Gapped fins....

[SolarYellow]

I haven't rabbit-holed the documents, but I have a vague general idea that OR is just applying the Barrowman equations to determine the CP. Do they look at anything but surface area? Like a good symmetrical airfoil being more efficient at producing lift as the angle of attack departs from zero than a flat fin with rounded edges?

 

[H. Craig Miller]

When split fins are created using a single freeform fin set in OpenRocket, the split is ignored, and a "jagged fin" warning is displayed; using this technique results in the same drag as if the gap did not exist. When split fins are created using two separate fin sets, OpenRocket (and RockSim) apply the Barrowman (or its own drag) equation separately to each fin set; the drag is the same as if one fin set is at 0 degrees and a second fin set is rotated mid-way between the fins in the first set.

To generally see how this works, create two touching three-fin fin sets that have the same dimensions, and look at where the CP is. Now rotate one of the fin sets, the CP does not move. Now delete the forward fin set and increase the bottom fin count to six. The CP should be in about the same place as with both fin sets, though not exactly (because of the fin set placement on the body tube).

Neither OpenRocket nor RockSim account for gapped fins based upon the size of the gap.

 

[KenECoyote]

What I'm saying is that, unless I'm mistaken, a change in airflow that has a tendency to increase (or decrease) stability is not always a change in the CP per sé.

True. In fact, the dimpling actually helped to increase airflow and efficiency rather than add drag. Still, I wonder if there's something else going on and I tried to find video of wind tunnel tests on split fins, but was unsuccessful.

 

[lakeroadster]

When split fins are created using a single freeform fin set in OpenRocket, the split is ignored, and a "jagged fin" warning is displayed; using this technique results in the same drag as if the gap did not exist. When split fins are created using two separate fin sets, OpenRocket (and RockSim) apply the Barrowman (or its own drag) equation separately to each fin set; the drag is the same as if one fin set is at 0 degrees and a second fin set is rotated mid-way between the fins in the first set.

To generally see how this works, create two touching three-fin fin sets that have the same dimensions, and look at where the CP is. Now rotate one of the fin sets, the CP does not move. Now delete the forward fin set and increase the bottom fin count to six. The CP should be in about the same place as with both fin sets, though not exactly (because of the fin set placement on the body tube).

Neither OpenRocket nor RockSim account for gapped fins based upon the size of the gap.

And now we know.

Thanks @H. Craig Miller for taking the time to provide guidance on this.

 

[neil_w]

Would folks want to see a warning when there are multiple sets of inline fins, indicating that sim results might inaccurate?

 

[OverTheTop]

For laminar flows (around zero AoA) you can get a boundary layer forming, which gets thicker as it moves along the surface. Think about how the viscosity of the air causes it to slow down as it gets near the fin. The gap would break that laminar flow and reset the boundary layer thickness. I have not managed to convince myself that this is what is going on here, but maybe it is.

It is the same reason heatsinks in electronics have textured/wavy surfaces, to break up the boundary layer and increase the thermal coupling to the moving air.

 

[SolarYellow]

For laminar flows (around zero AoA) you can get a boundary layer forming, which gets thicker as it moves along the surface. Think about how the viscosity of the air causes it to slow down as it gets near the fin. The gap would break that laminar flow and reset the boundary layer thickness. I have not managed to convince myself that this is what is going on here, but maybe it is.

It is the same reason heatsinks in electronics have textured/wavy surfaces, to break up the boundary layer and increase the thermal coupling to the moving air.

I think the posts from people who are part of the "OpenRocket Development Team" have made it pretty clear that OR isn't doing any calculations regarding the physics you're speculating about. OR doesn't have a magic wind tunnel built into it that will allow one to discover new phenomena.

 

[neil_w]

I think the posts from people who are part of the "OpenRocket Development Team" have made it pretty clear that OR isn't doing any calculations regarding the physics you're speculating about. OR doesn't have a magic wind tunnel built into it that will allow one to discover new phenomena.

It would be nice to imagine that sometime in the future, OR could be doing its simulations with CFD, thereby incorporating all these effects, I think. Regrettably, that is not the near future; who knows when that would be feasible.

For now, it's Barrowman, with all the caveats and disclaimers that go with it. Fortunately, even with its limitations, it's still quite useful.

 

[Nateairman]

It is certainly an interesting bit of designing to be aware of.

I originally looked at rockets with gapped fins that are known flyers to take some of the guess work out.

