Open Rocket sim on LPR 1/2 altitude of Rocksim

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Jerome, post the rkt file. I've seen openrocket produce wacky data for some imported files - especially if there are cg overrides. Please tell me where rocksim thinks the cg is.

Kevin
 
I've tried a couple of things, but it seems to sim in OR to about 500ft on a C6. What do you expect to simulate to?

Kevin
 
Rocksim is telling me about 1051 ft. Being such a small and light weight rocket, I have to imagine it going more than 500 ft on a C6.

I created the sim using Rocksim.. imported it in Open Rocket. Both RS & OR are showing the CG & CP in the same spot.


Jerome
 
Rocksim is telling me about 1051 ft. Being such a small and light weight rocket, I have to imagine it going more than 500 ft on a C6.
Look at the Cd of the fins in "component analysis" -- for whatever reason, it thinks they are very draggy if set to square (they are pretty thick for a small rocket.) If I set the cross-section to rounded, I get over 1000 feet on a C6.
 
I noticed the fins were really draggy as well - but I thought it was something with the inner tube extending so far to the rear.

Does rocksim let you specify fin cross section? What does it think it's set to?

Kevin
 
Hmm.. I set the cross section in both to "rounded".. Rocksim calc'd apogee at 1066' (15' more than squared)... Open Rock calc'd 1057'.

Interesting....

Jerome
 
Jerome,

Could you post the rkt file saved with square fins and with aerofoils?

Thanks

Kevin
 
I'm starting to see a pattern here... seems with several LPR sims in OR, if I leave the fins as "squared" I get half the altitude as "round" or "airfoil".


Jerome
 
Jerome,

Sampo agrees that the OpenRocket square edge Cd might be too pessimistic. This summer I'm going to do some flight experiments to get some better data.

Kevin
 
Jerome,

Thanks again for the ork file. Not only did you provide me with a little R&D project, you also provided me with a little "inspiration".

Kevin

IMG_20130502_210404_067.jpg
 
Anyone determined which sim results to trust? Openrocket apparently increases drag on square leading edged fins much greater with increased velocity than does Rocksim. Which is correct? Note the diverging values between Openrocket and Rocksim as max v is increased on my already-built rocket shown below:

G126-WT Openrocket Rocksim

Apogee 601 ft 769 ft
Max. Vel. 220 ft/s 278 ft/s
Max. Accel. 330 ft/s^2 342 ft/s^2
Time to Apogee 6.09 s 7.09 s

H160-SK Openrocket Rocksim

Apogee 1224 ft 1812 ft
Max. Vel. 373 ft/s 403 ft/s
Max. Accel. 345 ft/s^2 348 ft/s^2
Time to Apogee 8.15 s 10.37 s

I470-WT Openrocket Rocksim

Apogee 2386 ft 4382 ft
Max. Vel. 789 ft/s 910 ft/s
Max. Accel. 973 ft/s^2 983 ft/s^2
Time to Apogee 9.78 s 14.13 s

J270-GR Openrocket Rocksim

Apogee 2842 ft 5139 ft
Max. Vel. 726 ft/s 914.77
Max. Accel. 451 ft/s^2 470 ft/s^2
Time to Apogee 11 s 15.59 s

14035601914_d39494e344_b.jpg
 
They are both sims and as such are approximations of reality.

You need to first realize that no matter what program the results will be dictated by the quality of the inputs.

That being said, what diameter and weight is your test rocket. Considering the velocities you list you should be ok using barrowman equations. My guess is that Rocksim is probably about 5-10% optimistic based on a lot of experience.



Mark Koelsch
Sent from my iPhone using Rocketry Forum
 
Those are large differences. AFAIK, both simulators use the same basic physics for simple flight simulations so I suspect that you are using a slightly different input file for each simulation. Very carefully check your input an output files for differences. If you look hard enough you will find something.

Several initial suggestions.

Are the engine files you use identical? If not then you can expect a difference.

Are you using database values for component weigh? If so the can be differences in the parts data base. Use the mass override option and rerun the simulations so see it the differences disappear.

Make sure all parts are attached to the rocket if you graphically designed the rocket. Small graphical input differences may result in a configuration change that was not intentional.

Bob
 
Those are large differences. AFAIK, both simulators use the same basic physics for simple flight simulations so I suspect that you are using a slightly different input file for each simulation. Very carefully check your input an output files for differences. If you look hard enough you will find something.

Several initial suggestions.

Are the engine files you use identical? If not then you can expect a difference.

Are you using database values for component weigh? If so the can be differences in the parts data base. Use the mass override option and rerun the simulations so see it the differences disappear.

Make sure all parts are attached to the rocket if you graphically designed the rocket. Small graphical input differences may result in a configuration change that was not intentional.

