fuel tank dimple dies

The issue at hand is assumed knowledge. Just because airplanes haven?t fallen out of the sky due to the use of tank dies does not imply that the method is correct or a best fabrication/maintenance practice. If it is not a standard practice, and there is no empirical data to back it up, I would expect to be challenged on it. If a supplier of tools creates and sells a product that changes the mechanical bonding interface of the aircraft structure, I would expect said supplier to be able to back their design up with documentation and/or testing. Additionally, I would be skeptical if they were not able to produce them. It all goes in hand with being an informed consumer and builder.

There is a minefield of controversial choices that builders of experimental aircraft have to make (I won?t mention any so the thread doesn?t get hijacked) during their build, some of which may or may not be a best practice for the pros; but who is to say that we shouldn?t be pros and use best practices for our build? The pictures posted show that, in some cases, the riveted bond may not be what we expect. I believe that there is value in them, whether to simply focus on improving my technique, or proof of something bigger. However, nobody has convinced me yet that using the tank dies is a best practice and the pictures give me more data points to help make that decision.

1942 Dimple Instruction Video

The following link is to a 1942 riveting instruction video: http://www.youtube.com/watch?v=dyXEjn7f330

To me it's interesting that in both cases where the top sheet is pre-dimpled (methods 3 and 4), it uses a 100 degree dimple for the top sheet, and a 110 degree dimple or countersink for the lower sheet. It seems that for WWII aircraft it was proper to use a different dimple for substructure materials than for surface sheets.

Perhaps this has some relevance to the current discussion.

Bill,

I thought I would show some results that I came up with a while back. I made up several different samples to see which die produced a better result. Here is the thread.

http://www.vansairforce.com/community/showthread.php?t=70080

Back from Key West. Let's address a few details before continuing.

Pmerems said:

I believe the cracks shown on the rivet head are cracks in the anodization not the base material.

Click to expand...

I had the same initial thought at the time (circa 2009), but the MS20426 specifications suggest anodizing is unlikely.

A 2117T4 alloy rivet ("AD") is usually given a chemical surface treatment, i.e. alodine, and the color is yellow/gold. A 2117T4 can be anodized, but Class I anodize is clear; the rivet color is aluminum gray. Class II is color anodizing, but why would anyone choose the same color as alodine? However, we need not guess; we have the ID system. An anodized rivet would have a letter code notation following the length. No code in this position means yellow/gold alodine.

Sam Buchanan said:

And if there are no safety concerns about the tank dies, and if it is indeed just cosmetic concerns....then why all the fuss about joints not reaching design strength?

Click to expand...

Blisters are just a cosmetic concern until they are all over your tanks, right after you spent $5K~$15K for paint.

No fuss here. I've never brought up design strength, but we can do so if you want to talk about it. I've merely asked you to consider how a dimpled joint with space under the rivet head will behave in single shear. Imagine how the three components will shift and stretch. Then consider the sealant elongation data in post #30.

The relative movement is complex, but the short answer is under moderate loads they would un-nest a little bit, in particular with space under the manufactured head, or even just lacking the pre-tension of a properly set rivet.

clevtool said:

To answer the question about shear, in your example and using your term direct shear, there is no difference in your drawing if the rivet did not even have a head or a tail on it, the rivet would fill the hole and the shear would be based on the material and the cross sectional area of the rivet.

Click to expand...

My term (post #16) was single shear, as illustrated in post #10. And actually, no, a dimpled joint with MS20426AD's doesn't fail by shearing the rivet shank. The head and tail are vital. It's hardly a secret in the RV world. Bill Marvel published beautiful illustrations years ago. You can learn a lot about joint behavior under normal load by studying the failure mode.

http://www.kilohotel.com/rv8/rvlinks/marvelrivets.html