A better way to test stability of small rockets.

[beeblebrox]

All that stuff...I have been pretty good at eyeballing a rocket and figuring out if it will fly....It isn't that complicated especially for a standard fins, one tube, and a nosecone. When you get into strange shapes and designs like multiple tube fins or uber cool looking spaceships...they can be difficult or impossible to model with openrocket. (Example, openrocket does not even accurately allow you to place fins on a boattail...)

 

[Alan15578]

You know, I was just trying to help. I thought I had a good idea. Obviously I was wrong, at least at the height I suggested. My intention wasn't to tick off every model rocket builder that uses Open Rocket or some other program.
I never proposed driving around in my pickup, but that might actually work. I'd probably have to mount it to my roof rack. I'd have to get my cousin, Billy Bob to hang out the window and take video. Or my wife, Billy Joe, who is also my cousin. I've seen tents tested this way and, I believe, Myth Busters has done stuff like this.
I really am thinking about building a wind tunnel. Have been for weeks. Not sure what to use as a gimbal. Should probably be on a stem, so as not to mess up the airflow too much. I could mount the rocket between two strings, one above and one below.

Some rocketeer built a ram air wind tunnel that is strapped to a car and was run down the road. It did not work well and is hardly worth mentioning, except to say that it has been done before.

 

[Aeronerd]

[Edit] Removed comment.

 

[Steve Shannon]

Some rocketeer built a ram air wind tunnel that is strapped to a car and was run down the road. It did not work well and is hardly worth mentioning, except to say that it has been done before.

Scaled Composites tested models by simply attaching it to a jig atop a vehicle. No tunnel needed. It’s not a stupid idea.

 

[gldknght]

All that stuff...I have been pretty good at eyeballing a rocket and figuring out if it will fly....It isn't that complicated especially for a standard fins, one tube, and a nosecone. When you get into strange shapes and designs like multiple tube fins or uber cool looking spaceships...they can be difficult or impossible to model with openrocket. (Example, openrocket does not even accurately allow you to place fins on a boattail...)

Rocksim does.

 

[GlenP]

I don’t recall exactly now, but I thought the Orbital Sciences Pegasus rocket was designed mostly by computer analysis and no wind tunnel testing before its initial flight. I can’t find info on line about it now, but I just vaguely recall the Pegasus being touted as the poster child for CFD state of the art at the time. Unmanned small sat launcher, but still an accomplishment. Now they are part of Northrop Grumman I guess. Sorry I don’t have a specific link to a story to verify this story from my failing memory of the details.

 

[Ronz Rocketz]

I’m curious how much weight did you add to the nose cone for the string test? You said you added BBs so sounds like a lot.

 

[Kelly]

I might also point out that your computer programs predicted my arrow vanes wouldn't work as rocket fins, but, in the spin test they performed extremely well.

I'm interested in looking at this. Do you have an .orc or .rok file for the arrow vane rocket that you tested?

 

[Ronz Rocketz]

I'm interested in looking at this. Do you have an .orc or .rok file for the arrow vane rocket that you tested?

I tried creating one but no response I could see on its accuracy. Had to add a lot of weight to the cone to get the CG an inch in front of the CP.

https://www.rocketryforum.com/threads/another-arrow-vane-fin-design.159917/post-2009690

 

[Gunstar]

Perhaps throwing it like a dart at a target that won't harm it would be better.

 

[Jeff Lassahn]

Does anyone have any hard data about how accurate Barrowman CP calculations are under various circumstances? I hear anecdotes about rockets that Barrowman claims are stable that aren't, and anecdotes about everything working great. But I've never seen a description of a systematic comparison against measured data.

 

[Nytrunner]

Does anyone have any hard data about how accurate Barrowman CP calculations are under various circumstances? I hear anecdotes about rockets that Barrowman claims are stable that aren't, and anecdotes about everything working great. But I've never seen a description of a systematic comparison against measured data.

