What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

dimple comparison

Tom Martin

Well Known Member
I like to try different things on new projects and noticed that Cleveland tools now has a substructure dimple die set. I have noted in the past that sometimes dimples do not seem to nest properly, leaving an ever so slight gap between the parts. Thousands have airplanes have been built using just one set of dimple dies but I thought i would give it a try and so I ordered a new set of standard, and substructure set.
In use the parts do nest nicely together although what you see in the finished product is fine with or without the substructure dies. What I did notice is that the hole size increased with the substructure dies. I have a set of Avery dies and did a quick test of the three dies.
259zltf.jpg


"A" stands for Avery
"CS" stands for Cleveland Sub Structure
"C" stands for Cleveland standard dies
All dies are for 3/32 rivets (the small ones)

(In the picture the Avery hole looks bigger but that had to do with camera angle, note the smaller crisper look of the Avery dimple)

I used number drills to test the hole size

A Van's RV14 prepunched hole, right out of the box, will accept a #38 drill (.1015")
A normal Avery die produced a hole #37 (.1040")
A normal Cleveland die hole size the same #37 bit
A Substructure Cleveland die will accept a #34 drill (.1110")

Thus the hole size is 7 thou larger using the Substructure die. It works except that hole now takes a bit more rivet to fill it and any larger hole can be more difficult to rivet. What is more worrisome is that when used thicker material, 40 thou for example, the same as is used for the rear wing spar, I could see some small stress risers around the edge of the dimple hole. This was not evident on the thinner rib material.
Both the Avery and Cleveland normal dies produce a nice final product. The differences are purely up to the individual as to what looks better. The Avery die actually produces a crisper final hole with the outside of the dimple measuring 0.220" vs. and O.D. of Cleveland at 0.230".

As for myself I am not going to continue using the Substructure dies as I see no real visual improvement to the final product and there the two issues of increased hole size and subsequent stress risers on thicker material. The larger hole size was seen when used on thick or thin material.
I would suggest if you really feel the need for a substructure die use the cleveland die and nest an Avery die within it.
 
Last edited:
sub-structure dimple dies

I bought the sub-structure dies from Cleveland also, as well as their standard dimple die. What I found doing the practice kits (I did two) was that I was not satisfied with the results using the sub-structure dies. The clecos would not hold in the larger holes.

I've since learned that the sub-structure dies were intended to be used under the Cleveland tank dimple dies, which have a little larger dimple to account for Proseal under the rivet heads. Too many choices.
 
Dies

I've used all three Cleveland dies. According to Cleveland, they recommend the standard dies on skins and substructure on substructure parts unless it's a tank. They recommend tanks skins be dimpled with tank dies and substructure underneath.

I tested on samples and also had issues with clekos slipping out but what I found was a difference in clekos was causing the problem. Some have a smaller center pin which effects the grip in a dimpled hole.

Ultimately, I used tank/substructure on the tank and found really tight fit of parts. The proseal was a super thin layer and my tanks tested leak free. On other areas, I use standard/tank/substructure respectively depending on the number of layers. If I use the right clekos, I don't have any falling out.

This is only my opinion.
 
I also will not continue to use the sub-structure dimples

I also used the sub-structure dimples on my wings' structure...before I read about he hole being stretched out. While the parts do seem to nest together better before riveting, the 3-3.5 rivets called for in the plans were not enough to fill the larger hole and leave a tall enough shop head. I had to use 3.4s instead.
 
Section 5 indicates Vans does spec a slightly shorter rivet in some cases due to longer rivets folding over, knowing the shop head will be out of spec.


What Section 5 says, is that a slightly shorter rivet is sometimes specified, than what would have 1.5 D penetration prior to setting, and/or would result in a shop head that met the .5 D / 1.5 D Rule of thumb, as long as it still results in a shop head that does meet the MIL spec.

A rivet size is never (purposely) specified, knowing that it would likely result in an under spec. thickness or diameter shop head.

