RasAero II Apogee Discrepancy with OpenRocket

The Rocketry Forum

Help Support The Rocketry Forum:

Joined
Dec 19, 2021
Messages
14
Reaction score
2
I wanted to verify our apogee (and the rest of the flight) predictions between OpenRocket and RaeAseo II. After importing the rocket into RasAero II and making sure that everything matched the OpenRocket, I ran the simulation and found that the apogee varied between OpenRocket and RasAero II by around 1,000 ft, roughly 10% of the total apogee. Hopefully, I missed something in the setup which has caused the discrepancy but I have been unable to find it.

The rocket weighs about 48lb without motors and will be launching on an M1450 (probably). Peak velocity is currently only around M .75.

I have attached both files in the form of a zip. If someone could take a look and help find the source of the error that would be extremely helpful.
 

Attachments

  • Comp_Rocket.zip
    120.6 KB · Views: 8

mikec

Well-Known Member
Joined
May 9, 2009
Messages
2,664
Reaction score
481
Without looking at your files, how did you set the roughness and turbulence properties? Try tweaking those. A 10% difference is pretty typical.
 

JoePfeiffer

Well-Known Member
Joined
Jun 15, 2013
Messages
329
Reaction score
295
10% is doing really well. I see more variation than that from flight to flight of a real rocket.
 
Joined
Dec 19, 2021
Messages
14
Reaction score
2
Without looking at your files, how did you set the roughness and turbulence properties? Try tweaking those. A 10% difference is pretty typical.

I looked at the roughness properties and they should be the same between both. I get that 10% is reasonable, but I went digging a bit and from what I can tell RasAero thinks that the cd is much lower than that of the OpenRocket simulation.
1642987742994.png
1642987749687.png


Corresponding drag:

1642987767327.png

1642987773909.png


It was a bit of a pain to get the scales to line up well, but I think the differnce is pretty clear. Any reason why the cd values would be different between the two programs?
 

dhbarr

Amateur Professional
Joined
Jan 30, 2016
Messages
8,081
Reaction score
2,506
I looked at the roughness properties and they should be the same between both. I get that 10% is reasonable, but I went digging a bit and from what I can tell RasAero thinks that the cd is much lower than that of the OpenRocket simulation.
View attachment 501314 View attachment 501315

Corresponding drag:

View attachment 501316
View attachment 501317

It was a bit of a pain to get the scales to line up well, but I think the differnce is pretty clear. Any reason why the cd values would be different between the two programs?
Try different atmospheric models in OR ?
 
Joined
Dec 19, 2021
Messages
14
Reaction score
2
Try different atmospheric models in OR ?

Well, it appears that turning on the "all turbulent" setting has brought the apogees closer, now only about 400ft. About 4% is a difference I am comfortable with.

The reason this matters a lot is that this is for the SpaceportAmerica cup, with an alt. goal of 10,000 ft. Our team is running an airbrakes module, so the plan is to overshoot to around to just above 10K and use the module to slow it down during the coast phase to reach 10K. If we undershoot, we're screwed. Since an M1450 is a 99% M we really need to determine if we have to bump up to a 5 grain N motor, which means a lot needs to change. There is likely weight unaccounted for (paint, recovery hardware, little bits of drag from key switches and an external camera, etc.) so if we are looking for a large margin of error, hence the 11K. The airbrakes are sized to reduce the altitude from 11,000 to 9,500 plus a 1.5x factor of safety on their area in case our cd estimates are inaccurate. Given this, I am tempted to move to an N. However, if we can stay on the M1450 that would be ideal because it's an extra $150 per launch and the casing is more expensive.
 

dhbarr

Amateur Professional
Joined
Jan 30, 2016
Messages
8,081
Reaction score
2,506
Well, it appears that turning on the "all turbulent" setting has brought the apogees closer, now only about 400ft. About 4% is a difference I am comfortable with.