This hobby has a lot of anecdata floating around. Did the failed rockets' design comply with the assumptions under which Barrowman is valid?

 

[Jeff Lassahn]

For that matter, what _are_ the assumptions under which Barrowman is valid?
It breaks completely for some geometries, so there are definitely some nontrivial assumptions baked in. I've never seen a description of what those are.

 

[Ronz Rocketz]

 

[Jeff Lassahn]

That's useful information, thanks.
Is this list distilled from a more detailed reference, like a research paper? It would be interesting to see how these affect the calculation.

At first reading I think there's a couple of problems with this list as stated:
1. it doesn't say anything about the back end of the rocket. There's almost certainly an assumption analogous to #4 about what's supposed to happen at the rear stagnation point. And the calculation gives very different answers if you put a cone on the tail and take the limit towards zero length than it does if you just leave the tail empty, so something interesting happens there.

3. I don't think it's exactly true that there's no vortices. It must be modeling the fins as generating lift, which means there's got to be tip vortices. My _guess_ is that it models the fins as lifting vortex generators, but the body as vortex free with zero lift.

Most of the weird behavior I've seen in playing around with the calculation seems to come from a combination of bad things happening at converging boat tails and bad things happening because the body/nose is modeled as generating torque but not lift.

 

[Ronz Rocketz]

Apparently, this was Barrowman’s presentation.
https://www.nar.org/wp-content/uploads/2016/07/NARCON-2017-Stability-Barrowman.pdf

 

[GlenP]

That's useful information, thanks.
Is this list distilled from a more detailed reference, like a research paper? It would be interesting to see how these affect the calculation.

At first reading I think there's a couple of problems with this list as stated:
1. it doesn't say anything about the back end of the rocket. There's almost certainly an assumption analogous to #4 about what's supposed to happen at the rear stagnation point. And the calculation gives very different answers if you put a cone on the tail and take the limit towards zero length than it does if you just leave the tail empty, so something interesting happens there.

3. I don't think it's exactly true that there's no vortices. It must be modeling the fins as generating lift, which means there's got to be tip vortices. My _guess_ is that it models the fins as lifting vortex generators, but the body as vortex free with zero lift.

Most of the weird behavior I've seen in playing around with the calculation seems to come from a combination of bad things happening at converging boat tails and bad things happening because the body/nose is modeled as generating torque but not lift.

Potential flow is a mathematical term that refers to flow that can be described by the potential flow equation, ie a velocity potential field. Lift can be represented by a vortex element, a mathematical solution to the potential flow equation. Most smooth flows that follow a body have fairly regular streamlines, another mathematical derived shape computed from the velocity potential function. The vortices they refer to here are more likely those in a turbulent and random wake that varies where you have lots of mixed and rotational flow, where the potential flow equation is no longer valid. Vortex lift is a part of a potential flow model, but vortices in a turbulent wake are not.

Your guess, item 3 above, is one correct way to make a potential flow model.

 

[Jeff Lassahn]

The mathematician part of me gets all twitchy when you say things like that. Pedantically, a potential flow cannot have nonzero vorticity anywhere and cannot generate lift at all. There's some ways of hacking around that that work in certain circumstances, which is why you see vortex elements included in "potential" problems, but there's always a messy singularity somewhere.

Neither here nor there for the original topic of the discussion, though.

I've been chasing references this morning, and found the original Barrowman paper,
as well as some other stuff including this on body lift.


Two practical takeaways about interpreting stability calculations:
1. Barrowmans' method assumes lift from the rocket body is exactly zero. This is not a good assumption in a lot of practical cases, and tends to put the CP too far back.
2. Barrowmans' method over-emphasizes the strength of rear-facing transitions. If you've got one of those, like a boat-tail or a payload transition, be a little skeptical about what the computation is telling you. Fortunately, it tends to put the CP too far forward in these cases, so this one is less dangerous than #1.

 

[lakeroadster]

Some rocketeer built a ram air wind tunnel that is strapped to a car and was run down the road. It did not work well and is hardly worth mentioning, except to say that it has been done before.