The .5 D / 1.5 D rule of thumb, isn't a spec. (specification), it is just that, a general rule of thumb that can be used to simplify the rivet selection and rivet head checking process.

The actual MIL spec thickness and diameter values for rivet shop heads are smaller than .5 D and 1.5 D, respectively, for thickness and diameter.
A copy of the document, which contains a chart of the min. and max. thicknesses and diameters of shop heads can be viewed HERE

Additional info added because of questions received....
There is a lot of information in the linked document but the info related to my post is in Table III

Note that the minimum shop head diam spec for a 3/32 rivet is .122". If the 1.5 D rule of thumb is used instead, the minimum is .141" (1.5 X .094).
The minimum shop head thickness spec is .038". If the .5 D rule of thumb is used, the minimum is .047".
So, as you can see, a rivet can be short of meeting the general rule of thumb dimension, but still meet the actual MIL spec.
 
Last edited:
A customer sent me a link to this thread, interesting perspective that I had not heard directly. Tom, could you post the pictures on the top and bottom of the samples after they were riveted? I am just interested to see the difference (or lack of) that you had for results. Tom has certainly built enough to know what is going on here, but for others, keep in mind that when the rivets tighten up, the material deforms the dimples so they will be tight together. If you backrivet the samples, you are controlling the top skin and forcing the substructure to meet it. That is why backriveting wherever you can give such nice results.

It naturally will stretch further when dimpled deeper, there is no way around that. I agree that the quality of 'cleko' makes a difference in holding either in the SS die or the Tank die. Wedgelocs are the best quality of 'cleko' that we have found, very consistent, smooth operating, and all the ears track in line with one another.

Rivet length. Some quick math given the hole sizes stated, and a guess on my part that the affected area would be about .030" in height (controlled by the bottom of the standard die to the bottom of the SS die). Area of hole in normal die .0085in^2, area in SS die .0097in^2. Volumes .000255in^3 in standard, .00029in^3 in SS. That 0.000035^3 in the shank of a 3/32" rivet is
.005" in length. Where .5 in a dash length is .031". Somebody check me here, but I am confused by needing the longer rivet. I do know that with my projects, our rivet gauge (which is the nominal size) often slips over the rivet with a few thousands side play when using the called out rivet, and the nominal height. Again for new builders and lurkers, nominal and spec. are not the same thing. Thanks to Table III on the spec. previously posted, it looks like there can be .018" of side play... I think I have been too particular!
3/32" rivet, 0.122" min. head diameter, 0.038"-0.050" thickness.

There are certainly many opinions around, and that is why our return policy states, 'If you are not happy with anything you buy from us, return it for a full refund'. I, personally, hate so see money spent for a tool that is not used, or you are not satisfied with, so please return them to us. When Buzz first started, he bought lots of tools that were inferior. That was so frustrating that we started a tool company, and we certainly don't want to be involved in frustration or disappointment. After all building airplanes is one of the most fun things there is to do!

-Mike
 
I hope I'm not going off-thread too much here by bringing up the subject of dimples in countersinks. I've posted this question in a couple of other forums.

The plans call for countersinking the VS rear spar doubler to attach it to the (dimpled) rear spar. As a first-time builder I'm trying to get the correct countersink depth on the rear spar doubler. I'm practicing on scrap Aluminum.

When I use the manual's instructions on countersink depth (7 clicks beyond the flush rivet-head depth) and test fit the dimple into the countersink, the dimple is a long way from being fully into the countersink. When I put the rivet in and set it, then hold up the pieces to the light, there's still a small gap between the 2 sheets, the countersink is just too shallow to absorb the dimple. In fact, it's obvious that the amount of material in the dimple won't all fit into the small countersink volume, and the dimple material which won't be absorbed into the countersink will just serve to push the 2 surfaces apart. Again, as a first-time builder, excuse my ignorance, but I want to be sure the 2 surfaces are in good contact, which would seem a strict requirement for structural reasons.

What am I missing? Should I try the Avery (smaller dimple diameter and volume)? The dimple volume would seem to be a major factor in the countersunk case...