The reason this matters a lot is that this is for the SpaceportAmerica cup, with an alt. goal of 10,000 ft. Our team is running an airbrakes module, so the plan is to overshoot to around to just above 10K and use the module to slow it down during the coast phase to reach 10K. If we undershoot, we're screwed. Since an M1450 is a 99% M we really need to determine if we have to bump up to a 5 grain N motor, which means a lot needs to change. There is likely weight unaccounted for (paint, recovery hardware, little bits of drag from key switches and an external camera, etc.) so if we are looking for a large margin of error, hence the 11K. The airbrakes are sized to reduce the altitude from 11,000 to 9,500 plus a 1.5x factor of safety on their area in case our cd estimates are inaccurate. Given this, I am tempted to move to an N. However, if we can stay on the M1450 that would be ideal because it's an extra $150 per launch and the casing is more expensive.
M1939W, or are you married to CTI? An inch more of case, 500 more N, $25 cost increase?
 
Joined
Dec 19, 2021
Messages
14
Reaction score
2
M1939W, or are you married to CTI? An inch more of case, 500 more N, $25 cost increase?

The club definitely prefers CTI, but we can always out the M1939W in a CTI case. 1939 brings us to 10,900, which is still a bit short (I know this is nitpicky, but again I am assuming this is an over-prediction because of expected increase in weight and drag).
 

dhbarr

Amateur Professional
Joined
Jan 30, 2016
Messages
8,081
Reaction score
2,506
The club definitely prefers CTI, but we can always out the M1939W in a CTI case. 1939 brings us to 10,900, which is still a bit short (I know this is nitpicky, but again I am assuming this is an over-prediction because of expected increase in weight and drag).
Yup, that ought to work. Do you have any slimming on the aft end? Should be able to pad your comfort margin fairly easily.

 
Joined
Jun 30, 2015
Messages
1,485
Reaction score
1,703
Location
Australia
Thanks. Does OpenRocket take that geometry into account and mess with the drag calcs? I had asked somen about this a while back and I remember them saying that OpenRocet only does drag calculations for the external airframe +fins.
Probably better to design it so they're all inside and not have to ask that question. Fix it, resim and let us know.
 

Chuck Rogers

Well-Known Member
Joined
Jun 7, 2011
Messages
202
Reaction score
134
Some things to consider:

1) RASAero II varies thrust with altitude, OpenRocket does not. It's important to enter the nozzle exit diameter input into RASAero II, it is not only used for the Power-On CD, but it is also used to vary thrust with altitude. With the Spaceport America launch site at 4,000 ft above Sea Level, this can be a significant effect. Again, OpenRocket does not vary thrust with altitude. RASAero II assumes the rasp.eng thrust curve is a sea level thrust curve, and varies thrust with altitude using the nozzle exit area (calculated from the nozzle exit diameter). So what looks like an overshoot in the RASAero II altitude prediction is actually a more accurate altitude prediction.

2) Weight is going to be really, really important. When I compare predicted altitudes to actual altitudes, many (most) discrepancies turn out to be due to weight errors. You need to really carefully estimate and track the lift-off weight of the rocket. If the weight gets too high, you will undershoot on the 10,000 ft altitude.

3) The atmosphere model in RASAero II is likely calculating a more accurate Reynolds Number than OpenRocket. This could cause different friction coefficients for a given Mach number.

4) Adjusting the surface roughness in RASAero II will give you a more accurate altitude prediction. But if you want to be sure to not undershoot the 10,000 ft altitude, you should use the most optimistic surface roughness, which is Zero Roughness (Aerodynamically Smooth).

5) You can test your airbrake module by running RASAero II with the nominal rocket configuration, and get the predicted altitude, which will be above 10,000 ft. You can then increase the size of the Rail Guides or Launch Shoes, and see the change in the predicted altitude of the rocket for a given change in the rocket CD.

6) The RASAero II and OpenRocket predicted altitudes started 10% apart. I believe the RASAero II prediction is more accurate, see 1) above. But right off the bat, it would seem reasonable to be able to tolerate a +/-10% change in altitude with your drag modulation device. It would also seem prudent to have the nominal (zero deflection on the drag device) prediction be at least 10% over 10,000 ft.