Scaled Composites tested models by simply attaching it to a jig atop a vehicle. No tunnel needed. It’s not a stupid idea.

The biggest issue is getting the rocket up and out of the dirty air created by the vehicle. For a model rocket you could use some sort of a pole... but getting the rocket hoisted up through the dirty air would likely destroy it.

Or just buy this... XCOR Wind Tunnel Truck

 

[Ronz Rocketz]

.
2. Barrowmans' method over-emphasizes the strength of rear-facing transitions. If you've got one of those, like a boat-tail or a payload transition, be a little skeptical about what the computation is telling you. Fortunately, it tends to put the CP too far forward in these cases, so this one is less dangerous than #1.

I think Tim talks about this.
https://apogeerockets.com/education/downloads/Newsletter238.pdf

 

[n27sb]

I think Tim talks about this.
https://apogeerockets.com/education/downloads/Newsletter238.pdf

Thank you for this reference. I read through it and it somewhat explains what I am running into.

 

[GlenP]

The mathematician part of me gets all twitchy when you say things like that. Pedantically, a potential flow cannot have nonzero vorticity anywhere and cannot generate lift at all. There's some ways of hacking around that that work in certain circumstances, which is why you see vortex elements included in "potential" problems, but there's always a messy singularity somewhere.

Right, I was just suggesting that you what you said earlier, was actually not that far off, you are on the right track, but yes this subject is part of a college aerodynamics class. But the terms vorticity and vortex are not interchangeable in this context.

An irrotational vortex element is one of the exact solutions to the potential flow equation. Since the potential flow equation is linear, you can represent a more complex solution by a linear combination of elemental solutions with different strengths. The thin airfoil theory and the lifting line theory are examples of how vortex elements are used in a potential flow linear model to represent the lift distribution along the chord of a 2D airfoil or along the quarter-chord span of a 3D finite wing. The messiness is dealt with by imposing the mathematical boundary conditions that make the most sense in the interpretation of the physical model, where the flow is tangent to the airfoil or wing surface, and the "jump" condition at the trailing edge. Messy, yes, but not intractable.

https://en.wikipedia.org/wiki/Airfoil#Thin_airfoil_theorythis linear theory gives the 2D lift curve slope of 2Pi, which is pretty accurate for low angles of attack.

https://en.wikipedia.org/wiki/Lifting-line_theorywhich is the simplified theory that suggests the most efficient lift distribution, for minimum induced drag, is the elliptical distribution.
https://en.wikipedia.org/wiki/Lifting-line_theory#Elliptical_wings
sorry for the digression.

 

[Buckeye]

The biggest issue is getting the rocket up and out of the dirty air created by the vehicle. For a model rocket you could use some sort of a pole... but getting the rocket hoisted up through the dirty air would likely destroy it.

Or just buy this... XCOR Wind Tunnel Truck

A modern pickup truck is pretty aerodynamic, with the separated airflow confined to the open bed under a strong shear layer coming off the cab. The air is not very "dirty" above the vehicle. All that XCOR streamlining isn't really necessary if they are gonna mount the test board that high above the roof. A CFD simulation prior to building this contraption would have revealed that and saved them a lot of work. Oh, wait...

 

[lakeroadster]

A modern pickup truck is pretty aerodynamic, with the separated airflow confined to the open bed under a strong shear layer coming off the cab. The air is not very "dirty" above the vehicle. All that XCOR streamlining isn't really necessary if they are gonna mount the test board that high above the roof. A CFD simulation prior to building this contraption would have revealed that and saved them a lot of work. Oh, wait...

The goal was to clean up the airflow towards the rear of the truck, which the aero package does, and does very well.

Thus when they mount the prototype above the pickup the airflow will be as close to pristine as possible.

https://jalopnik.com/the-ford-f-250-trunnel-is-for-those-who-want-to-do-aero-1842492987