Anyone Static Test Stock Motors??

[LarryC]

So I have noticed, in accelerometer analysis, that flight data impulses are consistently short of standard impulses. (BTW, I have seen examples of only Aerotech composites in flights data worthy of impulse computation. That sampling bias is an accident of my experience.)

This effect has been variously attributed to

1) Bad computations and
2) flights that were actually somewhat off-vertical analyzed as if they were perfectly vertical.

Regarding item 1, I have generated faux accelerometer files in computer simulations and backed out the correct impulse. I reject number 1 a cause.

Regarding item 2, I can demonstrate that in this case, impulses should be slightly overestimated instead of very underestimated. I reject item 2 as a cause.

I have also looked at launch detect, noise, and the failure of thrust to register until it exceeds the launch weight of the vehicle. None of these effects accounts for the consistently lost impulse.

In addition, simultaneous altimeter data and observable performance are consistent with many of the inertial data.

Finally, accelerometer (subsonic) Cd curves level out later than corresponding DATCOM curves, but they level out at darn near the same value. If you use a standard thrust curve, you back out a significantly higher Cd for the flight – it is higher than the flight summary Cd you would get from DATCOM. This is a known defect, and it is normally attributed to the DATCOM method. I am beginning to wonder if it is actually a defect in the standard motor values.

That is, I wonder if the motors tested are sampled from the population of motors we actually buy. It would be irresponsible to make an accusation simply because I cannot think of other explanations, when best evidence is obtainable through other means. I am doing no such thing. I am merely running down another possible cause of low flight impulse values.

So, I ask: Has anyone taken composite motors (e.g.; the Aerotech G40) off of the shelf (rather than supplied by the manufacturer for the purpose) and static tested them for impulse? Are the values close to standard values?

Thanks and Regards,
-Larry C.

 

[Loki]

Larry,

Are you doing the thrust from acceleration computations manually, or are you using the ARTS DataAnalyzer software? If the latter, I have some suggestions for you that may help (email or call me).

As I do lots of static testing and flying, I can say that accerometer based curves can be very very good.

 

[LarryC]

Jeff,

I am doing them in my own program. I would appreciate any advice you may give me, but I am confident my computations are correct.

 

[LarryC]

OK, Jeff,

I was indeed using one of your fine flight computers, so I ran the data through your analysis program. For a G40 motor,

ARTS Data Analyzer Program: 82 ns
My program (OVAA2.xls): 94 ns
Tripoli Impulse Rating: 120 ns
RASP.eng rating 124 ns

It looks like your data analyzer might not be adding back the effect of drag. At any rate, both programs are well below 120 ns.

I have five more flights on the same kind of motor in two rockets. All have comparable impulses. I cannot run the others through your analysis program because they are off-vertical. The one reported here was a vertical launch.

Regards,
-Larry

 

[Rocketjunkie]

Originally posted by LarryC
OK, Jeff,

I was indeed using one of your fine flight computers, so I ran the data through your analysis program. For a G40 motor,

ARTS Data Analyzer Program: 82 ns
My program (OVAA2.xls): 94 ns
Tripoli Impulse Rating: 120 ns
RASP.eng rating 124 ns

How old is the G40? The newer ones (molded case) only have 100 N-s.

 

[metalwizard]

just wondring, how dose flight data have anything to do with a motors output?

there is no simulator in the world that can take into consideration EVERYTHING..

for instance..

dose your program take into consideration, relative Humidity? ambiant tempature, wind conditions at ALL altitudes, how about your rocket, dose it consider the finish (paint etc) variables in the paint surface will effect total performance. things like the airfoil in your fins, are the exactly the same (to say the 1000th inch?) motor tempature at its firing. your ignition, did it fire absolutly immediatly? or did it chuff a tiny bit (so small the nekkid eye and ear couldnt catch it..)

anyways I know some of that is out there but it ALL effects you rockets performance. in other words the only way to see the exact output of a specific motor would be to put it on a load cell and fire it... or should I say fira a big hand full and get an average. because even two motors made from the same mold and propellant etc are gonna be slightly different..


just some things to think about..

 

[LarryC]

Regarding whether my program takes everything into account - no, but I am using closure with altimeter altitude as a criterian. The day was calm enough that some rockets were landing on the launch pad. Also, the finish of the rocket has nothing to do with it, because at small angles of attack, the accelerometer reading is the sum of thrust and drag acceleration.