7) I'm sure whatever drag producing device you are using can be found in Hoerner, Fluid Dynamic Drag. So take a look at your drag device with zero deflection (zero change in CD), and with max deflection the change in the total rocket CD, and compare to the values that you used in the RASAero II altitude change with rocket total CD study in 5) above.

8) If you have a RASAero II predicted altitude 20% over 10,000 ft, and you can reduce the altitude 20% by using only one-half of the deflection of your drag device, based on the accuracy (and stability, phase and gain margin) of your controller, you should do very well on hitting very close to 10,000 ft.


Charles E. (Chuck) Rogers
Rogers Aeroscience
 
Last edited:
Joined
Dec 19, 2021
Messages
14
Reaction score
2
Yup, that ought to work. Do you have any slimming on the aft end? Should be able to pad your comfort margin fairly easily.


What do you mean by slimming? If anything, it'll get a tad wider because our tip-to-tip (yes, I know unnecessary at this scale but it's what the team likes to do. Plan is 2 or 3 layers of 4oz and 1 layer of 2oz) will increase diameter a bit.
 
Joined
Dec 19, 2021
Messages
14
Reaction score
2
Some things to consider:

1) RASAero II varies thrust with altitude, Open Rocket does not. It's important to enter the nozzle exit diameter input into RASAero II, it is not only used for the Power-On CD, but it is also used to vary thrust with altitude. With the Spaceport America launch site at 4,000 ft above Sea Level, this can be a significant effect. Again, Open Rocket does not vary thrust with altitude. RASAero II assumes the rasp.eng thrust curve is a sea level thrust curve, and varies thrust with altitude using the nozzle exit area (calculated from the nozzle exit diameter). So what looks like an overshoot in the RASAero II altitude prediction is actually a more accurate altitude prediction.

2) Weight is going to be really, really important. When I compare predicted altitudes to actual altitudes, many (most) discrepancies turn out to be due to weight errors. You need to really carefully estimate and track the lift-off weight of the rocket. If the weight gets too high, you will undershoot on the 10,000 ft altitude.

3) The atmosphere model in RASAero II is likely calculating a more accurate Reynolds Number than Open Rocket. This could cause different friction coefficients for a given Mach number.

4) Adjusting the surface roughness in RASAero II will give you a more accurate altitude prediction. But if you want to be sure to not undershoot the 10,000 ft altitude, you should use the most optimistic surface roughness, which is Zero Roughness (Aerodynamically Smooth).

5) You can test your airbrake module by running RASAero II with the nominal rocket configuration, and get the predicted altitude, which will be above 10,000 ft. You can then increase the size of the Rail Guides or Launch Shoes, and see the change in the predicted altitude of the rocket for a given change in the rocket CD.

6) The RASAero II and Open Rocket predicted altitudes started 10% apart. I believe the RASAero II prediction is more accurate, see 1) above. But right off the bat, it would seem reasonable to be able to tolerate a +/-10% change in altitude with your drag modulation device. It would also seem prudent to have the nominal (zero deflection on the drag device) prediction be at least 10% over 10,000 ft.

7) I'm sure whatever drag producing device you are using can be found in Hoerner, Fluid Dynamic Drag. So take a look at your drag device with zero deflection (zero change in CD), and with max deflection the change in the total rocket CD, and compare to the values that you used in the RASAero II altitude change with rocket total CD study in 5) above.

8) If you have a RASAero II predicted altitude 20% over 10,000 ft, and you can reduce the altitude 20% by using only one-half of the deflection of your drag device, based on the accuracy (and stability, phase and gain margin) of your controller, you should do very well on hitting very close to 10,000 ft.


Charles E. (Chuck) Rogers
Rogers Aeroscience


Wow okay this has given me a lot to think about. Let me address a few things.

