OpenRocket Cnalpha accuracy

[cragbot]

I've been reading through the 2009 paper and noticed that the simulated Cnalpha quantities are pretty far off of the wind tunnel data referenced. Does anyone know if that part of the aerodynamic model has been improved? If it hasn't, anyone have ideas for how to be confident in the dynamic stability of a rocket short of trying to tie OR into a CFD model or something? It looks like RasAero has a much better Cnalpha prediction, but we haven't been able to find as much detailed information about how the rest of the RasAero simulation operates. My team's funding is somewhat dependent on being able to describe exactly how our simulation works.

Thanks all,
Craig

 

[neil_w]

Forwarded to OR team.

 

[thzero]

I've been reading through the 2009 paper and noticed that the simulated Cnalpha quantities are pretty far off of the wind tunnel data referenced. Does anyone know if that part of the aerodynamic model has been improved? If it hasn't, anyone have ideas for how to be confident in the dynamic stability of a rocket short of trying to tie OR into a CFD model or something? It looks like RasAero has a much better Cnalpha prediction, but we haven't been able to find as much detailed information about how the rest of the RasAero simulation operates. My team's funding is somewhat dependent on being able to describe exactly how our simulation works.

Thanks all,
Craig

Read the code? It's open source.

 

[UhClem]

If it hasn't, anyone have ideas for how to be confident in the dynamic stability of a rocket short of trying to tie OR into a CFD model or something?

Are you really interested in dynamic stability? That is quite a different animal from the static stability you get from the CP-CG relationship.

 

[H. Craig Miller]

The primary limitation stated in the paper still exists:

"The design was replicated in OpenRocket as closely as possible, given the current limitations of the software. The most notable difference is that an airfoil profile was selected for the fins instead of the double-wedge that is not supported by OpenRocket.”

Other limitations may or may not have been resolved:

“However, a few aspects of the rocket break the assumptions made in the computation methods. First, the boattail at the end of the rocket reduces the drag by guiding the air into the void left behind it, while the simulation software only takes into account the reduction of base area. Second, the airfoil shape of the fins affects the drag characteristic especially in the transonic region, where it produces the slight reduction peak. Finally, at higher supersonic speeds the simulation produces less reliable results as expected, producing a too high drag coefficient.”

Even so, the author believed that:

“Overall, however, the drag coefficient matches the experimental results with reasonable accuracy, and the results of actual test flights... give credence to the drag coefficient estimation.”

Forwarded to OR Team for review.

 

[cragbot]

Are you really interested in dynamic stability? That is quite a different animal from the static stability you get from the CP-CG relationship.

Yes, you can determine a damping ratio using OR if you use some custom expressions and a simulation listener. However this is very dependent on Cnalpha, so if that's significantly wrong then we can't say much about the dynamic stability of the vehicle before flight testing. Just relying on static margin is too wishy washy for our team and importantly our department head who is funding us.

-Craig

 

[cragbot]

The primary limitation stated in the paper still exists:

"The design was replicated in OpenRocket as closely as possible, given the current limitations of the software. The most notable difference is that an airfoil profile was selected for the fins instead of the double-wedge that is not supported by OpenRocket.”

Other limitations may or may not have been resolved:

“However, a few aspects of the rocket break the assumptions made in the computation methods. First, the boattail at the end of the rocket reduces the drag by guiding the air into the void left behind it, while the simulation software only takes into account the reduction of base area. Second, the airfoil shape of the fins affects the drag characteristic especially in the transonic region, where it produces the slight reduction peak. Finally, at higher supersonic speeds the simulation produces less reliable results as expected, producing a too high drag coefficient.”

Even so, the author believed that:

“Overall, however, the drag coefficient matches the experimental results with reasonable accuracy, and the results of actual test flights... give credence to the drag coefficient estimation.”

Forwarded to OR Team for review.

We're happy enough with the accuracy of the drag coefficient based on the data from the paper for our velocity regime. The Cnalpha results are what I'm most worried about.

Thanks for taking the time,
Craig

 

[Jeff Lassahn]

Which paper are you referencing?

 

[cragbot]

Which paper are you referencing?

Both of the papers here show the same discrepancy in simulated normal force coefficient vs. wind tunnel data. They're both rather old so if this has been addressed in later releases it'll save my team some work. https://openrocket.info/documentation.html

 

[UhClem]

Why are you concerned about dynamic stability? This is rarely a problem with rockets of normal configuration.

 

[cragbot]

Why are you concerned about dynamic stability? This is rarely a problem with rockets of normal configuration.

For hobby purposes it's not a problem for the vast majority of flights I agree. As an engineering student though it's not really enough to just say "static margin looks good should be alright" especially if our team does more extreme flights. The Bare Necessities project is an extreme example. Our team needs to be confident in the dynamic stability so we know that the vehicle responds safely to a typical disturbance. Static margin really doesn't tell you that.

 

[rharshberger]

Why are you concerned about dynamic stability? This is rarely a problem with rockets of normal configuration.

Sounds like the individual is a college student, who's sponsor/prof thinks they need that data for what is probably a competition rocket.

 

[rharshberger]

For hobby purposes it's not a problem for the vast majority of flights I agree. As an engineering student though it's not really enough to just say "static margin looks good should be alright" especially if our team does more extreme flights. The Bare Necessities project is an extreme example. Our team needs to be confident in the dynamic stability so we know that the vehicle responds safely to a typical disturbance. Static margin really doesn't tell you that.

Welcome to free open-source software, the codes available so you can tweak it to your hearts desire, the target audience for OR is not a bunch of engineering students who should have access to much more powerful tools and programming ability as part of their project. Which college?