If a rocketeer downloads the frenzy .ork file and sims a bunch of different motors for flights using the caliber as a set up guide, it would be very good to know that the sim software cannot actually account for fluid dynamics in this situation and that Barrowman equations are actually used (not fin multiplication) and that CP could be much different than anticipated. IF you are building the frenzy would you be better off changing the sim file to a continuous fin to match the area of the gapped fins to better represent what may be happening or do you leave the fins as they are in the .ork file and fly a ?

I enjoy scratch building my rockets as there is a lot more planning and designing especially when it comes to flight design and characteristics. The gapped fin idea came to mind as an easy way to solve some CP issues with out adding a lot of weight.

Anyhow!

I really appreciate everyone chiming in and getting the conversation going on a scientific level. I am hopeful that there can be some more of that.

I am all for gapped fins.

It seems to work well on airfoils...... leading edge slats.

 

[bjphoenix]

This is interesting or disconcerting. I have a couple of rockets in construction with gapped fins and I was relying on OR sims for stability. They are built but awaiting painting so I don't know when they will fly. One is styled similar to the Frenzy, downscaled to BT55 with an 18mm mount, the other is styled similar to the Wildman Darkstar,, downscaled to BT60 with a 24mm mount.

 

[rharshberger]

This is interesting or disconcerting. I have a couple of rockets in construction with gapped fins and I was relying on OR sims for stability. They are built but awaiting painting so I don't know when they will fly. One is styled similar to the Frenzy, downscaled to BT55 with an 18mm mount, the other is styled similar to the Wildman Darkstar,, downscaled to BT60 with a 24mm mount.

As long as you use two separate sets of fins instead of one set with a complex shape (one fin with a cut to look like two fins) it should sim just fine, I have 3" Frenzy that sims reliably once I adjusted surface finish to get a modified Cd.

 

[OverTheTop]

I think the posts from people who are part of the "OpenRocket Development Team" have made it pretty clear that OR isn't doing any calculations regarding the physics you're speculating about. OR doesn't have a magic wind tunnel built into it that will allow one to discover new phenomena.

True, but possibly worth adding to the discussion.

 

[Nateairman]

This is interesting or disconcerting. I have a couple of rockets in construction with gapped fins and I was relying on OR sims for stability. They are built but awaiting painting so I don't know when they will fly. One is styled similar to the Frenzy, downscaled to BT55 with an 18mm mount, the other is styled similar to the Wildman Darkstar,, downscaled to BT60 with a 24mm mount.

How did you make your fins in the sim file? As two fins with a gap or as a free form fin with a gap?

To be safe it seems obvious to use the less aft CP result (that is one fin that incorporates the gap) but, that is like only going halfway with a design.

 

[Nateairman]

Has anyone been able to reproduce the ginormous CP move shown in @Nateairman 's post #5? I'm seeing modest CP movement backwards when the fin is split but nothing nearly as crazy as shown in that post.

Neil,

I simply took photos of two rockets I had open and would need to go through the details to be 100% sure I didn't make a mistake in the two different designs.

I used the frenzy.ork after seeing what had happened with the others. I noted about a caliber difference by removing the gap and increasing the fin size to include the gap.

 

[BigMacDaddy]

Quick mock-up - I also get a shift (4cm in this design) by putting a gap in the default fin on a BT-60 tube with a default cone (not sure if CP shift is exaggerated).



EDIT: If this is modeled as a free-form single fin with a gap the shift is not nearly as significant (and is in the opposite direction).

 

[neil_w]

At this point we're pretty clear that this is a direct result of the way the Barrowman equations work. Rocksim yields similar results.

What is not clear is how we can improve the way we handle this situation, in part because we don't really know what the most *correct* answer is.

We could put in a warning when you have inline fins, that CP may be optimistic. Or we could sim gapped fins as a single complex freeform fin. But without knowing what results are closest to reality, it's a bit dicey to code in heuristics like this.

I'm almost wondering if, in cases such as this, we should specify a range of possible CP values (almost like error bars) to indicate the uncertainty, but that could be a bit messy.

Suggestions welcome.

 

[Grog6]

All sims have problems somewhere, in areas the model doesn't cover. In electronics, sims suck below 1nS, or over 10M ohms. But, I've used sims to make circuits below 10pS, and 100Gohm; It can be done, if you are aware of the limitations.

 

[BigMacDaddy]

At this point we're pretty clear that this is a direct result of the way the Barrowman equations work. Rocksim yields similar results.

