One would think they cannot help, but some people say they make no difference.
And someone told me recently that somebody did try replacing them with flush rivets and found the plane went slower. Is that possible? One can imagine they might work as vortex generators, and certainly the skins of racing boats have a dimpled surface.
Laminar flow is less draggy than turbulent, but the wing needs to be designed for laminar flow, eg Mooney, Comanche.
Draggier, thicker wings may benefit from turbulent flow being energised by rivets or VGs? Although this would be effective mainly at higher AoA? So should it have an effect on cruise speeds? VGs typically have a small speed penalty.
The Wright brothers threw away the maths theory and spent all their effort in the wind tunnel – controlling why the ’plane slowed down after filing down the rivets would need careful testing, but in theory the effect should be similar to a good wash and polish – a couple of knots?
I heard of an early Boeing that was accidentally made with protruding rivets on the rear half of the fuselage, before somebody noticed. Allegedly it went a few knots faster than the ‘correctly’ built versions. I’ll search out the story in the book again if I have time.
When the Spitfire was being developed for production, one of the big questions was whether the rivets all needed to be flush or whether they could use protruding rivets instead (which were a lot easier and cheaper to use). So a Spitfire was built using flush rivets only and then halved dried peas were glued to various locations on the wings and fuselage, in order to discover by trial and error which areas cost them speed, using the protruding rivets.
They then went on to assemble the largely aluminium airframe using a mix of flush and protruding rivets made of magnesium which, with the addition of water, over time produced a nice electrolytic reaction. Which is why, when the press writes about the Lincolnshire farmer who is digging holes around Burma looking for “perfectly preserved Spitfires wrapped in tarred, greaseproof paper, in wooden crates”, the task of re-commissioning those phantom aircraft is being heroically underestimated.
The Bonanza and Baron use flush rivets and ones with protruding heads, depending on where they are located. Most, but not all of the rivets on the top surface of the wing are flush.
Can anyone suggest how the paperwork for this could be done?
At worst, you would need a complete flight test regime – because the rivets might be generating some “useful” turbulence.
The labour would of course be very significant but that’s a different topic. Everybody knows how to do flush rivets!
Not worth it – in time. How are you going to dimple/countersink all those ribs/spars? Remove all skins and start again?
There was some tests done somewhere, where only a small part of the leading edge mattered.
Interesting tuft test stallin a RV7 (must be the lens – it looks like a 9 wing):
Peter wrote:
Can anyone suggest how the paperwork for this could be done?
You would need to rebuild your while plane to do this. The skin is too thin for countersinking, which means dimpling is needed. It certainly will look (much) better, but any improvement will be in max speed, 2-3 knots maybe? There are lots of discussion about this for the Sonex line of aircraft. They can be built either way. I have not seen any conclusive evidence that is not offset by minor changes in rigging and trimming.
Didn’t the Convair880, competitor to the B707 and DC8, have non-rivetted wings, which surprised the designers by being less efficient than rivetted wings?
There was a Boeing (I think 707) accidentally built with raised rivets over the tail, which went a few knots faster than the correctly built aircraft. It’s recounted in this book:
Which is otherwise one of the most boring aviation books on my bookshelf.
