@JoePfeiffer, can you tell me if this is expected. Below testing done with the jar file from this PR.

Here is a set of tube fins with 1:1 aspect ratio:

[As an aside, the first thing I notice is that the drag for the "Tube fin set" still only shows the drag of one fin. It seems to me that Component analysis should show the total drag for the set, so the sum of the columns works correctly.]

Anyway: pressure drag is .056 for one of these tube fins. At 1:1 aspect they should be reasonably transparent to the airflow.

Here is the same rocket with the length of the tube fins increased to 12", for a 12:1 aspect ratio:

The friction drag has scaled linearly, while the pressure drag has increased to .088, a 57% increase.

However, everything I've ever read suggests that as tube fin aspect ratio gets longer, eventually they become completely opaque to the airflow, and act like a capped tube. I would assume that my 12:1 ratio tube fins behave like that. If that's true, then I would have expected two changes:

1. the pressure drag would have increased more than the observed 57%
2. The friction drag would not have increased linearly, because there is no longer flow on the inside of the tubes. Therefore, the exterior component of friction drag should have increased linearly, but the interior friction drag should have gone to zero. Assuming interior and exterior friction drag started about equal, then I would expect a 12x increase in aspect ratio to result in (about) 6x friction drag.

The 12:1 tube fins, with no airflow through them, should also have virtually no impact on CP anymore. That's harder for me to judge by fiddling in OR, though; I don't know exactly what I should be seeing for the different version of the tube fine.

It is (extremely) possible that everything is working correctly and my own understanding is wrong, but if so please explain why so I don't have to ask this stupid question again. :)

I believe what you're seeing with pressure drag and friction drag is correct (at least, it's coming from what seems to be the standard formulas used by real aerodynamicists).

I'm very confident in the friction behavior: the friction on the interior of the tubes isn't actually calculated separately; it drives the pressure drop and so it comes out implicitly in the pressure drag.

I'm less confident in the pressure behavior itself: while we're all under the impression that airflow pretty much stops in a tube relatively quickly, the equations I'm finding for the pressure drop all say it's linear with length. That implies that it takes a much longer tube than we suspect for that "stopped" behavior to manifest. So I'm not liking the results there, but they are what they are.

At the moment, looking at the two rockets I've got data from (back in the original #1333), I'm now seeing pretty good results on the low power rocket (though by no means the suspiciously good results in that PR), but the high power rocket sims are 'way too low. I've been wasting a lot of time trying to incorporate cross section shape (ie rounding on forward and aft edges of the tubes); I'm thinking I might be better off just pushing the current version and see what we get from users.

Wait, were you trying to address this issue or not: https://www.rocketryforum.com/threads/announcement-openrocket-version-22-02-final-is-now-available-for-download.177835/post-2391950. I think back when you filed #2065 I was talking about the linked issue while you may or may not have been talking about the same thing.

That was where it started -- but those good results in the six-tube case relied on the errors in every other case.

It's definitely behaving better... 3 tube fins goes higher than 6 now. :)

@neilweinstock thanks -- your questions pulled me out of the rabbit hole I'd descended trying to model rounded cross sections for the tubes, at which point I finally noticed that the .ork files for both of the test rockets had the wrong tube thickness. It's awfully hard to get pressure drag right when the thickness is wrong!

So this PR now:

1. Corrects the interstices calculations that were causing nonsensical results with numbers of fins other than six.
2. Refines the drag calculations.
3. When a TubeFinSet is first added to a model, it copies the tube thickness from the body tube to the TubeFinSet. It doesn't track it after that; if you change the thickness of the body tube you need to adjust the thickness of the tube fins by hand.
4. Fixes [Bug] error in calculation of interstices area for tube fins #2065

Here are comparisons between the test rockets and the simulated results:

I've attached the corrected .ork files that got these results.
Blue Tuber v3-mod.zip
tubular.zip

Outstanding, a really great fix for the upcoming release.