Puzzling OpenRocket inaccuracies on Aerotech Cheetah

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

4regt4

Well-Known Member
TRF Supporter
Joined
Mar 11, 2022
Messages
972
Reaction score
861
Location
Southern Oregon
First off, I've simmed about 20 or so rocket on OR. And I'm impressed with the results. Except...

For some reason, OR overestimates apogee and flight time on my AT Cheetah by 20-25%. It's consistent with any motor tried. For example, a 24/40 F39 gives 1486ft and 9.03 seconds, while in reality I get 1123ft and 7.3 seconds. I suppose my altimeter could be off some, but the apogee time is supported by a video of the flight. An F24 simmed at 1425ft and 9.13 seconds, I got 1083ft and 7.6 seconds. I realize that OR isn't going to be perfect, but on all my other rockets it has been much closer than this,

I'm using a RocSim file supplied by Aerotech, ported to OR. I've checked the principal dimensions, especially diameter, length, and fin size and shape. I had to add 0.7oz to match my actual measured weight (probably too much paint). So I've verified size and weight, what's left?

FWIW, I'm using the same settings for wind, etc. as I use on my other (more accurate) sims.

Hans.
 
First off, I've simmed about 20 or so rocket on OR. And I'm impressed with the results. Except...

For some reason, OR overestimates apogee and flight time on my AT Cheetah by 20-25%. It's consistent with any motor tried. For example, a 24/40 F39 gives 1486ft and 9.03 seconds, while in reality I get 1123ft and 7.3 seconds. I suppose my altimeter could be off some, but the apogee time is supported by a video of the flight. An F24 simmed at 1425ft and 9.13 seconds, I got 1083ft and 7.6 seconds. I realize that OR isn't going to be perfect, but on all my other rockets it has been much closer than this,

I'm using a RocSim file supplied by Aerotech, ported to OR. I've checked the principal dimensions, especially diameter, length, and fin size and shape. I had to add 0.7oz to match my actual measured weight (probably too much paint). So I've verified size and weight, what's left?

FWIW, I'm using the same settings for wind, etc. as I use on my other (more accurate) sims.

Hans.
You can trim it with the paint settings and fin airfoiling. Rounded finds and "regular" paint works to get my L2 a little closer to the actual measurement.

Doesn't really match reality but there's no setting for the amount of airfoiling and such. I was going to say check the weight but you already did. I would assume that a lot of people forget about glue, fillets, etc when building those files so the weights are always off even if the material specs are perfect.

Also, are you getting straight flights? Are you flying in wind? If it isn't an exactly vertical flight you aren't going to get the performance numbers that you are expecting. Weather conditions and elevation of the launch site may also be playing a role. The first 10k feet of sky near the ground there's kind of a lot going on.
 
You can trim it with the paint settings and fin airfoiling. Rounded finds and "regular" paint works to get my L2 a little closer to the actual measurement.

Doesn't really match reality but there's no setting for the amount of airfoiling and such. I was going to say check the weight but you already did. I would assume that a lot of people forget about glue, fillets, etc when building those files so the weights are always off even if the material specs are perfect.

Also, are you getting straight flights? Are you flying in wind? If it isn't an exactly vertical flight you aren't going to get the performance numbers that you are expecting. Weather conditions and elevation of the launch site may also be playing a role. The first 10k feet of sky near the ground there's kind of a lot going on.
HI, and thanks.

I was launching in the Willamette Valley, Oregon, so it's only a few hundred feet above sea level. Was not a terribly windy day, less than 5mph. Flights were very straight, as was the scary descent on the F39, which ejected over 5 seconds after apogee. That was partly due to relying on OR's estimate of flight time, and also 2+ seconds of "bonus delay". (As measured from the video, the 9 second delay became an 11 second delay... Should have used the F39-6, not the F39-9.)

AT's pre-molded plastic fins are in an airfoil shape, maybe I'll change that to rounded.

Hans.
 
I'll give you one guess what additional piece of information would be extremely helpful to us so we can look into this. ;)
Yeah..... Let me tinker a bit more first.

