MESOS Flight to 293K ft - RASAero II Prediction 290K ft

[Chuck Rogers]

To All:


I've completed a comparison of the RASAero II Flight Simulation prediction with the Flight Data for the Kip Daugirdas Project MESOS 293K ft Flight flown at Black Rock. The summary of the comparison of the RASAero II Flight Simulation prediction with the Flight Data are as follows:

GPS Apogee Altitude = 293,488 ft
RASAero II Predicted Apogee Altitude = 289,789 ft (-1.26% Error)

Accelerometer-Based Maximum Velocity = 4,047 ft/sec
RASAero II Predicted Maximum Velocity = 4,095 ft/sec (Mach 4.23)
(+1.19% Error)


Note that the error in the predicted altitude was only -1.26%.


The comparison plots for the predicted and actual altitude versus time and velocity versus time are presented below.






The complete analysis for the RASAero II Flight Simulation Comparison with the Kip Daugirdas MESOS 293K ft Altitude Rocket Flight Data is included in the attached PDF file.



Charles E. (Chuck) Rogers
Rogers Aeroscience

 

[waltr]

Looks like the RASAero II Flight Simulation is very, very good. I think 1.2% difference is easily accounted for in motor variation.

 

[Chuck Rogers]

To All:

Attached is the RASAero II .CDX1 input file for the Kip Daugirdas Project MESOS 293K ft Flight that was used in the RASAero II flight simulation comparison. Also attached are the rasp.eng motor data files which were used for the flight simulation. These motor files need to be edited into the rasp.eng motor data file in the Documents - RASAero II directory for the simulation to run. (See the RASAero II Users Manual for adding new motors to the RASAero II rasp.eng file.)


Charles E. (Chuck) Rogers
Rogers Aeroscience

 

[robopup]

Excellent analysis, thank you!

 

[OverTheTop]

You would have to be really happy with the simulation results Chuck. These flights that push the envelope really show up where the hard work has been done in the simulations.

 

[Buckeye]

Note that the error in the predicted altitude was only -1.26%.

This data summary is impressive, but a bit deceiving without the detailed report. The 1.26% agreement was achieved AFTER significant tuning of the simulation with the measured flight data. Ignition delay, launch angle, and aerodynamic drag were all manipulated after the fact. Most troubling is the choice of "smooth" finish to get a good match, even though the finish was damaged during the flight. This calls into question the whole surface roughness approach used by this simulator and others.

Was zero wind speed a valid assumption? A mere 5 mph takes off 8,000 ft from the max altitude. Since you forced a desired downwind range with the launch angle, I guess wind speed had to be eliminated as a variable.

I know this was an extreme case and the tuning helps validate other parts of the code, but most of us don't use simulations in this way.

 

[FredA]

Nice work Chuck!
Great to see you can get this close.

Don't have time to totally digest today, but I'm looking forward to checking into how you modeled the interstage adapter and the overlapping bodies....
Always more to learn - thanks for posting the files too.

 

[Kip_Daugirdas]

This data summary is impressive, but a bit deceiving without the detailed report. The 1.26% agreement was achieved AFTER significant tuning of the simulation with the measured flight data.

I think RASAero is pretty incredible. Really the only major thing that was changed on this sim is the surface finish to achieve the actual flight altitude. The paint/coating damage only occurred at high altitudes/high Mach numbers were their effect was minimal. It would be cool if the simulation would automatically switch surface finishes after a rocket attained these speeds.

I will say the most important things done for these sims was actually ground testing the motors (with data) and weighing the rocket, propellant, etc. This allowed for me to maximize the coast between stages and make good decisions regarding the flight and the rocket’s stability.

Finally, the thrust curves had to be adjusted to sea level - the motors were ground tested in Utah at 5,000’ MSL. Chuck also had to do some thrust adjustments to the second stage motor to account for the larger expansion cone on the nozzle exit. This was all done well BEFORE the flight. No tweaking of .ENG files was done after the fact.

 

[Chuck Rogers]

This data summary is impressive, but a bit deceiving without the detailed report. The 1.26% agreement was achieved AFTER significant tuning of the simulation with the measured flight data. Ignition delay, launch angle, and aerodynamic drag were all manipulated after the fact. Most troubling is the choice of "smooth" finish to get a good match, even though the finish was damaged during the flight. This calls into question the whole surface roughness approach used by this simulator and others.