What is not clear is how we can improve the way we handle this situation, in part because we don't really know what the most *correct* answer is.

We could put in a warning when you have inline fins, that CP may be optimistic. Or we could sim gapped fins as a single complex freeform fin. But without knowing what results are closest to reality, it's a bit dicey to code in heuristics like this.

I'm almost wondering if, in cases such as this, we should specify a range of possible CP values (almost like error bars) to indicate the uncertainty, but that could be a bit messy.

Suggestions welcome.

Sorry, don't have suggestions but did extend my example / experiment to do the free-form fin as you suggested and added result above.

 

[Nateairman]

At this point we're pretty clear that this is a direct result of the way the Barrowman equations work. Rocksim yields similar results.

What is not clear is how we can improve the way we handle this situation, in part because we don't really know what the most *correct* answer is.

We could put in a warning when you have inline fins, that CP may be optimistic. Or we could sim gapped fins as a single complex freeform fin. But without knowing what results are closest to reality, it's a bit dicey to code in heuristics like this.

I'm almost wondering if, in cases such as this, we should specify a range of possible CP values (almost like error bars) to indicate the uncertainty, but that could be a bit messy.

Suggestions welcome.

Neil,


Many rockets have inline fins that are not as concerning, for example the NikeX by estes. The warning may be best when the fins are in line and less that 1 caliber apart.

As for testing the true CP it seems that some flights may be the best way.

 

[KenECoyote]

At this point we're pretty clear that this is a direct result of the way the Barrowman equations work. Rocksim yields similar results.

I actually tested this in Rocksim myself this morning as well. I used CHAD test by taking a standard Madcow 4" Patriot file and lengthening the rear fins so there wasn't any gap and here are my results (original with gap followed by no gap):


No suggestions on a fix other than testing on actual rockets with both fin versions and recording the results.

 

[neil_w]

I actually tested this in Rocksim myself this morning as well. I used CHAD test by taking a standard Madcow 4" Patriot file and lengthening the rear fins so there wasn't any gap and here are my results (original with gap followed by no gap):
View attachment 557599

No suggestions on a fix other than testing on actual rockets with both fin versions and recording the results.

Do you have Rocksim set to Barrowman or something else?

 

[KenECoyote]

Do you have Rocksim set to Barrowman or something else?

Rocksim; however changing it to Barrowman didn't affect the CP.

Note: I don't normally change or touch that, so maybe someone else can also verify?

 

[Nateairman]

I actually tested this in Rocksim myself this morning as well. I used CHAD test by taking a standard Madcow 4" Patriot file and lengthening the rear fins so there wasn't any gap and here are my results (original with gap followed by no gap):
View attachment 557599

No suggestions on a fix other than testing on actual rockets with both fin versions and recording the results.

Ken,

Lengthening a fin to close the gap has an effect, however, making a single fin the size of the gapped fins has the most notable difference. I did experiment with making the gap smaller and smaller to overlapping and the result was not the same as a single fin.

 

[neil_w]

Ken,

Lengthening a fin to close the gap has an effect, however, making a single fin the size of the gapped fins has the most notable difference. I did experiment with making the gap smaller and smaller to overlapping and the result was not the same as a single fin.

Thanks, I overlooked that.

The proper comparison is between a single fin, and two gapped fins with the same outer profile.

 

[KenECoyote]

Agreed, as noted I used CHAD designing.

 

[neil_w]

Agreed, as noted I used CHAD designing.

NO ONE INVITED CHAD INTO THIS DISCUSSION

 

[KenECoyote]

(Ignore... I missed something already noted.)

 

[KenECoyote]

NO ONE INVITED CHAD INTO THIS DISCUSSION

 

[jqavins]

Suggestions welcome.

A few people around here, I don't remember who, use AeroCFD. and, of course, there are other CFD programs that someone might have access to. My suggestion is to request that someone determine the (mostly) true CP.

Alternatively, knock together a quick and easy 3/4FNC with split fins, and do multiple swing tests with varying nose weights to determine the true CP. Or, for that matter, someone here might have access to an actual wind tunnel.

Then, for OR, use whichever method is demonstrated to give the better answer AND add a warning.

 

[bjphoenix]

How did you make your fins in the sim file?

I have 2 separate freeform fin sets.
When I've finished with painting and added the recovery system I can measure the actual CG and compare with the calculated CP. I'll probably go ahead and swing test too just for grins.