Structural guy needing help with parallel tolerance

[andriver]

As a structural guy I rarely deal with tight tolerances, but I have a situation where I would like two holes on separate plates to be parallel to each other.

Problem: Lifting a relatively light tank (80 kips) but it has the weirdest lifting eyes (5.5in Diameter hole) which would require a very large shackle for a good hole/pin ratio. Due to clearance with the side of the tank, only a 35-ton shackle would fit which blows up the lifting eye in bearing.

Solution: I decided to design some link plates that go on the both sides of the original lifting eyes, a machined pin will go into the oversized lifting eye hole, and transition to a smaller hole for an 85-ton shackle at the top of the link. The two links are joined by a small pipe to help with any slight out of plane loads (a spreader bar matching distance between lift eyes will be used to minimize any out of plane loads). Note: see attached for diagram.

Question: I want the hole on one link plate to be parallel with the holes on the opposite link plate so that both my machined pin, and shackle pin are bearing "equally" on both link plates. How do I call this out, and what are reasonable tolerances in this regard?

Thank you in advance for your input.

 

[drawoh]

andriver,

It looks to me like you need accurate positioning. Parallelism looks much less important. If you machine rather than punch, you have hole edges perpendicular to the faces, and you have accuracies within a few thousandths of an inch.

The required tolerance comes from your structural calculation I should think.

--
JHG

 

[andriver]

If we flame cut the holes, and then bore to the finished diameter will this still be perpendicular to faces? If that is the case, than I agree with you that positioning tolerance would be all I need.

 

[drawoh]

andriver,

If you have access to a machine shop, I would recommend letting them do all the fabrication of your link. I am not familiar with your fabrication tools. If you can position your bore accurately perpendicular to your face, then it will work.

--
JHG

 

[powerhound]

Just remember that callling them out parallel means they will be parallel, not necessarily coaxial. They could be offset by 1/2" and still be parallel. Perpendicular is the same way. Both holes can be perpendicular to a face but still be 1/2" away from each other. Position is what you're after here.

John Acosta, GDTP Senior Level
Manufacturing Engineering Tech
SSG, U.S. Army
Taji, Iraq OIF II

 

[HGMorgan]

If not using GD&T:
2 holes in-line with-in .xxx
GD&T could use Position or Concentricity.

Harold G. Morgan
CATIA, QA, CNC & CMM Programmer

 

[desertfox]

Hi

Well I opened the attachtment only to find a blank page

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

 

[DiegoLGraves]

For a one-off like this, can you just stack the plates and cut through both at the same time?

Diego

 

[andriver]

Diego,

My thoughts exactly, I was wondering the same thing. This will be sent to a machine shop, when we choose one I can follow up with a contact there to express my needs. I just thought I would get a little guidance on how to specify on a drawing my wishes for the holes to match up (one of the engineers here said to add a note to match them). To me this doesn't tell them how close, obviously perfect is not realistic (or is it, if you machine at same time?)

My current call out of the holes being (+/-) 1/32" of an inch actually surprisingly leads to a sloppy fit that would put an eccentric load into the lifting eye. I also was looking for info on what is attainable, or realistic tolerances without racking up the bill too much.

 

[drawoh]

andriver,

A machinist can work way more accurately than [±].032". [±].005" is a standard machining tolerance. They can get down below [±].001" if you need them to.

It sounds like your tolerances are driven by structural concerns. Work out what you need, and specify that. Tolerances should be driven by your requirements, not your process.

--
JHG