Was zero wind speed a valid assumption? A mere 5 mph takes off 8,000 ft from the max altitude. Since you forced a desired downwind range with the launch angle, I guess wind speed had to be eliminated as a variable.

I know this was an extreme case and the tuning helps validate other parts of the code, but most of us don't use simulations in this way.

Buckeye:

The question for a preflight prediction for a rocket like this is how much damage to the rocket external surfaces will occur from the aerodynamic heating, and the erosion effects of very high dynamic pressure during ascent. This is obviously very hard to predict. Some of the single-stage Mach 3, and Mach 3+ rockets flown at Black Rock have been recovered really roughed up. The MESOS rocket was going to reach Mach 4 to Mach 4.25, so some damage from aerodynamic heating was expected.

You learn from these flights and adjust your models. Much was learned from this flight.

Part of the difference may be that for these multi-stage rockets, the upper stages reach Mach 3 to Mach 4 at higher altitudes, and thus have less damage from the aerodynamic heating. The AeroPac 104K ft two-stage rocket was more accurately modeled by Rough Camouflage Paint, but the Jim Jarvis 175K ft three-stage rocket and the 293K ft flight of the MESOS rocket were more accurately modeled by Smooth Paint. That's learning from the flight data.

For predictions for future high-altitude attempts, the altitude range can probably be bracketed by running the rocket with the Surface Finish set to Smooth Paint (the Nominal prediction), and then running the rocket with the Surface finish set to Rough Camouflage Paint (the Conservative prediction). This is one of the reasons I included the Rough Camouflage Paint prediction in the charts in the PDF file. From the example of this rocket, it looks like this will vary the predicted apogee altitude of the rocket by about 10% (Nominal Prediction to Conservative Prediction).

But if I was asked to give a prediction, for a multi-stage Mach 3+, Mach 4 rocket, I would say use a Surface Finish of Smooth Paint.

As for running the flight simulation with zero wind, and adjusting the launch angle to match the downrange distance at apogee, this is a standard procedure for a postflight analysis when there is no balloon wind data taken during the launch. Taking NOAA wind data from a site some distance away won't give you local wind conditions at low to medium altitudes. Taking balloon wind data is done very, very rarely during our launches. With no balloon wind data, an exact wind profile cannot be determined. So yes, postflight, the launch angle is varied to match the apogee downrange distance. The original January/February 2022 flight simulations, and later flight simulations, were run with a straight-up launch angle. With no balloon wind data, to check how well the models are working, it's best to run the flight simulation with no wind, and adjust the launch angle to match the downrange distance from the flight.


Charles E. (Chuck) Rogers
Rogers Aeroscience

 

[Buckeye]

You learn from these flights and adjust your models. Much was learned from this flight.

I get the learning part and applying it to future simulations. The motor ground testing was great work. Finding the proper drag model was also good work. If I ever plan to fly to Mach 4, I will choose Smooth Paint in RasAero II.

Where you are losing me is the reverse engineering of the ignition delay and launch angle from the measured data and then boldly claiming 1.26% agreement with said measured data. I should hope so. This seems like correlation merely for correlation's sake. What is learned here that can be applied to the next flight? If I have to launch the model every time in order to complete the inputs to the simulation, then what's the point?

Don't get me wrong. I am a simulation guy. I worked 30+ years in FEA and CFD modeling. Too many of my colleagues were hung up on matching the simulation to the test results rather than designing the product. I think honestly bracketing the simulation with 2 or 3 sigma variation before the build is more useful than trying to fudge 0% error after the product is complete.

 

[jsdemar]

Does RasAero II allow adding a custom atmospheric stack for wind speed and direction vs altitude?

 

[Chuck Rogers]

I get the learning part and applying it to future simulations. The motor ground testing was great work. Finding the proper drag model was also good work. If I ever plan to fly to Mach 4, I will choose Smooth Paint in RasAero II.

Where you are losing me is the reverse engineering of the ignition delay and launch angle from the measured data and then boldly claiming 1.26% agreement with said measured data. I should hope so. This seems like correlation merely for correlation's sake. What is learned here that can be applied to the next flight? If I have to launch the model every time in order to complete the inputs to the simulation, then what's the point?