That said, I have found the answer to my question at the Aerotech web site:

=======
Aerotech now lists the total impulse of the G40 motor as 100ns. My values were 94-97.

Aerotech now lists the impulse of the F25 as 73 ns. Mine was 70.
=======

If you go to Tripoli or the NAR, you see the G40 listed as 123 ns.

It looks like the specs have changed, and the values were pretty good after all.

So... Never mind.

 

[metalwizard]

glad you got it figured out, people somtimes think that numbers should be exact every time. thats why I gave the extreme example of things to take into consideration..

 

[LarryC]

OK, I'm still flummoxed. The Aerotech catalog listed on their website is 2005-2006. The test date for the G40 listed on www.thrustcurve.org is 30 May 2006 and has a total impulse of 113 ns. Once again, the catalog lists the impulse as 100 ns, and that is most consistent with my flight impulses. (Actually, I was glib in my previous post. I'm getting an average of 96 ns over eight flights I scrounged from my records.)

I can see the launch rod eating up ~5 ns, but I cannot see 17 ns going up this way. The ARTS unit has enough lead-in data that I can tell, by the thrust curves, that launch detect isn't a factor. Altimeter data are consistent with inertial data. The motor is single use so assembly is not a factor.

Finally, I've never before seen a manufacturer understate the impulse of a motor, and Aerotech's listing for the G40 is significantly short of the tested value.

Again, I'm making no accusation, but I must ask the simple question again:

===
Are the motors that our national organizations test sampled from the same population as that of the motors we fly?
===

Does anyone here test off-the-shelf motors? That is best evidence in this matter.

Thanks,
-Larry Curcio

 

[n3tjm]

Aerotech has changed the design of the F25, F50, G40, G80 so many times. The F's used to be close to 80 and the G's 120. However, they switched to their molded cases that required a liner, so the impulse dropped significantly.

 

[billspad]

Originally posted by LarryC

Again, I'm making no accusation, but I must ask the simple question again:

===
Are the motors that our national organizations test sampled from the same population as that of the motors we fly?
===

Does anyone here test off-the-shelf motors? That is best evidence in this matter.

Thanks,
-Larry Curcio

The motors manufacturers submit for testing are supposed to come from a regular production lot. Whether they are or not is anybody's guess. Same thing for motors tested for recertification.

I haven't tested many but I have done some to see how my test stand was working. They were within limits.

As Doug explained, Aerotech has made many changes in the past couple of years. You really have to look at the date the motor was manufactured and check the NAR website to see if you can find an engine data sheet for that particular version. Take a look at NFPA 1125. There's a section that addresses design change. See if you can figure out how much they can change a motor before it has to be recertified.

 

[metalwizard]

Again, I'm making no accusation, but I must ask the simple question again:

===
Are the motors that our national organizations test sampled from the same population as that of the motors we fly?
===

Does anyone here test off-the-shelf motors? That is best evidence in this matter.

Thanks,
-Larry Curcio

the motors that go to be tested are the same as the motors you fly, (at least from aerotech) I build them exactly the same no matter were they are going.
again. the only way to get a true reading of the output of any motor is to put it on a load cell and fire it..
I do that all the time. We test at LEAST 1 out of 100 motors built, as long as they fall within spec they go out. if not (very very rare) I test at LEAST two more.
again if they fall in spec they go out, if not (very very rare) we look into why...

bottom line you get the same motor that S&T gets. If we make a change to the design, and it changes the spec of the motor then it goes to S&T and get recertifed. Just for an example. the F23FJ was just recertified. why because people wanted a G33 rms that was shippable (under62.5 G's) we made changes to that casting/propellant. since the f23 econojet uses the same grain as the G33 (two grain version) we changed it to.(now they are both shippable) It actually came into spec, but since it was a change to the propellant we had the f23FJ econ recertified... and I'll bet you never new we did that.. heres the information release on that specific motor.. https://www.rocketryforum.com/showthread.php?t=36532

 

[LarryC]

Thanks for the reply.

Bob Korman was kind enough to steer me to the retest of the G40 by NAR. (Google r110 in r.m.r.) I found no official notice of redesign or tetesting of the F25. Perhaps you can provide a reference?