Bob
Both files are identical, originating from Rocksim (i.e, the Rocksim file was loaded into Openrocket). This is the very issue that caused this thread to be started in the first place a year ago with a different rocket and motor showing radically different results between Rocksim and Openrocket with the only difference being the simulation with square leading edge fins, but the issue was never adequately resolved here.

Openrocket results for rockets other than those with square leading edges fairly closely matches Rocksim results when using Barrowman Equations. Not so with square edge fins where the results are radically different, the difference increasing with the maximum velocities reached. The simulated drag of square edge fins must be much greater in Openrocket than Rocksim, that difference increasing with velocity. Results within Rocksim for rounded and square edge fins are nearly identical using both the Barrowman and Rocksim equations within Rocksim.

Trouble is, which is most correct, Rocksim or Openrocket? The differences are highly significant if one is using motor delay deployment, especially on a 4" OD rocket.

I'm building a scaled down (BT-70) version of the six-fin 4" rocket shown above and will fly it with motors which take it to significantly different velocities carrying an Altimeter II, then I'll see which simulation software is closest to being correct with square fin leading edges. I don't know what issue, if any, the different Reynolds Number might introduce, but I'll hopefully get a ball park idea of which simulation software is more correct.

As fun as my experimentation might be, there must be all kinds of people who have launched rockets with square leading edge fins with an altimeter on-board who can check their altimeter-measured results against Openrocket's and Rocksim's predictions and let us know which simulation software was closest to correct. I tend to think it will be Rocksim.
 
Those are large differences. AFAIK, both simulators use the same basic physics for simple flight simulations so I suspect that you are using a slightly different input file for each simulation.

"The same basic physics" is quite mouthful when talking about aerodynamics. Both use the Barrowman equations for basic stability computations, so those should be relatively close for "typical" rockets. Both extend the stability computations in various ways. The drag computation methods can differ quite a lot - there's no general standard way of computing that.

The OpenRocket methods are documented in the technical documentation (https://openrocket.sourceforge.net/documentation.html). The RockSim methods are unknown and undocumented.


My feeling is that the OpenRocket methods for thick, square fins are too pessimistic. Personally, I'd trust the values for rounded fins. Will be good to get some test flight data from Kevin.


Cheers,
Sampo N.
 
I'm afraid my test data is quite incomplete. I was hoping to get the opportunity to have additional flights this year to help fill out the data a little more.

Of course, anybody can contribute data to this effort :) Speaking of which I have some HPR flights I could easily add to the mix as well.

Kevin

"The same basic physics" is quite mouthful when talking about aerodynamics. Both use the Barrowman equations for basic stability computations, so those should be relatively close for "typical" rockets. Both extend the stability computations in various ways. The drag computation methods can differ quite a lot - there's no general standard way of computing that.

The OpenRocket methods are documented in the technical documentation (https://openrocket.sourceforge.net/documentation.html). The RockSim methods are unknown and undocumented.


My feeling is that the OpenRocket methods for thick, square fins are too pessimistic. Personally, I'd trust the values for rounded fins. Will be good to get some test flight data from Kevin.


Cheers,
Sampo N.
 
My feeling is that the OpenRocket methods for thick, square fins are too pessimistic. Personally, I'd trust the values for rounded fins. Will be good to get some test flight data from Kevin.
On the first point, same here. My test data will be done within the next few weeks, assuming the weather cooperates and the rocket and its altimeter aren't lost.

On the Rocksim method, Apogee Newsletter 238 excerpts:

"Paul came up with a really unique method of finding the CP of free-form shapes. In fact, we copyrighted it! If you are interested in reading how we do it, you can purchase Technical Publication #17... What I like to tell people is that the RockSim Method is the Barrowman Method. The difference is that we went back to the original equations and tried to remove as many assumptions as possible. The result is that you can design a wide variety of configurations and the rockets will be stable."

From that, I gather that the special sauce in the Rocksim method is mainly the ability to find the CP of free form shapes, something that the Barrowman Equations can't do in some cases.

Here's a mention of an Openrocket departure from the Barrowman method used for calculating fin pressure drag, but I have no idea if it's relevant to the apparent difference between Rocksim (using Barrowman Equations) and Openrocket when a square leading and trailing fin profile is used:

From OpenRocket technical documentation for OpenRocket version 13.05, 2013-05-10, Sampo Niskanen:

3.4.4 Fin pressure drag (excerpts):

Barrowman estimates the fin pressure drag by dividing the drag further into components of a finite thickness leading edge, thick trailing edge and overall fin thickness [4, p. 48{57]. In this report the fin thickness was already taken into account as a correction to the skin friction drag in Section 3.4.2. The division to leading and trailing edges also allows simple extension to the different profile shapes.

In the case of a rectangular fin profile the leading edge pressure drag is equal to the stagnation pressure drag as derived in equation B.2 of Appendix B.1 (equation 3.90)

The fin base drag coefficient of a square profile is the same as the body base drag coefficient in equation 3.94.
 