Don't get me wrong. I am a simulation guy. I worked 30+ years in FEA and CFD modeling. Too many of my colleagues were hung up on matching the simulation to the test results rather than designing the product. I think honestly bracketing the simulation with 2 or 3 sigma variation before the build is more useful than trying to fudge 0% error after the product is complete.

Buckeye:


Only 5 items were changed from the last preflight flight simulation to the postflight simulation presented above.

1) The Launch Site Temperature actually was not changed, but it was confirmed that the Launch Site Temperature was 65 deg F for the launch.

2) The Launch Site Elevation was changed from 3917 ft to 3910 ft, to match Kate data.

3) Based on the on-board accelerometer data, the delay in the ignition of the Sustainer Stage (Stage-2) after it had separated from the Booster Stage (Stage-1) was changed from 16 sec to 15.16 sec.

4) The preflight simulations all used vertical (zero degrees from vertical) launch angles. As noted previously, as no balloon wind data was taken, the flight simulations were run with no wind with the launch angle from vertical adjusted to match the downrange distance at apogee from Kate data. The launch angle from vertical was 2.77 deg.

5) The preflight simulations were run with the Surface Finish set to Rough Camouflage Paint, but as will be addressed below, there was already data available that Smooth Paint might be more appropriate.


As Kip noted, the only really significant "switch" that was changed from the preflight to postflight simulations was the change in Surface Finish from Rough Camouflage Paint to Smooth Paint.

For previous single stage Mach 3 and Mach 3+ rockets flown at Black Rock, the rockets came back with damage from aerodynamic heating, and Rough Camouflage Paint provided a better match to the flight data. For two stage and three stage Mach 3 and Mach 4 rockets, because of the delay in igniting the upper stage (or upper stages) to gain altitude by coasting before igniting the next stage (delay staging), these upper stages reach Mach 3 to Mach 4 at much higher altitudes. This reduces the aerodynamic heating, reducing the damage to the rocket and any possible increase in the surface roughness.

An earlier rocket, the AeroPac 104K ft two stage rocket, did not show this effect. This was actually first seen on the Jim Jarvis 175K ft Three Carb Yen three stage rocket. See the plot below, a Surface Finish of Smooth Paint provided an excellent match to the flight data. The 175K ft Three Carb Yen rocket had much less damage from aerodynamic heating than the AeroPac 104K ft rocket.

From visual inspections of the rockets after flight, and comparing simulations with different Surface Finish settings, it was beginning to appear that two stage and three stage Mach 3-4 rockets behaved differently than single stage Mach 3-4 rockets in terms of damage from aerodynamic heating and the subsequent increase in Surface Roughness (increased roughness in the Surface Finish setting).




So indeed, in the pre-flight simulations for the 293K ft MESOS rocket the Surface Finish setting was set to Rough Camouflage Paint, but it was already becoming apparent that a Surface Finish of Smooth Paint might be more appropriate.

That's where it stood, and the rocket was launched. Now we have a second data point for a two stage or three stage Mach 3-4 rocket. For the multi-stage Mach 3-4 rockets, Smooth Paint appears to be the more appropriate setting for Surface Finish.


Charles E. (Chuck) Rogers
Rogers Aeroscience

 

[Chuck Rogers]

Does RasAero II allow adding a custom atmospheric stack for wind speed and direction vs altitude?

RASAero II only allows one wind speed to be entered, and the wind speed is held constant with altitude. The rocket is assumed to be pointed directly into the wind.

The wind speed model in RASAero II is focused on having an accurate pitch dynamics model when leaving the launch rail (including jet damping which I believe many of the other rocket simulation software packages do not include), and for modeling the weathercocking that occurs early in the flight.

A wind model with magnitude and direction with altitude is something we'll probably be adding in later versions of the program. This is really most useful for dispersion analysis, it would be part of turning RASAero into a Monte-Carlo program. But as I noted, when comparing to flight data for a particular flight, balloon wind data is very, very rarely taken at launches. Very few people are even measuring ground wind speed at launch.


Charles E, (Chuck) Rogers
Rogers Aeroscience

 

[Chuck Rogers]

Where you are losing me is the reverse engineering of the ignition delay and launch angle from the measured data and then boldly claiming 1.26% agreement with said measured data. I should hope so. This seems like correlation merely for correlation's sake. What is learned here that can be applied to the next flight? If I have to launch the model every time in order to complete the inputs to the simulation, then what's the point?