1) This is really intersting. I know I set the nozzle size correctly. Most of our launches have beenin Ohio (essentially sea level) so we have not had the opportunity to have an OR prediction be wildly off, they're usually pretty accurate.

2) When we complete construction, we always manually measure mass and CG to override the predictions so that we get accurate pre-launch flight path predictions. In the past (see point 1) many of our launches have been within 1% or 2% of our OR predictions, but again we have only launched from effectively sea level.

3) Neat.

4) I think I am a little confused; wouldn't zero roughness produce less drag and lead to a higher altitude prediction? If we want to make sure we do not undershoot shouldn't we assume that it's worse than what we hope, meaning more drag hence lower apo, so that if anything we overshoot and have our airbrake take care of it?

5) I'll definitely relay this information to our airbrakes people

6) In our design requirements we have 11,000 ft minimum natural apogee. Airbrakes are sized to reduce apogee from 11,000 ft to 9,500ft assuming they deploy to their full angle (more on this below) one second after motor burnout. Competition rules say that we cannot turn on airbrakes during motor burn, and I wouldn't want to anyway. In addition to this oversize. we are also rounding up to the nearest reasonable number (usually multiples of .25") just as a little extra FOS.

7) We are estimating our airbarkes to have a cd of around 1.2 based on this:
1642992641711.png


We will verify this experimentally and make changes if necessary.

8) The idea is as follows:
Shoot for an apo prediction a bit more than 10% over 10,000
Size airbrakes to reduce apo by about 15%

The question, I guess, comes to whether or not to trust OR or RasAero II. I understand your points in favor of RasAero and I like that they help explain some of the differneces that we are seeing. I do think, however, that trusting OpenRocket means that we are at worst under-estimating our apogee, in which case the extra can be accounted for by our airbrakes or, after our test flights, a piece of steel bolted to a bulkhead if absolutely necessary.

So, what it comes down to is that if we go this direction, then we need an OR that gets us to 11K without too much overshoot. Running on an N2200 I can get 11700, which is a little high, but I think that we can make heavy bulkheads or slightly larger airbrakes.

That was a lot, I hope it makes sense. If anything I said about our airbrakes is confusing, this page may help clear it up. Just scroll down to the airbrakes section for an explanation of how our system works.
 

dhbarr

Amateur Professional
Joined
Jan 30, 2016
Messages
8,081
Reaction score
2,506
What do you mean by slimming? If anything, it'll get a tad wider because our tip-to-tip (yes, I know unnecessary at this scale but it's what the team likes to do. Plan is 2 or 3 layers of 4oz and 1 layer of 2oz) will increase diameter a bit.
I only meant that you appear to have room for a tailcone on the backend to reduce base drag.

I was also surprised at the low height of your fins, but presumably they were selected at that span for a reason.
 
Joined
Dec 19, 2021
Messages
14
Reaction score
2
I only meant that you appear to have room for a tailcone on the backend to reduce base drag.

I was also surprised at the low height of your fins, but presumably they were selected at that span for a reason.

Would the drag benefits outweigh (haha) the added weight of the tailcone? The fins are small because there is an 8.8lb payload all the way at the top so the CG is far up.
 

Chuck Rogers

Well-Known Member
Joined
Jun 7, 2011
Messages
202
Reaction score
134
The question, I guess, comes to whether or not to trust OR or RasAero II. I understand your points in favor of RasAero and I like that they help explain some of the differneces that we are seeing. I do think, however, that trusting OpenRocket means that we are at worst under-estimating our apogee, in which case the extra can be accounted for by our airbrakes or, after our test flights, a piece of steel bolted to a bulkhead if absolutely necessary.

It's pretty clear that you've been tweaking OpenRocket to get good matches (1%-2% error), then you do just a few runs on RASAero II and say there is a 10% difference. Tweaking RASAero II will also get you to 1%-2% error.