The test rocket design is as shown and photographed. Measured mass and CG override was used. Actual launch altitude and weather conditions were used in both Rocksim (RS) and Openrocket (OR) simulations. Ambient temperature was 63 F.

Motor used was an Aerotech E20W-7 single use. In 10 E20 NAR certification samples: Total Impulse spec 35 N-s; measured average 35.01 N-s; Std Dev % only 0.8%!

NAR-reported standard deviation % for Max Thrust, Avg. Thrust, and Burn Time were 2.6%, 1.8%, and 1.7% respectively

Any delay time error in the motor used is irrelevant to the maximum altitude actually achieved since deployment was post-apogee.

So, in all respects, the single motor used can most likely be considered to be in more than adequate compliance with the motor data used by both RS and OR. Anyway, good enough to get a ballpark comparison of RS and OR predictions versus measured results.

A second flight data acquisition attempt was to use an F32T-8, but it became too windy for optimum vertical ascent due to weathercocking. Only minor weathercocking was observed during the E20W-7 flight. The F32T-8 flight will be done at some near-term future date followed by another E20W-7 and F32T-8 flight to acquire more data points.

Note that this square fin leading edge test rocket was designed to be an extreme example with its six thick, unswept, and flat leading edges. I'll leave it to others to test the square vs rounded/streamlined leading edge simulation differences versus measured results between RS and OR for less extreme, far more typical examples with fewer and swept fins.

Results with Aerotech single use E20W-7, only the most important values for rocket survival using motor delay train deployment method listed:

Apogee
------
1137 ft - Rocksim (145% of measured)
713 ft - Openrocket (91% of measured)
783 ft - Measured (by Altimeter II v2)


Deployment Velocity
-------------------
19 ft/s down - Rocksim
58 ft/s down - Openrocket
48 ft/s down - Calculated from the measured apogee to deployment time of 1.5s using this on-line calculator and my calculated air resistance (k) value of 0.001471 kg/m from k=0.5*rho*Cd*A assuming a Cd of 0.75:

https://keisan.casio.com/exec/system/1224830797

Given the low velocity achieved prior to deployment and the very low drag of a rocket, that complex path gives effectively the same result as the simple vacuum velocity formula v=a*t.

Descent Velocity (parachute)
16.5 ft/s - Rocksim
15.6 ft/s - Openrocket
16 ft/s - Measured

EDIT: Results with Aerotech single use F32T-8 to conduct test at higher max velocity:

Apogee
------
1891 ft - Rocksim (162% of measured)
1065 ft - Openrocket (91% of measured)
1167 ft - Measured (by Altimeter II v2)


The Rocksim's apogee value overshot measured apogee by a greater amount (162% vs 145%) as the maximum flight velocity increased indicating that it was underestimating the increase in fin drag at higher velocities.

Openrocket's predictions remained the same (91% vs 91% of measured altitude) at both lower and higher max velocities indicating that it was properly compensating for the increased drag of the six flat leading/trailing edge fins at higher velocities.

14029218019_71dd9b27f7_o.jpg


14215884765_5669810955_c.jpg
 
Last edited:
Test - rounded fin leading edges - square trailing edges
--------------------------------------------------------------

Because of the tail wag seen near max-v in the above listed flights of what should have been a very stable rocket, I suspected that the flat leading edge of each fin was causing major turbulence that was reducing fin effectiveness.

To test this, I bought a bag of perfectly straight 12" lengths of very accurately and uniformly machined 3/16" wooden dowel from Hobby Lobby (the 3/16" dowels at Home Depot and Lowes were horribly machined, warped garbage), epoxied the appropriate lengths to the leading edge of each fin, filled in the gaps, and painted. The result was a perfectly rounded leading edge with only a very slight increase in fin area.

Results with Aerotech single use E20-7:

Apogee
------
1124 ft - Rocksim (125% of measured)
1081 ft - Openrocket (120% of measured)
903 ft - Measured (by Altimeter II v2)

Max-v
-----
290 fps - Openrocket
299 fps - Measured (by Altimeter II v2)


The flight was straight as an arrow and rock stable with absolutely no hint of tail wagging. Note that neither Rocksim nor Openrocket allow the setting of rounded leading edges only, explaining at least some of the overprediction of altitude. But also note that there wasn't nearly as wide of a difference in their apogee predictions as there was for the square leading and trailing fin edge flights.
 
Last edited:
This is fascinating work. Thank you!
You're most welcome. I really enjoy opportunities to experiment.

My tests indicate that for rockets with unswept flat leading edged fins, Openrocket predicts apogee more accurately than Rocksim. Turbulence from same may reduce fin effectiveness as it apparently did in this case.

Intuitively, I think that fins with swept flat leading edges probably produce less drag because of the outward radial airflow along the swept leading edge. Rocksim might be more accurate in that much more typical case. Swept fins with flat leading edges might experience less reduction in fin effectiveness from turbulent airflow for the same reason.
 
Back
Top