 

[UhClem]

Sounds like the individual is a college student, who's sponsor/prof thinks they need that data for what is probably a competition rocket.

Actually it sounds more like they don't really understand dynamic stability. Hopefully they have access to a copy of Topics in Advanced Model Rocketry.

I have had only two rockets exhibit dynamic stability issues. Which vanished with higher thrust motors.

 

[cragbot]

Welcome to free open-source software, the codes available so you can tweak it to your hearts desire, the target audience for OR is not a bunch of engineering students who should have access to much more powerful tools and programming ability as part of their project. Which college?

Florida institute of technology. I'm pretty confident RasAero does what we need, but we're also trying to avoid working with a black box simulator as much as possible so OpenRocket seems like a good place to start, and we can modify it if we need to.

 

[rharshberger]

Florida institute of technology. I'm pretty confident RasAero does what we need, but we're also trying to avoid working with a black box simulator as much as possible so OpenRocket seems like a good place to start, and we can modify it if we need to.

Enjoy the project and if you get an improvement share it with the OR team.

 

[cragbot]

Actually it sounds more like they don't really understand dynamic stability. Hopefully they have access to a copy of Topics in Advanced Model Rocketry.

I have had only two rockets exhibit dynamic stability issues. Which vanished with higher thrust motors.

I'm only just now taking control systems so I'm still learning about this. I have read through the apogee newsletters on the topic though, and the normal force coefficient is hugely important so if it's wrong in the simulation how can we confidently say anything about the dynamic stability of the rocket?

 

[user 30299]

I'm only just now taking control systems so I'm still learning about this. I have read through the apogee newsletters on the topic though, and the normal force coefficient is hugely important so if it's wrong in the simulation how can we confidently say anything about the dynamic stability of the rocket?

Can you post your OR file so that we can review it ?

 

[cragbot]

Can you post your OR file so that we can review it ?

I'll ask my team, probably post it tomorrow if they're alright with it I don't see why they wouldn't be

 

[user 30299]

I'm only just now taking control systems so I'm still learning about this. I have read through the apogee newsletters on the topic though, and the normal force coefficient is hugely important so if it's wrong in the simulation how can we confidently say anything about the dynamic stability of the rocket?

Are you being this thorough with Fin Flutter? You have a greater chance of shredding your fins then having a rocket go unstable.

 

[OzHybrid]

Forwarded to OR team.

Would a quick easy solution be to enable OR to export a model that could be imported into RAS Aero

 

[cragbot]

Are you being this thorough with Fin Flutter? You have a greater chance of shredding your fins then having a rocket go unstable.

For this project they're quite a bit thicker than finsim suggests they need to be, we're more worried about them breaking on landing than flutter. Our department head also recognizes that flutter analysis is above our paygrade so he's ok with us using finsim for that.

 

[cragbot]

Would a quick easy solution be to enable OR to export a model that could be imported into RAS Aero

That may be helpful for other users. In our case it was easy enough to re-create the rocket in RasAero, the problem is that RasAero isn't as open so we can't say to our department head that we know exactly how it runs. I have no reason to doubt RasAero's output, and the comparisons to wind tunnel data for RasAero look better but that's not quite enough.

 

[user 30299]

For this project they're quite a bit thicker than finsim suggests they need to be, we're more worried about them breaking on landing than flutter.

How did you determine what thickness to use? (Engineering method or someone's years of experience) Or could you have improved the recovery system to reduce fin breakage?

 

[OzHybrid]

That may be helpful for other users. In our case it was easy enough to re-create the rocket in RasAero, the problem is that RasAero isn't as open so we can't say to our department head that we know exactly how it runs. I have no reason to doubt RasAero's output, and the comparisons to wind tunnel data for RasAero look better but that's not quite enough.

Historically Open Rocket has been good enough for sub sonic amateur rocketry. Supersonic fliers have always used RAS Aero as an add on to the initial design in Open Rocket. Perhaps getting your top brains in the team to participate in the OR development team could be a way forward. As your team and HOD need to understand how the analysis is done , it seems reasonable that you would understand the analysis and be able to contribute. I think it's unlikely that RAS Aero will go open source as the larger version forms a commercial product.
The new version of OR needs to be released soonish and it's unlikely to include all that you want/need for this next release. Maybe the next one with your teams assistance.
Norm

 

[UhClem]

I'm only just now taking control systems so I'm still learning about this.

You should have more than enough math background by now to follow the differential equations in Topics... Out of print but a good college library out to have a copy or be able to find one via an inter-library loan.

I should note that I have been unable to replicate the dynamic stability issues of my rockets in Rocksim or Openrocket. Although Rocksim does show that the damping ratio is around the 0.05 lower limit proposed in Topics.

 

[cragbot]

How did you determine what thickness to use? (Engineering method or someone's years of experience) Or could you have improved the recovery system to reduce fin breakage?

We did an impact force estimation and applied a 1.5 safety factor. Came out thicker than kits similarly sized to our rocket so we're taking that to mean we have margin on that front.

 

[OzHybrid]

Is this using grid fins as an air brake? Are you going for the exact altitude comp? Is that why there is a need for dynamic stability analysis?
Norm

 

[cragbot]

Is this using grid fins as an air brake? Are you going for the exact altitude comp? Is that why there is a need for dynamic stability analysis?
Norm

For this year it's just about making sure we really understand the flight dynamics. We'll prove that we do with our flights this year and add complexity next year.

 

[JoePfeiffer]

Would a quick easy solution be to enable OR to export a model that could be imported into RAS Aero

It's something I'd certainly like to see (see OR issue #875 ). Is the RASaero file format documented?