Are you adjusting the surface roughness or launch angle in OpenRocket? Are you adjusting the launch angle to match a measured downrange distance? A tweak running OpenRocket to better match altitude data may be covering up a less accurate model. Perform the same process with RASAero II, you'll get down to 1%-2% error. From your posts above you've already eliminated most of the 10% difference by surface roughness and all turbulent flow settings, you have to dig deep to make sure you are really running a one-to-one comparison between RASAero II and OpenRocket.

I'm not making a criticism here. Adjusting your simulation model is a key part of the altitude prediction process. Jim Javis's Mach 3 rockets have been coming back from flight with less thermal damage than other rockets. For his rockets the surface roughness setting in RASAero II is different than for other rockets, Jim has calibrated it from prior flights. It makes his predictions for future flights more accurate.

I'd recommend tweaking RASAero II until you have the same 1%-2% match, because the lack of variation of thrust with altitude in OpenRocket is going to cause you problems as you move from near sea level to 4,000 ft at Spaceport America.


4) I think I am a little confused; wouldn't zero roughness produce less drag and lead to a higher altitude prediction? If we want to make sure we do not undershoot shouldn't we assume that it's worse than what we hope, meaning more drag hence lower apo, so that if anything we overshoot and have our airbrake take care of it?

There are two possible strategies. The strategy above is to have the most pessimistic altitude prediction, in which case I would use the rough camouflage paint surface roughness setting and all turbulent flow. This way you make sure you never undershoot.

The strategy I proposed is that you should plan to overshoot in altitude, and the real issue is whether your modulating drag producing device has enough authority (like control authority) to generate enough drag to overcome the most optimistic altitude prediction.

Remember, overshooting by too much, and your drag modulating device can't overcome the overshoot, is also a potential failure mode.

So those are your two possible strategies.


6) In our design requirements we have 11,000 ft minimum natural apogee. Airbrakes are sized to reduce apogee from 11,000 ft to 9,500ft assuming they deploy to their full angle (more on this below) one second after motor burnout. Competition rules say that we cannot turn on airbrakes during motor burn, and I wouldn't want to anyway. In addition to this oversize. we are also rounding up to the nearest reasonable number (usually multiples of .25") just as a little extra FOS.

7) We are estimating our airbarkes to have a cd of around 1.2 based on this:

We will verify this experimentally and make changes if necessary.

As you are doing, I strongly recommend you perform flight tests to verify your CD change as a function of how you are modulating the drag.


As a note, a fellow aerospace engineer co-worker of mine who went on to become a University professor, had a team in a contest like this where they were the first to use a real-time drag modulating device to hit a specific altitude. They won the contest. I won't provide details here, this is of course a competition. :) What really impressed me was that they put the rocket with the drag modulating device in a wind tunnel (!!!) with the drag modulating device at different settings for different wind tunnel runs. You'll do fine with flight tests, but I was really impressed that they put the rocket in a wind tunnel.


Charles E. (Chuck) Rogers
Rogers Aeroscience
 
Last edited:
Joined
Jun 30, 2015
Messages
1,485
Reaction score
1,703
Location
Australia
It's pretty clear that you've been tweaking Open Rocket to get good matches (!%-2% error), then you do just a few runs on RASAero II and say there is a 10% difference. Tweaking RASAero II will also get you to 1%-2% error.

Are you adjusting the surface roughness or launch angle in Open Rocket? Are you adjusting the launch angle to match a measured downrange distance? A tweak running Open Rocket to better match altitude data may be covering up a less accurate model. Perform the same process with RASAero II, you'll get down to 1%-2% error. From your posts above you've already eliminated most of the 10% difference by surface roughness and all turbulent flow settings, you have to dig deep to make sure you are really running a one-to-one comparison between RASAero II and Open Rocket.

I'm not making a criticism here. Adjusting your simulation model is a key part of the altitude prediction process. Jim Javis's Mach 3 rockets have been coming back from flight with less thermal damage than other rockets. For his rockets the surface roughness setting in RASAero II is different than for other rockets, Jim has calibrated it from prior flights. It makes his predictions for future flights more accurate.

