Tolerance Stackup on Bolted Cover where a Composite Positional Tolerance is Used

[RealSaladsamurai]

Hello All -

I am trying to do a tolerance stackup on an assembly to insure that it meets its design intent. In the attached image is a schematic of the assembly. It is essentially a box with a cover bolted to it. It's function is to hold a test article hanging from the lid during a vibration test. Note that the "box" is in fact a weldment made up of four walls welded together.

In the Section and detail view, you can see that the cover has a shoulder that rests in a rabbet style groove in the box. The idea is that during testing, if/when the lid shifts left/right and up/down due to vibrations, the shoulder will take the loading instead of the bolts. This means that the gap G2 between the shoulder and rabbet-wall should always be smaller than the gap G1 between the bolt and the clearance hole in the cover. That way, the shoulder will always strike the wall and the bolt will never contact the inside of the thru hole.

I will walk through what I have done so far, but note that my confusion mainly stems from what happens when we introduce a composite position.

Again, the problem statement is to ensure that G2 MAX is always less than G1 MIN

Calculate G2 MAX
G2 MAX = (D2 MAX)-(D1 MIN)

Calculate G1 MIN
To calculate G1 MIN, I calculate the Virtual Condition of the thru hole (VC_H) in the cover. I then calculate the VC of the fastener (VC_F) in the box (the weldment) and then take the difference.

Calculate VC_H

VC_H = D - T_H

where:
D = MMC thru hole diameter
T_H = tolerance of position of thru hole
​

Calculate VC_F
From appendix B.5 of ASME 14.5 2009 (derived from fixed fastener formula when no projected tolerance zone is used):

VC_F = F + T_F*(1+2P/D)

where:
F = MMC fastener diameter
T_F = positional tolerance of tapped hole
P = maximum thickness of thru hole (cover)
D = minimum depth of tapped hole
​

Calculate G1 MIN = VC_H - VC_F

​

My confusion is what to use for the value of T_H and T_F?

Both are positioned using a composite positional tolerance and so there is an upper feature control frame (FCF) that gives the feature locating tolerance of .021 and there is a lower FCF that gives the feature relating tolerance of .005.

I cannot seem to wrap my head around which one of these effects the gap size ... my gut is telling me that only the locating tolerance of .021 matters, but I could use a little guidance on this.

Thank you for looking.

________________________
FEMAP v11.1.0
MSC Nastran v2013

 

[drawoh]

RealSaladsamurai,

A composite tolerance is used when the hole pattern itself is more important than its location to all the datums. In other words, you want the screws to work, but you don't care where the datum features wind up. This is useful if your holes are drilling or punched, located from an inaccurate edge like a bent piece of sheet metal or a welded flange.

I don't bother with the standard. This is easy to work out from first principles.

Your tapped hole has a positional tolerance. A screw inserted into the tapped hole occupies a space consisting of the screw diameter, plus the positional tolerance. You need to account for the screw being non-perpendicular and extending above the face with the tapped holes. Either specify a projected tolerance, or increase your occupied space a little.

Your clearance hole at MMC/MMB must not intrude into the space occupied by your screw. Your hole, located at exact nominal requires a diameter at least equal to your occupied space. Given a positional tolerance of your clearance hole, the hole must equal the screw occupied space, plus your positional tolerance. If you specify your positional tolerance at MMC/MMB, your bonus positional tolerance is however far your hole exceed minimum diameter.

--
JHG

 

[RealSaladsamurai]

Hi drawoh -

Thank you for your reply. If I have understood you correctly, what you have described insures that the 2 parts will bolt together. You have described, in words, how to intuitively compare the two virtual conditions.

However, beside insuring that 2 parts bolt together, I also need to insure that the bolt never contacts the inside of the thru hole. This can be accomplished by designing the thru hole and tapped hole configuration such that at worst case (size, form, orientation, and location of all features involved), we do not have line to line design. We need to have a gap (G1). And that gap needs to always be larger than the gap over at the shoulder (G1). This is the same as saying that the shoulder will always "hit first".

That is where I am having trouble. Thanks again.



________________________
FEMAP v11.1.0
MSC Nastran v2013

 

[drawoh]

RealSaladsamurai,

Add your required gap size to your clearance hole.

--
JHG

 

[desertfox]

Hi RealSaladsamuri

If the lid is meant to move when bolted down then what you're doing is fine but if the lid is not meant to move up down or side to side then you need to increase the preload of the bolts.

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[pmarc]

RealSaladsamurai,

Based on information provided, it looks like for this specific analysis you should most likely use values defined in upper segments of composite position callouts.

Why am I saying 'most likely' and not 'for sure'?
Because your description does not actually clarify which datum features the holes in both parts are positioned from. I assume both composite position callouts reference to a primary datum plane which is derived from mating face, right? Again, I assume the secondary and tertiary datums are derived from width/height of the cap shoulder and box rabbet-walls respectively, right? If that is true, .021 should be used in calculations of gap G1.

Another unclear, yet very important, point... Assuming that the datum features have been selected the way I described above, at what material boundary have the secondary and tertiary datum features been referenced in both composite position callouts? RMB (no modifier after datum letter) or MMB (M modifier after datum letter)?

I am asking because your calculations of G1 will work only if the secondary and tertiary datum features are referenced RMB. But if the secondary and tertiary datum features are defined at MMB, you will have to take into account additional factor that is commonly called 'datum shift' or 'datum feature shift'. And in that (MMB) case I don't think the virtual-condition-sizes-comparison approach will work properly.

The bottom line is, without seeing the drawing or knowing exactly how the features in question are dimensioned and toleranced it will be difficult to give you clear and meaningful answer.

 

[RealSaladsamurai]

drawoh: Thanks again for our help;that seems the most sensible solution.

pmarc: Thanks for adding to the discussion. You are correct that my datums have been chosen in this fashion. I guess I never asked anyone explicitly before, but I have always quietly assumed that the fastener formulas will ONLY work if the datums are chosen in this "corresponding feature" way. Other wise, there would be additional tolerance stack tracing back to the "corresponding features."

________________________
FEMAP v11.1.0
MSC Nastran v2013

 

[pmarc]

RealSaladsamurai,
So at what material boundary, RMB or MMB, the secondary and tertiary datum features are referenced in the position callouts?

 

[drawoh]

RealSaladsamurai,

I have notes on hole tolerances on my website. I have worked all this out in some detail. Once you have sufficient datums to fixture your part, the equations apply. It does not matter how you apply your datums, as long as your part cannot move. If your secondary datum is hole[ ]1 and your tertiary datum is hole[ ]2, the equations apply to hole[ ]3.

--
JHG

 

[RealSaladsamurai]

pmarc: They are RMB. We deal with all one-off items, so I think that RMB is very typical around here ... nobody is looking to save any money (must be nice...)

drawoh: Great documentation and great presentation; I appreciate the reference.

________________________
FEMAP v11.1.0
MSC Nastran v2013