I changed the fins from "airfoil" to "rounded", as the AT fins are not truly airfoil. But this only lowered altitudes by ~3%. Changing paint to "regular" made a bigger difference, lowering estimates by ~5%. But I'm still nearly 20% too much.

I'll dig around further in some of the setting areas that I normally don't venture into. It's possible that the RocSim file had some sort of (bad) assumptions built in that carried over.

Hans.
 
OK, I'm attaching the file. As noted, I have changed fin profile and paint from what originally referenced in the OP.

Hans.
 

Attachments

  • aerotech_cheetah.ork
    10.1 KB · Views: 0
Thanks! There must be something about this particular file if you're getting better results with your other files. No ideas yet but we'll give it a look and see if anything jumps out.
 
BTW what version of OR are you using?

Also, can you post the original Aerotech Rocksim file, or point me to where I can get it? Is it the one on Rocket Reviews?
 
I'm using beta .4 OR. Let me find the RockSim file. It came from Aerotech's website. As I remember, they are kind of hidden and I had to download a .zip of all their rockets to get it.

Hans.

..Hey, just found it in my Google Drive folder.

Edit: I think I had to do a minor adjustment to diameter. The nose cone was 1.96" and the body 1.9", or vice-versa.
 

Attachments

  • aerotech_cheetah.rkt
    96.7 KB · Views: 0
I also have some rockets that sim to higher apogee in OR. I adjusted the 'paint' finish and that did get the sim closer. Other rockets seem to sim much closer.

Possibly the launch lugs/buttons drag or other component isn't right.

I got your .ork sim to hit 1070 feet & 6.25sec delay by setting paint to Rough and square fins.
 
I also have some rockets that sim to higher apogee in OR. I adjusted the 'paint' finish and that did get the sim closer. Other rockets seem to sim much closer.

Posibly the launch lugs/buttons drag or other compenent isn't right.
Yes, launch lugs....

On one of my own designs, I forgot the launch lugs. Adding them made a *huge* difference. Draggy little things.

To the point of my original post, Aerotech uses some gigantic square profile lugs. The original RockSim file had them as 1/4", regular tube-type lugs. That's one of the changes I made to the original file, enlarging the O.D. to 0.375 (leaving I.D. at 1/4"), which is what the square lugs are. But they are still tube type. I don't know how to change their cross section to square. I could fudge the numbers a bit and make the wall thickness even bigger. Maybe I should try that and see what happens.

Hans.
 
For some reason, OR overestimates apogee and flight time on my AT Cheetah by 20-25%.

I realize that OR isn't going to be perfect. . .

RockSim estimates that your configuration will achieve an even higher apogee than OpenRocket. . .

Compare.OR-RS.png
Probably not really an OpenRocket issue. . . , construction and finish can play a significant role in performance.
 

Attachments

  • Cheetah.RS.AT_F39T-8.pdf
    141.6 KB · Views: 0
Last edited:
I got your .ork sim to hit 1070 feet & 6.25sec delay by setting paint to Rough and square fins.
OK, I changed paint to "unfinished" and square fins. Now the sims for the F39 and F24 are reasonably close.

However, it seems to have screwed up the sims on smaller motors. For example, actual altitude on an E26 Qjet was 681ft, but after changing paint and fin settings, the sim is now 575ft, or roughly 15% too low.....

Hans,
 
On a side, but somewhat related note. . . mass overrides should not be used to override a mass component (especially when the mass and mass override values are the same). And, CG overrides should not be used when a mass component's density is uniform across its length (especially when the CG and CG override values are the same).

Mass Objects.png

I say somewhat related because correcting, removing the redundant overrides does not change the simulation estimations.
 
Last edited:
On a side, but somewhat related note. . . mass overrides should not be used to override a mass component (especially when the mass and mass override values are the same). And, CG overrides should not be used when a mass component's density is uniform across its length (especially when the CG and CG override values are the same).

View attachment 539112

I say somewhat related because correcting, removing the redundant overrides does not change the simulation estimations.
Actually, all those overrides were in Aerotech's original file. I didn't mess with them.

Hans.

Edit: I didn't bother to drill down into those "mass components" because the overall mass and CG were very close. If it ain't broke, don't fix it. As I previously mentioned, I had to add 0.7oz, likely because of heavy paint.
 