Several additional things were learned from this flight:

1) The time delay from sending the signal to ignite the Sustainer Stage (Stage-2) and the actual detection of the change in acceleration from the onboard accelerometer data. This could be factored into the next flight using Kate and a similar ignitor and motor.

2) Wind is not predicted before the flight. The effect of various variables, including wind, resulted in a no wind launch angle of 2.77 deg. A no wind launch angle of 2 deg from vertical has often been used for simulating these flights, again based on experience from postflight analysis of the prior flights. So running no wind launch angles from 0 deg to 3 deg may be a good way to bracket the altitude.

Don't get me wrong. I am a simulation guy. I worked 30+ years in FEA and CFD modeling. Too many of my colleagues were hung up on matching the simulation to the test results rather than designing the product. I think honestly bracketing the simulation with 2 or 3 sigma variation before the build is more useful than trying to fudge 0% error after the product is complete.

For predictions for future high-altitude attempts, the altitude range can probably be bracketed by running the rocket with the Surface Finish set to Smooth Paint (the Nominal prediction), and then running the rocket with the Surface finish set to Rough Camouflage Paint (the Conservative prediction).


In the end there may a range of settings that bracket the altitude, but from experience there is a specific set of settings which over several rockets gets the best match in altitude.


If I were to run a rocket like this again, for the pre-flight simulation I would use:

1) Ignition delay from this flight if Kate was used, and the ignitor and motor were similar. If the ignitor and motor were different, I'd assume no additional ignition delay.

2) A Launch Angle from Vertical of 2.0 deg with No Wind.

3) A Surface Finish of Smooth Paint.


Overall, the goal is not to match a particular flight, but over a whole series of flights to center the line through the plot of Actual Altitude versus Predicted Altitude. See the plot below (from the RASAero web site, www.rasaero.com ).





Charles E. (Chuck) Rogers
Rogers Aeroscience

 

[Kip_Daugirdas]

1) The time delay from sending the signal to ignite the Sustainer Stage (Stage-2) and the actual detection of the change in acceleration from the onboard accelerometer data. This could be factored into the next flight using Kate and a similar ignitor and motor.

Stage speration is another difficult unknown to simulate as all of my sustainers are designed to drag separate. The booster can still be generating some thrust but if drag outpaces thrust....separation happens. Kate recognized a longer coast time than what was intended because the second stage separated before the booster completely burned out. This adds further complexity to the RASAero simulation.

The time to ignite the second stage motor @T+23 seconds was <0.1 seconds to reach full-thrust.

Below is a snap shot showing the changes in G loading close to motor burnout. Notice the abrupt acceleration change at 7.0 seconds. I'm calling T+7 seconds separation. But according to the thrust curve from motor testing, the O4500 has another second of burntime left. One caveat is that motors tend to burn quicker with acceleration so it could also mean that the O4500 burned out faster. I have attached the booster motor thrust curve and the knock down curve that was used for the simulation below too for reference.

Kate Data (Redline shows separation):


O4500 Thrust Curve as tested at 5,000 ft MSL in Utah:


O4500 Sea Level Knockdown Curve (Used for RASAero Simulation):

 

[VernK]

<snip> The booster can still be generating some thrust but if drag outpaces thrust....separation happens. Kate recognized a longer coast time than what was intended because the second stage separated before the booster completely burned out. This adds further complexity to the RASAero simulation.
<snip>
Notice the abrupt acceleration change at 7.0 seconds. I'm calling T+7 seconds separation.

Below is a plot of the acceleration recorded by Kate during booster motor burn. There is definitely something that happened at T+7 sec. A glitch can be seen in the plot. As you say, that could be the actual staging uncoupling event.

Kate estimated booster motor burn time to be 6.2 sec simply because that is when the acceleration transitioned from positive to negative. However, as you pointed out, the booster motor can still be burning past 6.2 seconds but just not providing enough thrust to overcome the drag. I don't know of a better way to detect motor burn out in real time during a flight. Kate will then fire the sustainer motor igniter based on the coast time the user specifies. In this case, that coast timer started at T+6.2 seconds. What was the coast timer set for on MESOS?

It is also amazing how similar your thrust curve and the acceleration curve look. Despite the fact drag complicates any direct comparison.

 

[Kip_Daugirdas]

What was the coast timer set for on MESOS?

