Concerned about under dimpling, Please let me know your thoughts

[rvbuilder2002]

Originally Posted by wawrzynskivp

The deformation of the lower layer is going to happen regardless right?

Not from what I have seen, but if it did, wouldn't it negate the need for forming a bigger dimple in the underlying part?

[wawrzynskivp]

Originally Posted by rvbuilder2002

Not from what I have seen, but if it did, wouldn't it negate the need for forming a bigger dimple in the underlying part?

(I don't think the difference is so much in size as in shape. The skin thickness at the center of the dimple must be less as we stretch that material away from plane. Consequently the angle of the lower layer dimple is the base 100 deg plus approx 10 degrees owing to the additional cone effect of the sheet getting stretched toward the center.)

If the concern about using a sub structure dimple is because the original hole would get bigger, then consider that using a larger angle as is done with substructure angles stretches the material LESS not more. This would allow the stretched hole of the substructure dimple to be smaller than the surface dimple NOT larger. A benefit rather than a concern.

If the sheets are laying against each other after riveting then the dimples would have to have nested, I think that's intuitive.

To nest dimples we have to accommodate the underside of the dimple in the topside of the next layer. We can do this before riveting in a known manner or hope that it fully happens during riveting in a way that we don't control. The need to form a bigger angle in underlying layers is only ever negated if we lack clamping pressure during the drive and the rivet sets before the sheets come into contact.

The chance of an un-nested set seems much more likely when we build in a bad nest. Again, do a couple of good dimples in moderate stock and see just how much gap needs to be taken up by rivet compression.

If we intend to work the underlying shape during riveting rather than in our preparation we are intentionally work hardening that material twice.

You can actually see a difference in sheets combined with nesting dimples vs those combined without nested dimples. It just comes out 'flatter.'

So if the lower layers must accommodate an angle greater that the original surface (100 deg), and they must accommodate the new angle if the sheets are to nest, then why not do it ahead of time and ensure mated sheets before riveting? If it begs testing then why not do that? It would be a pure guess on my part but I would tend to think Lockheed did that already.

[alpinelakespilot2000]

Before you get concerned about the negativity toward the DRDT above, pretty much my entire -9 was done with it and I never had troubles with not getting a nice crisp dimple. Nor did I ever see a difference when I used my impact c-frame. I'm a big fan of the DRDT.

Now, that said, I have no knowledge of Isham dies, but I did use Cleaveland dies which, according to Wally Anderson of Synergy Air, were among the best.

[rvbuilder2002]

Originally Posted by wawrzynskivp

If the sheets are laying against each other after dimpling then the dimples would have to have nested, I think that's intuitive.

True, but they don't.

And from an engineering perspective, we expect that to be the case.

[wawrzynskivp]

Originally Posted by rvbuilder2002

True, but they don't.

And from an engineering perspective, we expect that to be the case.

Mis-written. Meant to say if the sheets are laying flat after RIVETING.

[rvbuilder2002]

Originally Posted by wawrzynskivp

Mis-written. Meant to say if the sheets are laying flat after RIVETING.

I assumed that is what you meant, and that is what my answer was a reply too.
Riveting does not reform the lower dimple to make them nested together in a net fit.

[wawrzynskivp]

Originally Posted by rvbuilder2002

When a .098" diam. (#40) hole is dimple counter sunk, the displacement of material due to the shape change, opens up the hole diameter to right near the maximum limit specified by the Mil Spec.
If an even bigger dimple is formed on the same sized hole, the hole diameter is opened up even further..

I had to edit my last post and it probably got overlooked.

A substructure dimple isn't really 'bigger' it is a different shape, specifically a bigger angle. That bigger angle means less stretch, a hole much closer (smaller) to the original not bigger.

[wawrzynskivp]

Originally Posted by rvbuilder2002

I assumed that is what you meant, and that is what my answer was a reply too.
Riveting does not reform the lower dimple to make them nested together in a net fit.

I think something's getting lost here.

I take two 1" wide strips of .024 or other similar stock. Drill, dimple with the same die and stack. Lots of wobble between sheets, can't get the sheets to mate. Then I drive a rivet and if I get good rivet compression along its shank then the finished stack appears to be very close to mating. Something had to change in those dimples for that to happen.

If I do the same with a sub structure shape on the lower sheet the sheets mate completely before dimpling and look better after driving. How is this not a good thing?

[rvbuilder2002]

Originally Posted by wawrzynskivp

I think something's getting lost here.

I take two 1" wide strips of .024 or other similar stock. Drill, dimple with the same die and stack. Lots of wobble between sheets, can't get the sheets to mate. Then I drive a rivet and if I get good rivet compression along its shank then the finished stack appears to be very close to mating. Something had to change in those dimples for that to happen.

If I do the same with a sub structure shape on the lower sheet the sheets mate completely before dimpling and look better after driving. How is this not a good thing?

I have done test samples were a rivet joint was sectioned and inspected. It showed that riveting does not reform the lower dimple to produce a net fit between the two (even tough to your eye it may look like it makes it fit better But as i mentioned previously, if it did, that would be another reason for not using sub structure dies.

I already explained how it could potentially be a bad thing (maybe you missed that post), but that I can't say for sure because we have never test it.

We have tested using the same sized dies so there is no debate on the strength using that method.

With that being the case, I see no point in using sub structure dies and adding in an unknown factor, when there is no clear benefit (in my mind anyway), but it does add a potential unknown factor.

Apologies to to everyone else for the thread drift (though I guess it is somewhat related to the general theme of the discussion)

[ctremie]

UPDATE!!!!!

CLEAVELAND DIES CAME IN TODAY AND BANGED IT PERFECTLY STRAIGHT!!

IT WAS THE DIES!!



Thank god i figured this out early before putting anything else together!!

Thanks everyone for your input!!