PS When I fit 2 dimpled surfaces together, the dimples still don't nest completely, but the riveting process seems to cinch things up nicely, not so in the countersink case.
 
Countersinks

I hope I'm not going off-thread too much here by bringing up the subject of dimples in countersinks. I've posted this question in a couple of other forums.

The plans call for countersinking the VS rear spar doubler to attach it to the (dimpled) rear spar. As a first-time builder I'm trying to get the correct countersink depth on the rear spar doubler. I'm practicing on scrap Aluminum.

When I use the manual's instructions on countersink depth (7 clicks beyond the flush rivet-head depth) and test fit the dimple into the countersink, the dimple is a long way from being fully into the countersink. When I put the rivet in and set it, then hold up the pieces to the light, there's still a small gap between the 2 sheets, the countersink is just too shallow to absorb the dimple. In fact, it's obvious that the amount of material in the dimple won't all fit into the small countersink volume, and the dimple material which won't be absorbed into the countersink will just serve to push the 2 surfaces apart. Again, as a first-time builder, excuse my ignorance, but I want to be sure the 2 surfaces are in good contact, which would seem a strict requirement for structural reasons.

What am I missing? Should I try the Avery (smaller dimple diameter and volume)? The dimple volume would seem to be a major factor in the countersunk case...

PS When I fit 2 dimpled surfaces together, the dimples still don't nest completely, but the riveting process seems to cinch things up nicely, not so in the countersink case.

It is pretty hard to go wrong following the plans and build manual. Many of us, myself included, make a sport out of over thinking many tasks. This tends to slow down building progress. If it makes you feel better, use a few more clicks on the cage. Won't make a bit of difference one way or another.
 
As a follow up to this discussion I wanted to look at the final product. I really do not care how nicely the parts nest before riveting but what the final product will look like. Both the Cleveland and Avery dies were what they call spring back dimple dies.
I took two pieces of .025 metal and drilled them all at the same way, they were all dimpled with a C frame.
I used standard Avery dies, Standard Cleveland dies and a standard cleveland die with a Sub Surface die on the lower sheet.

Before riveting the Cleveland dimple with the Subsurface dimple did in fact nest together better then the Avery dies, or the Cleveland dies.
In this picture the left one on your screen is the Cleveland with the Sub Surface die
The centre picture is the standard Cleveland dies
The right one on your screen is the standard Avery dies
There was some warpage during the cutting process but the one on the left did "nest" together better. The hole of the Sub Surface die was visually larger, similar to what was found in my original post on this thread.

vh69p0.jpg


Then I riveted the parts using a squeezer and set the heads all the same. The part was cut and the surface buffed.
My apologies for the orientation of the picture
The top picture is the Avery dies.
The middle picture are the Cleveland dies
The bottom picture is the Cleveland die on top and the Cleveland Sub Surface die on the bottom.
Note that all three test pieces are nested firmly together. For overall appearance my preference would be the Avery dimples
I draw your attention to this bottom picture, note the small gap between the top and bottom sheet, next to the rivet. I believe that using the slightly larger Subsurface die on the lower sheet means that there is not enough "material" on the top sheet, with the smaller Cleveland dimple, to fill the hole between the parts and the rivet.
Over the years I have purchased tools from both Cleveland and Avery, they are both excellent suppliers and I have no affiliation with either supplier.
Based on my experiences with the Sub Surface dies I can not recommend their use.

2uzp9tx.jpg
 
Tom, Verrrrry interesting, great work.

I would really like to see similar x-section comparisons on a dimpled surface nested into a countersunk surface with various countersink diameters/depths (see my previous post on this thread)
 
Testing

Interesting. Thanks for posting it.
I don't use a SS die unless the dimple going into it is a tank die dimple. I used them on the tank and areas with three layers where the top layer is standard die, middle layer tank die and bottom SS die. The wing walk was a good example and they seemed to nest together very well but I have not cut samples to test. It would be interesting to see. For two layers, I dimple the bottom layer with tank dies. All are Cleveland.
 
Last edited:
Back
Top