Vern thank you for sharing the plot! That’s awesome to see. The steps in the acceleration tail-off are Grains 5 and 6 burning out. Grains 5 and 6 have larger cores (less propellant). Separation happened somewhere during the burnout of the remaining four grains.

Second stage ignition was set at 23 seconds after liftoff. So ignition is irrespective of coast time.

 

[VernK]

Second stage ignition was set at 23 seconds after liftoff. So ignition is irrespective of coast time.

Okay, that makes more sense now. Here is the acceleration plot for the sustainer motor. Acceleration begins at T+23.1 seconds so the motor lit right away.

By the way, both the booster motor and sustainer motor acceleration plots are very "clean". I take that to mean the burn was very smooth. Often times I see all sorts of small artifacts in the acceleration plots that I believe come from motor burn "irregularities".

 

[robopup]

Kate estimated booster motor burn time to be 6.2 sec simply because that is when the acceleration transitioned from positive to negative. However, as you pointed out, the booster motor can still be burning past 6.2 seconds but just not providing enough thrust to overcome the drag. I don't know of a better way to detect motor burn out in real time during a flight. Kate will then fire the sustainer motor igniter based on the coast time the user specifies. In this case, that coast timer started at T+6.2 seconds. What was the coast timer set for on MESOS?

In post processing scripts for my altimeter data, I define burnout as the time of minimum acceleration. Obviously because of drag, this is still not quite correct, but I think it's a better approximation than acceleration < 0, at least for the purposes of backing out motor performance from a flight.

In my firmware, currently I detect burnout with the same logic you use. I'm considering improving this though and I think something iterative like below could work (although for the purposes of staging etc., acceleration < 0 might be what you care about anyway):

• look for acceleration < 0
• then while acceleration < 0 check if acceleration_i > acceleration_i-1. When true, at burnout

Edit: Sorry for the thread hijack...

 

[Alan15578]

I think RASAero is pretty incredible. Really the only major thing that was changed on this sim is the surface finish to achieve the actual flight altitude. The paint/coating damage only occurred at high altitudes/high Mach numbers were their effect was minimal. It would be cool if the simulation would automatically switch surface finishes after a rocket attained these speeds.

What paint/coating did you actually use? What was learned about paint and coatings over other high performance flights? Did you record any temperature or heating rate data during the flight? When you set your staging delay was it specifically for a heating constraint or was it just to maximize altitude?

Most of my professional work has been with single flight vehicles, but a recoverable aeromodel capable of more than a single flight is something else. There is certainly nothing wrong with touching up the paint after the flight, but there should be enough remaining to be a good prop for show and tell.

 

[Alan15578]

Buckeye:

The question for a preflight prediction for a rocket like this is how much damage to the rocket external surfaces will occur from the aerodynamic heating, and the erosion effects of very high dynamic pressure during ascent. This is obviously very hard to predict. Some of the single-stage Mach 3, and Mach 3+ rockets flown at Black Rock have been recovered really roughed up. The MESOS rocket was going to reach Mach 4 to Mach 4.25, so some damage from aerodynamic heating was expected.

You learn from these flights and adjust your models. Much was learned from this flight.

Part of the difference may be that for these multi-stage rockets, the upper stages reach Mach 3 to Mach 4 at higher altitudes, and thus have less damage from the aerodynamic heating. The AeroPac 104K ft two-stage rocket was more accurately modeled by Rough Camouflage Paint, but the Jim Jarvis 175K ft three-stage rocket and the 293K ft flight of the MESOS rocket were more accurately modeled by Smooth Paint. That's learning from the flight data.


Charles E. (Chuck) Rogers
Rogers Aeroscience

Nice work. Does RASAero provide any aero heating simulation data? It would be useful to at least see a simple stagnation temperature at the nose tip during the simulation.

I am not overly concerned with surface finnish and matching flight data. At some point it becomes a bit silly l like adding base cone stabilisation to Rocksim. Ideally you would want to just input Cd Vs Mach data from whatever source may be available, and have the simulator use that instead of internal estimation. I would be more interested if RASAero simulates or warns or the high altitude coning.

Alan

 

[Chuck Rogers]

I would be more interested if RASAero simulates or warns or the high altitude coning.