I'd recommend tweaking RASAero II until you have the same 1%-2% match, because the lack of variation of thrust with altitude in Open Rocket is going to cause you problems as you move from near sea level to 4,000 ft at Spaceport America.




There are two possible strategies. The strategy above is to have the most pessimistic altitude prediction, in which case I would use the rough camouflage paint surface roughness setting and all turbulent flow. This way you make sure you never undershoot.

The strategy I proposed is that you should plan to overshoot in altitude, and the real issue is whether your modulating drag producing device has enough authority (like control authority) to generate enough drag to overcome the most optimistic altitude prediction.

Remember, overshooting by too much, and your drag modulating device can't overcome the overshoot, is also a potential failure mode.

So those are your two possible strategies.




As you are doing, I strongly recommend you perform flight tests to verify your CD change as a function of how you are modulating the drag.


As a note, a fellow aerospace engineer co-worker of mine who went on to become a University professor, had a team in a contest like this where they were the first to use a real-time drag modulating device to hit a specific altitude. They won the contest. I won't provide details here, this is of course a competition. :) What really impressed me was that they put the rocket with the drag modulating device in a wind tunnel (!!!) with the drag modulating device at different settings for different wind tunnel runs. You'll do fine with flight tests, but I was really impressed that they put the rocket in a wind tunnel.


Charles E. (Chuck) Rogers
Rogers Aeroscience
All I'll add to which to trust is Chucks modelling was recently updated to match real world testing results. Open Rocket has not had it's modelling updated for some time. I do my primary modelling in Open Rocket as it's easy to use. Final modelling is done in RAS Aero. As it is the most accurate in terms if simulation results.
Norm
 
Joined
Dec 19, 2021
Messages
14
Reaction score
2
It's pretty clear that you've been tweaking OpenRocket to get good matches (1%-2% error), then you do just a few runs on RASAero II and say there is a 10% difference. Tweaking RASAero II will also get you to 1%-2% error.

Are you adjusting the surface roughness or launch angle in OpenRocket? Are you adjusting the launch angle to match a measured downrange distance? A tweak running OpenRocket to better match altitude data may be covering up a less accurate model. Perform the same process with RASAero II, you'll get down to 1%-2% error. From your posts above you've already eliminated most of the 10% difference by surface roughness and all turbulent flow settings, you have to dig deep to make sure you are really running a one-to-one comparison between RASAero II and OpenRocket.

I'm not making a criticism here. Adjusting your simulation model is a key part of the altitude prediction process. Jim Javis's Mach 3 rockets have been coming back from flight with less thermal damage than other rockets. For his rockets the surface roughness setting in RASAero II is different than for other rockets, Jim has calibrated it from prior flights. It makes his predictions for future flights more accurate.

I'd recommend tweaking RASAero II until you have the same 1%-2% match, because the lack of variation of thrust with altitude in OpenRocket is going to cause you problems as you move from near sea level to 4,000 ft at Spaceport America.




There are two possible strategies. The strategy above is to have the most pessimistic altitude prediction, in which case I would use the rough camouflage paint surface roughness setting and all turbulent flow. This way you make sure you never undershoot.

The strategy I proposed is that you should plan to overshoot in altitude, and the real issue is whether your modulating drag producing device has enough authority (like control authority) to generate enough drag to overcome the most optimistic altitude prediction.

Remember, overshooting by too much, and your drag modulating device can't overcome the overshoot, is also a potential failure mode.

So those are your two possible strategies.




As you are doing, I strongly recommend you perform flight tests to verify your CD change as a function of how you are modulating the drag.


As a note, a fellow aerospace engineer co-worker of mine who went on to become a University professor, had a team in a contest like this where they were the first to use a real-time drag modulating device to hit a specific altitude. They won the contest. I won't provide details here, this is of course a competition. :) What really impressed me was that they put the rocket with the drag modulating device in a wind tunnel (!!!) with the drag modulating device at different settings for different wind tunnel runs. You'll do fine with flight tests, but I was really impressed that they put the rocket in a wind tunnel.


