I had a "well duh" moment this morning while thinking about the countersinking issue yesterday. When the dimple and countersink were tested together initially, I am certain we didn't take into account the thickness of the dimpled material. Different thicknesses of sheet will have dimples that are slightly different in size on the back side. I'm pretty sure we just grabbed a piece of scrap and checked the fit, not thinking about this aspect. That being said, I'm not worried about it in the slightest. The rivets are fine, and there's almost no perceptible wave of the skin in between the rivets, meaning it isn't being pushed up and down between rivets. I got more information from my question on VAF (what an incredible resource). Every question that could possibly be asked, has been at some point. They pointed me to a post from years ago directly from Van's. Van's cautions going too deep on the countersink, but says that a skin not laying perfectly flush really has no strength concerns. They showed examples of what they mean by doing a cutaway of different depths. Bottom line, go flush or up to .007" deeper than flush and it'll be fine. I must have missed that in the manual, because I don't recall seeing that called out. Probably just information overload and I didn't connect the dots. From Van's post on VAF:
If you go too deep, you will still have no gap between the skin and the sub structure but the strength is likely compromised, With countersinks that are slightly too shallow (as in the attached photo) the strength is no different, but there is assurance that it is not too deep. That is the reason for specific recommendations in Section 5 of the manual. The depth was derived from doing samples of different depths and then machining off 50% to allow magnified inspection of the rivet joints.
Countersinks only deep enough for a rivet head still produced totally acceptable riveted joints (though not quite as nice aesthetically). Countersinks any deeper than about .009 " beyond what would be needed for a flush rivet began to produce visible gaps in the riveted joint.
My outcome on the rear spar/skin looks almost identical to their first picture of "Net CS, Nom Dimple" in terms of how much space I see under the skin. Remember that this rivet shank shown in the pictures below is only 3/32" wide, which puts the gap in perspective. Yes, I know I'm a perfectionist and nobody else will ever notice. I'm learning that perfectionism should mean "structurally sound" and not actually aesthetically perfect, otherwise I'll never finish the plane. The aesthetics can improve on plane #2.