Alan:

I recommend that RASAero II be used to predict the Supersonic Center of Pressure (CP) of the rocket, and that a Minimum Stability Margin of 2.0 Calibers be used for all Supersonic Mach Numbers. Rockets which have met this requirement to my knowledge have a nearly perfect track record for having flights where there were no Supersonic stability issues up to Mach 3 and over Mach 3, and now up to Mach 4.23.

One exception was the Jim Jarvis Three Carb Yen 175K ft flight, the plot for which was presented above, where the rocket started coning at 140K ft where the rocket was slowing down at high altitude and the dynamic pressure was very low. How to prevent this still needs work, although the coning did start at high altitude/low dynamic pressure, so it didn't take a lot of altitude off the flight, although you can clearly see the effect in the plot which was presented above.

The Supersonic Stability Analysis for the MESOS 293K ft flight was presented on Chart 8 of the PDF file. The Stability Margin dipped below 2.0 Calibers at the maximum Mach number (Mach 4.23), but not by very much. And the rocket of course flew successfully.


Charles E. (Chuck) Rogers
Rogers Aeroscience

 

[Kip_Daugirdas]

What paint/coating did you actually use? What was learned about paint and coatings over other high performance flights?

Cotronics 4525 was used on the fin leading edges. Then the entire rocket was painted in Rustoleum BBQ paint.

Before you laugh, Rustoleum BBQ paint behaves like very expensive intumescent paints. I apply 6 coats and do a final post cure to 225F. Stagnation temps at Mach 4.2 and 60,000ft are approximately 1,300F if I remember correctly. Spicy 🌶 but luckily air density is low.

This approach seems to work pretty well. I plan to look at different methods/coatings for the leading edges for the next flight. The rest of the airframe/nosecone was well protected via the paint.

MESOS is designed to be fully reusable except for the rocket motor liners and nozzles. In addition, the paint does need to be stripped and reapplied between flights.

Fin leading edge pic:

 

[wclaybaugh2]

Cotronics 4525 was used on the fin leading edges. Then the entire rocket was painted in Rustoleum BBQ paint.

Before you laugh, Rustoleum BBQ paint behaves like very expensive intumescent paints. I apply 6 coats and do a final post cure to 225F. Stagnation temps at Mach 4.2 and 60,000ft are approximately 1,300F if I remember correctly. Spicy 🌶 but luckily air density is low.

This approach seems to work pretty well. I plan to look at different methods/coatings for the leading edges for the next flight. The rest of the airframe/nosecone was well protected via the paint.

MESOS is designed to be fully reusable except for the rocket motor liners and nozzles. In addition, the paint does need to be stripped and reapplied between flights.

Fin leading edge pic:

Kip:

May I ask if you spin balance the payload (or whole vehicle)?

I find myself increasingly thinking that coning (both during boost and at high altitude) is in part due to non-asymmetric mass offsets. It is, of course, normal for an exoatmospheric logintudinally spin stabilized vehicle to transition to a flat spin due to momentum transfer, but any mass offest along the logintitudinal axis seems--to me--likely to abet that process.

I'm actively looking for a shop in N. Colorado that can do a dynamic balance of my current payload, if anyone has any suggestions....

Bill

 

[Finicky]

It is, of course, normal for an exoatmospheric logintudinally spin stabilized vehicle to transition to a flat spin due to momentum transfer, but any mass offest along the logintitudinal axis seems--to me--likely to abet that process.

Best sentence ever

 

[Buckeye]

I am not overly concerned with surface finnish and matching flight data. At some point it becomes a bit silly l like adding base cone stabilisation to Rocksim.

Not quite the same thing. The base drag cone trick is to overcome limitations in the Barrowman static stability prediction. It is not used in flight simulation (I think) and besides, there is no flight data measurement for CP to compare to, anyway. Experimental CP would have to come from wind tunnel or CFD experiments, which are ground tests before flight.

Ideally you would want to just input Cd Vs Mach data from whatever source may be available, and have the simulator use that instead of internal estimation.

Agree, this would be a nice feature in the software, like the old wRASP. The problem is that there are not many sources of Cd vs. Mach Number for the hobbyist. So, the hobbyist has to rely on the internal models. I have seen Cd backed out from accelerometer flight data and maybe some old wind tunnel measurements here and there. Do you have an available source?

