the second part of my earlier post - it may not apply in this case, but I would think there could be situations where you would not want these to be simply reference "non-mandatory" dimensions and actually have them inspected snd controlled at the "assembly" level ("assembly" in quotes because in this case welding actually creates a single inseparable part). In that case one could apply the same or looser tolerances than at the component level - the QC department might complain about double inspection however if it is not inspected that could be ignoring the many potential effects that welding has on the part.

3DDave,
See my above comment - in this case what would be your suggestion? A custom note? Or do you think its not worth the trouble? I can imagine a situation where this applies - for example a tab as shown with a hole that is very large relative to the tab itself with one or more thin sections around the hole's perimeter. In this case any warping of the tab post-weld could change the local size of the hole, not to mention any size/form changes imparted to the tab itself.

 

[pmarc]

chez311,
I agree with you. There could be situations as you described.

donatim24,
If you are more concerned with how the holes are positioned relative to each other, you could, for example, go with a composite position tolerance having loose tolerance in the upper segment relative to "global" datums, and tighter tolerance in the lower segment with no datum references controlling mutual spacing between the holes (similar to fig. 7-49 in Y14.5-2009). This is just one idea of how this could be tackled.

 

[donatim24]

Thank you all again for your help!

I originally set out to tolerance the holes in a very similar manner to what pmarc described in his most recent post. I got thrown by my assumption that the datum reference for coaxiality was controlled on a separate drawing. I now see the validity of restating the controlling diameter at the assembly. Thanks to your input, I plan to restate the hole diameters at the assembly level in the event of any distortion during welding and use a composite position tolerance to control both location and coaxiality.