Charles E. (Chuck) Rogers
Rogers Aeroscience

I've got a lot to think about. Tomorrow the team is going to sit down and try to solve this problem, so hopefully, it'll all get sorted out. For now, it looks like we'll be able to get the M1450 at the right altitude.

For our drag-producing device, we have plans to model it and run CFD analysis numerous times, and we have a custom system running that can update what it believes that cd should be during flight to make. Plus, we can review that data after our test launches. Unfortunately, our wind tunnel is far too small to accommodate our rocket (or even just the lower portion) but we're working on finding others in the area.

I think the upper-bound failure condition is very interesting. It is not something we have considered before but of all the outcomes of a test flight that is the "least bad" of the potential failures. It'll be much easier to make modifications to reduce the apogee than it would be to increase it. Regardless, we will definitely keep an eye out for that and make sure we aren't maximizing apogee without watching what we are doing.

As far as running software tweaks, that is not something we have delved into for RasAero. We have another 6" rocket flight coming up much sooner, and I would very much like to tweak the software to match that flight data so that we can get an accurate model. It is built in the same style as our competition rocket.

All I'll add to which to trust is Chucks modelling was recently updated to match real world testing results. Open Rocket has not had it's modelling updated for some time. I do my primary modelling in Open Rocket as it's easy to use. Final modelling is done in RAS Aero. As it is the most accurate in terms if simulation results.
Norm

I'll make sure to let my team know about that. We really haven't been using RasAero at all, but I had wanted to verify with it this time around since it seemed like a good idea. After the flight, I'll tweak our model settings and hopefully, we'll be able to get more accurate simulations in the future.

Thanks for the help, everybody!
 

Chuck Rogers

Well-Known Member
Joined
Jun 7, 2011
Messages
202
Reaction score
134
All I'll add to which to trust is Chucks modelling was recently updated to match real world testing results. Open Rocket has not had it's modelling updated for some time. I do my primary modelling in Open Rocket as it's easy to use. Final modelling is done in RAS Aero. As it is the most accurate in terms if simulation results.
Norm

Norm:

That is indeed the technique many rocketeers use. Initial design of the rocket (layout of the components) on OpenRocket or Rocsim, final runs done on RASAero II.


Below are the model updates, which were based on comparisons with flight data and wind tunnel data, which have been made to RASAero and RASAero II since the initial release in 2008. These are not all of the updates to the software, these are just the updates which were improved prediction models. Oldest is on the bottom, newest is on top. (This information is also on the RASAero II software download page.)


RASAero II release history:


Version 1.0.2.0 – Release Date May 22, 2019
Corrected errors in the nose cone wave drag models for LV-Haack, parabolic, and elliptical nose cones. There were no errors in the nose cone wave drag models for tangent ogive, Von Karman ogive, conical, and power law nose cones. Nose cone wave drag occurs at transonic, supersonic, and hypersonic Mach numbers.

Added new protuberance drag models (for missile raceways, camera shrouds, fin brackets, etc.); streamlined no base drag, streamlined with base drag, and inclined flat plate.

Combined protuberance drags; rail guide, launch lug, launch shoe, streamlined no base drag, streamlined with base drag, inclined flat plate; into one combined protuberance drag coefficient (CD) output.

Made further improvements in the extensions to the power-on base drag model for very large nozzle exit diameters at supersonic and hypersonic Mach numbers, with the nozzle exit area filling a large portion of the rocket base area, for more accurate power-on CD predictions for first and second stages of satellite launch vehicles. No change in the power-on CD for most model, high power, and amateur rockets.

Made mods to the Rogers Modified Barrowman Method nose cone subsonic potential CNAlpha and Center of Pressure (CP) equations for increased accuracy. In the Rogers Modified Barrowman Method the Nose Cone and the body tube which follows the nose cone are treated aerodynamically as a single unit for subsonic potential CNAlpha and CP.