 

[Chuck Rogers]

I recommend that RASAero II be used to predict the Supersonic Center of Pressure (CP) of the rocket, and that a Minimum Stability Margin of 2.0 Calibers be used for all Supersonic Mach Numbers. Rockets which have met this requirement to my knowledge have a nearly perfect track record for having flights where there were no Supersonic stability issues up to Mach 3 and over Mach 3, and now up to Mach 4.23.

When running the RASAero II software, if during the Flight Simulation run the Stability Margin of the rocket at Subsonic Mach numbers falls below 1.0 Calibers, or for Transonic and Supersonic Mach numbers the Stability Margin falls below 2.0 Calibers, the Marginal Stability warning message shown below will be displayed. The rocket will continue the Flight Simulation run, but it is recommended that the User go back and check the Center of Gravity (CG) and the Center of Pressure (CP) and assess whether to make changes to the rocket. This assessment for the MESOS 293K ft Flight Rocket is what was shown on Chart 8 of the PDF file that was attached to an earlier post. If you run the RASAero II .CDX1 input file for the MESOS 293K ft Flight Rocket which was attached to a previous post you will see the Marginal Stability warning message below.





By checking the CG and CP with time, the Stability Margin with time, and the Mach Number with time, the User can assess if there will be stability issues with the rocket. The Stability Assessment for the MESOS 293K ft Flight Rocket which was presented on Chart 8 of the PDF file showed that the Stability Margin dipped below 2.0 Calibers, but not by very much.

A Technical Note; the Stability Margin check and warning message is only for angles of attack less than 5 deg. (-5 deg < Angle of Attack < +5 deg) This is because, particularly for Subsonic Mach numbers at launch when the rocket is leaving the launch rail, or at apogee, there can be momentary excursions in angle of attack, and as the RASAero II software models the forward movement of CP with angle of attack there can be momentary reductions in Stability Margin below the limits.


Charles E. (Chuck) Rogers
Rogers Aeroscience

 

[Buckeye]

Overall, the goal is not to match a particular flight, but over a whole series of flights to center the line through the plot of Actual Altitude versus Predicted Altitude. See the plot below (from the RASAero web site, www.rasaero.com ).

I will make a mental note that these comparisons are after extensive back calculating from the measurements after the flight is complete. First time predictions are probably more like +/-10% agreement which is pretty typical of the softwares and good enough for design work.

 

[Chuck Rogers]

Nice work. Does RASAero provide any aero heating simulation data? It would be useful to at least see a simple stagnation temperature at the nose tip during the simulation.

Alan:

RASAero II does not include any aerodynamic heating models. I have an excellent stagnation heating model in another program where I can get stagnation temperature with time and the total stagnation heat load. I plan to take a look at the various flights to do at least a stagnation temperature comparison and a stagnation heat load comparison between the flights.

One issue is that this stagnation heating model is an equilibrium model, i.e., you'd use it for cruising at Mach 3 for 30-60 minutes, rather than a rocket going Mach 3 for tens of seconds, it will overpredict the heating. The analogy I give for our rockets is rather than soaking in an oven for 30-60 minutes, you put the rocket in the oven for tens of seconds and then quickly pull it out.

But this will answer the question of how much less is Mach 3 to Mach 4 aerodynamic heating at 10K ft to 20K ft, versus at 30K ft to 60K ft.


Charles E. (Chuck) Rogers
Rogers Aeroscience

 

[Chuck Rogers]

Agree, this would be a nice feature in the software, like the old wRASP. The problem is that there are not many sources of Cd vs. Mach Number for the hobbyist. So, the hobbyist has to rely on the internal models. I have seen Cd backed out from accelerometer flight data and maybe some old wind tunnel measurements here and there. Do you have an available source?

Buckeye:

There is a whole Comparisons with Wind Tunnel Data Page on the RASAero web site (see the left-hand buttons on the site), and I put the full reports there from which the wind tunnel data was obtained.

I've also done comparisons with CD's backed out from inflight acceleration measurements during coast, see Comparisons with Flight Data, the Adrian Adamson Violent Agreement flights.

Actually, while high power rocketeers use both the aerodynamic prediction and flight simulation capabilities of RASAero II, RASAero II is extensively used to generate aero data for other flight simulation programs, including space launch vehicle ascent to orbit programs. Many undergraduate and graduate students and small space start-ups use RASAero II for their aerodynamic data, and then use that aerodynamic data in other flight simulation programs.


Charles E. (Chuck) Rogers
Rogers Aeroscience