OR sims are short of actuals

I've created a Open Rocket file for one of my old rockets so I can do sims on some EX propellant. The first step is to make sure the model and sims are accurate with known commercial loads.

What I found was that the OR sims were all short of actual flights.

I flew a AT K550W to 6400 ft and the sim shows 6150, a 250 ft diff
I flew a CTI K675SK to 7380 ft and the sim shows 7180, a 200 ft diff
I flew a CTI L935W to 9560 ft and the sim shows 9400, a 160 ft diff

The OR file has the physical shape of the rocket correct, including fin profiles. The fins are set to airfoiled and all surfaces are set to Polished, or the sims would be even further away from actual. There are almost no internal components, but a mass object is in place to give the rocket the correct weight and balance.

Anyone have any idea why the sims are off and what can be done to get them to be accurate?

View attachment Grapeshot.ork

I think your sims are pretty accurate and would not be concerned.

There are too many factors to consider in consistently obtaining more accurate simulations, namely differences in motor performance vs. motor profile, winds, temp, humidity, even the friction on the rail, etc. I see you have already adjusted the surface or CD to the maximum as well as applied an airfoil profile to the fins. I did notice that if you play with the fin thickness it makes a difference however not consistently across all three simulations.

For example I went down to 9/64 (.1406) and it was really close on the K550 (6388' AGL) but the bigger the motor the bigger the difference, which is reversed from your original sim. K675 7449' AGL and L935 9873" AGL. Possibly something in between would be a happy medium.

Personally I would be happy with your results and just keep that in mind when planning your flights.

If you are within 10% thats acceptable, most of my rockets that have been flown with altimeters seem to be within that perentage.

That seems to be within the margin of error on the motors themselves. The reaction to wind is hard to sim because even if you put in the ground level wind accurately, you don't accurately know what's going on up above. I don't know OR but the few times that I've been concerned about how my rocket reacts to wind, I ran a whole set of sims with different wind speeds and variability.

Another thing most people dont modify on the sim in OR is the temperature, it can make a significant difference on a sim as it affects air density.

Dude. All your simulations/flights are less than 4%. Move along, nothing to see here.

Buckeye said:

Dude. All your simulations/flights are less than 4%. Move along, nothing to see here.

Click to expand...

You're probably right. OCD kicking here I guess. I was just wondering why I had to go with polished surfaces and airfoiled fins to even get the sims close to the actuals. Most people complain about the sims being higher than the actual flights. Since mine were below actual flights and the Cd inputs were set to max the altitude, I figured I must have something significantly off.

Handeman said:

You're probably right. OCD kicking here I guess. I was just wondering why I had to go with polished surfaces and airfoiled fins to even get the sims close to the actuals. Most people complain about the sims being higher than the actual flights. Since mine were below actual flights and the Cd inputs were set to max the altitude, I figured I must have something significantly off.

Click to expand...

Yeah, just giving ya the business. :wink:

I have seen this before, too. Are your flights supersonic? The Cd models may overestimate the drag around and above Mach 1. Rocksim is absolutely horrible in this regard. OR is much better, but nothing is perfect.

I think that most people complain that sims predict much higher than they realize is because they don't have anything close to actual weights in the sims, nor quite as good a finish as they presume, nor the correct environmental factors as their actual launch. That and they actually believe the "Flies to 10 Bajillion feet" propaganda printed on model rocket kit face cards. :lol:

There might be something to the way the sims compute things with regards to the motor performance, too. My sims usually predict about 5-8% high for my black powder powered models (LPR) but anything that's burning AP it's usually predicting about 5-10% low. I've done back to back flights on 29mm BP and AP on two seperate launches, with two separate altimeters (estes, jolly logic alt2) just to see if I was crazy. (My wife confirms that I'm crazy, but the rocket data shows a well defined trend!). The Alt 2 and a Stratologger in other MPR/HPR vehicles is proving out that my AP burning rockets go higher than OR predicts, as well, even when I have the sim to the gram accurate on specific components and overall weight distribution.

4%? I'd call that well within natural variation, never mind variations induced by winds aloft!
As long as you're not within striking range of your waiver, kick the tires and light the fires.......and rely on the data from the vehicle! :fly:

Banzai88 said:

There might be something to the way the sims compute things with regards to the motor performance, too.

Click to expand...

No. The simulations do the same thing regardless of the motor type or any other input to the code.

But you hit the nail on the head. Thrustcurves vs. real motor burns are the biggest sources of error in any correlation effort, overshadowing most mass, Cd, and weather discrepancies. A bad motor file can screw up a simulation by 20%. Just look at the natural standard deviations in the certification test fire data sheets. Also, OR pulls motor files from Thrustcurve.org which has a lot of multiple entries and unconfirmed mass data in many of the submitted motor files. Not sure how OR decides which motor file to use in its database.

Motor variation is larger than that variance. Chances are you did not measure the rail angle of actual flights, which could easily account for this. There are a multitude of things that could account for this.

Your sims sound like they are on the money to me.

markkoelsch said:

Motor variation is larger than that variance. Chances are you did not measure the rail angle of actual flights, which could easily account for this. There are a multitude of things that could account for this.

Your sims sound like they are on the money to me.

Click to expand...

I would say you're right, except the sims are constantly low when compared to actual flights. What concerns me is that to get that close, I have to set all finishes to Polished and the fins to airfoiled. With most of my other sims, I set the finish to paint, or rough to get the sims to come down low enough to match actuals. Using Polished for all finishes and still not simming as high as actuals is an anomaly I haven't seen before. I figured I'm not doing something right.

I doubt you are doing it wrong.

How fast is the rocket traveling?

With sims that close, I'm jealous.

A good example of low sims would be my L2 cert. Expected ~5,500' under perfect conditions and nominal flight path. Actual result was 5,999' with a big weathercock as soon as it cleared the rail.

Flew the same rocket again a month or so later with a larger motor. Expected 6,500'. Actual was 7,300' with an almost perfect flight.