Made mods to the Rogers Modified Barrowman Method expansion section subsonic potential CNAlpha for increased accuracy. No mods were made to the Rogers Modified Barrowman Method expansion section subsonic potential CP.

Corrected error in the Barrowman Method subsonic CP for LV-Haack nose cones.

For rounded and square airfoils with All Turbulent Flow corrected errors in the fin supersonic and hypersonic friction drag.

Version 1.0.1.0 – Release Date August 5, 2016
Changed the recommended stability margin for subsonic Mach numbers to a stability margin of 1.0 calibers, kept the recommended stability margin for transonic and supersonic Mach numbers at 2.0 calibers, for the marginal stability warning message.

Added extension to power-on base drag model for very large nozzle exit diameters at supersonic and hypersonic Mach numbers, with the nozzle exit area filling a large portion of the rocket base area, for more accurate power-on drag coefficient (CD) predictions for first and second stages of orbital launch vehicles.

Version 1.0.0.0 – Release Date September 12, 2015

Initial Release - New and improved models and capabilities compared to the RASAero software:
New and improved supersonic Center of Pressure (CP) models. Rocket supersonic CP typically moved forward approximately 1 caliber compared to previous RASAero models.

New and improved dynamic stability models.

New and improved transonic CD models.

New and improved supersonic boattail wave drag and base drag models for increased altitude prediction accuracy for rockets with boattails.

For very steep or very short boattails, new flight data-based boattail wave drag and base drag models which eliminate unrealistically high altitude predictions for very steep or very short boattails.

Added power law, LV-Haack, parabolic and elliptical nose cones, including subsonic, supersonic, and hypersonic CD, Cnalpha and CP models for the new nose cones.

Added the capability for 5, 6, 7 and 8 fins, to the existing 3 and 4 fins.

Added improved viscous crossflow models to the Rogers Modified Barrowman Method and the supersonic body CP methods for improved predictions for the forward movement of the rocket CP with angle of attack.

New and improved Rogers Modified Barrowman Method subsonic CP models at low angle of attack.

Added improved, increased accuracy launch shoe subsonic, supersonic, and hypersonic CD models.

Added new and improved square airfoil leading edge wave drag model.

Made corrections to square airfoil and rounded airfoil friction drag models for all turbulent Made corrections to square airfoil and rounded airfoil friction drag models for all turbulent flow.

Rocket reference area changed from base area of the nose cone to the maximum cross-sectional area of the rocket body.

RASAero release history:


Version 1.0.2.0 - Release Date March 15, 2011
Made corrections to fix error where on very high altitude flights (above 200,000 ft) extremely low Reynolds numbers would cause unrealistically high drag coefficients. Correction added a low Reynolds number cut-off to avoid very low Reynolds numbers (under 10,000) at high altitude.
Version 1.0.1.0 - Release Date January 7, 2009
Made corrections to fix errors in equivalent sand roughness (surface finish) skin friction coefficient calculations for very rough surface finish rockets. Errors caused very rough surface finish rockets to fly higher than same rocket with a smoother (reduced roughness) surface finish.
Version 1.0.0.0 - Release Date September 11, 2008



Charles E. (Chuck) Rogers
Rogers Aeroscience
 

Chuck Rogers

Well-Known Member
Joined
Jun 7, 2011
Messages
202
Reaction score
134
As far as running software tweaks, that is not something we have delved into for RasAero. We have another 6" rocket flight coming up much sooner, and I would very much like to tweak the software to match that flight data so that we can get an accurate model. It is built in the same style as our competition rocket.

I'll make sure to let my team know about that. We really haven't been using RasAero at all, but I had wanted to verify with it this time around since it seemed like a good idea. After the flight, I'll tweak our model settings and hopefully, we'll be able to get more accurate simulations in the future.

Below are some altitude prediction comparisons for RASAero. Other altitude comparisons for specific rockets, and comparisons with wind tunnel data, are on the RASAero web site ( www.rasaero.com ).

1643158319452.png


Charles E. (Chuck) Rogers
Rogers Aeroscience
 
Top