Last edited:
I used Aerotech 's file. Hopefully they figured the masses correctly, I'm guessing that's why there are several corrections to some of the components.

The overall mass was only 0.7 oz lighter than my finished rocket, likely because of a rather thick layer of primer and paint. And, yes, I corrected the final weight.

Hans.
 
Given that:
1) You've had good results with OR on other models
2) Rocksim gives similar results to OR
3) This is a simple 3FNC rocket that shouldn't present OR (or Rocksim) with any particular challenges

I'm inclined to chalk this one up to this particular rocket simply underperforming in the measured flights, for unknown reasons. Maybe unexpected atmospheric conditions, I dunno. But at the moment there doesn't seem to be anything pointing to a problem in OR here.

In other words, it is a puzzling inaccuracy. :)

Please follow up with your results when you fly this rocket again in the future.
 
Given that:
1) You've had good results with OR on other models
2) Rocksim gives similar results to OR
3) This is a simple 3FNC rocket that shouldn't present OR (or Rocksim) with any particular challenges

I'm inclined to chalk this one up to this particular rocket simply underperforming in the measured flights, for unknown reasons. Maybe unexpected atmospheric conditions, I dunno. But at the moment there doesn't seem to be anything pointing to a problem in OR here.

In other words, it is a puzzling inaccuracy. :)

Please follow up with your results when you fly this rocket again in the future.
Huh? An anomaly in model rocketry?

Yes, I'll chalk it up to "just one of those things".

I'm unlikely to be at a field that will support F motors soon. But I expect to fly several times in the next month or so on E's.

Thanks much for your efforts. It's appreciated.

Hans.
 
Does your altimeter log the ascent data - or just the final results. I export both the altimeter and the sim data and combine them so I can overlay them. It gives you a better idea of whether you're off by acceleration or drag. I tune a sim by adding mass to get the boost phase to overlay, and adding drag to get the coast to match. Changing the mass is pretty easy - to make big drag adjustments, I add a zero length transition, usually to the bottom, or sometimes the top of the fincan and adjust the diameter to tune the drag.
 
Does your altimeter log the ascent data - or just the final results. I export both the altimeter and the sim data and combine them so I can overlay them. It gives you a better idea of whether you're off by acceleration or drag. I tune a sim by adding mass to get the boost phase to overlay, and adding drag to get the coast to match. Changing the mass is pretty easy - to make big drag adjustments, I add a zero length transition, usually to the bottom, or sometimes the top of the fincan and adjust the diameter to tune the drag.
That's a very helpful idea.

I have a few different altimeters, some log the flight data, some don't. In this case, I am using the Firefly, and reading the data with the uber-rare unobtainium readout. So, unfortunately, I don't have a full flight profile. But I recently scored a PNut altimeter on Ebay, so I should switch that in for this rocket and test it. I already have the interface cord.

Hans.
 
Another thought-- does the Altimeter bay have large enough vent holes?
If not then it is possible the Baro never sees the pressure drop of apogee and is under reporting altitude.
 
HI, and thanks.

I was launching in the Willamette Valley, Oregon, so it's only a few hundred feet above sea level. Was not a terribly windy day, less than 5mph. Flights were very straight, as was the scary descent on the F39, which ejected over 5 seconds after apogee. That was partly due to relying on OR's estimate of flight time, and also 2+ seconds of "bonus delay". (As measured from the video, the 9 second delay became an 11 second delay... Should have used the F39-6, not the F39-9.)

AT's pre-molded plastic fins are in an airfoil shape, maybe I'll change that to rounded.

Hans.
What part of S.Oregon? I'm from Jackson County.
 
Another thought-- does the Altimeter bay have large enough vent holes?
If not then it is possible the Baro never sees the pressure drop of apogee and is under reporting altitude.
Certainly possible. Flights with E class motors seem to match the sims.

I have 4 1/8" holes about 1" below the level of the nose cone shoulder. Instructions from PerfectFlight indicate that should be enough for a 1.9" tube section about 9" long, but.......

Hans,
 
Back
Top