In the case of the F25 and G40, my impulses are close to the ones that Aerotech quotes, and the G40 is almost identical to the new NAR impulse. Still, if references are not freely available, that does little good. If I may say so, it is hard to know what we are buying anymore.

I have also been analyzing many data sets from the ARLISS project. I have yet to see a flight impulse of the M1419 motor in many data sets that has an impulse quite like the ones quoted. TRA lists the impulse as 7755 ns and Aerotech lists it as 7680 ns. The impulses from flight data are all on the order of 6800 ns. I'm relying, of course, on other people's launch weights, but why should they be all short and all around the same value? Has this motor been redesigned as well?

(Again, off-vertical trajectories analyzed as vertical ones would slightly over-estimate impulse; it doesn't grossly under-estimate impulse. That old rationalization doesn't wash.)

Forgive my skepticism. I'm looking at data and I'm listening to people. I'm trying to reconcile. Made some progress with the G40 and F25. How about the M1419?

Thanks again,
-Larry C

 

[metalwizard]

I don't know for sure about the M1419, but i don't believe its been recertified latley.. but I'll look into that for you.

 

[LarryC]

Much obliged!

 

[Art Upton]

..
I have also been analyzing many data sets from the ARLISS project. I have yet to see a flight impulse of the M1419 motor in many data sets that has an impulse quite like the ones quoted. TRA lists the impulse as 7755 ns and Aerotech lists it as 7680 ns. The impulses from flight data are all on the order of 6800 ns. I'm relying, of course, on other people's launch weights, but why should they be all short and all around the same value? Has this motor been redesigned as well?...
-Larry C

One possibility is they all could have all been weighed on the same non-calibrated scale?

Being they were all flown at the same site, this is a variable that needs to be taken into account.

Just Curious, are you working on a project to add motor sims to wRasp or digitrak or something ?

 

[bobkrech]

"OK, I'm still flummoxed. The Aerotech catalog listed on their website is 2005-2006. The test date for the G40 listed on www.thrustcurve.org is 30 May 2006 and has a total impulse of 113 ns. Once again, the catalog lists the impulse as 100 ns, and that is most consistent with my flight impulses. (Actually, I was glib in my previous post. I'm getting an average of 96 ns over eight flights I scrounged from my records.)"

I'm going to guess that you didn't actually read the instruction sheets that were provided with each motor. All motors are required to be shipped with an instruction sheet containing a thrust curve and the actual motor parameters. If you read them, what did the instruction sheets list as the total impulse for those motors?

"Are the motors that our national organizations test sampled from the same population as that of the motors we fly?"

They are supposed to be. You should also be aware that until NFPA 1125-2007 was issued, motor manufacturers could modify the total impulse by a fair amount of their motors if the average thrust did not change significantly without recertifying the motor. The G80 was an example of this as the old G80 had a total impulse of 120 NS the new design had only 100 NS total impulse. For a time, the 120 NS thrust curve was the only one that was certification tested and the new 100 NS version was not tested by S&T until the recertification date. The proper manufacturer's thrust curve and motor parameters were however included with the 100 NS motor instructions, so it really is important to read the instructions and check to see that you really know what motor you are flying.

"Does anyone here test off-the-shelf motors? That is best evidence in this matter."

Yes. The manufacturer is required to test 1% of his motors continuously to insure that they comply with his certification values.

Bob

 

[LarryC]

Art,

You may have a point, and I'll go you one better. There are a lot of people out there who take their launch weights from rocket design programs, rather than from scales. Although these estimates are very useful in hypothetical contexts, they do not substitute for measurements before launch.

I don't know it this pertains in this case. Just chalking up another possiblility.

Thanks and Regards,
-Larry C

 

[cjl]

I was just recently performing some similar calculations on my Deuce, and it appears that the K530's that it had were significantly underperforming too. I can't put too much faith in this data, but I will recalculate the coefficient of drag as well as reweigh it at the next flight and see if the data still supports a lower-than-advertised performance. Assuming that the measurements were correct, the data seems to support the K530's being 82% J461's instead of the advertised 1412Ns K529's (10% K's)

I'll have to make more measurements and such before this is certain though...

Here is the thrust curve I measured:


And here is the rated thrust curve:

 

[n3tjm]

Don't forget that since the Duece's motors are canted, you are going to loose some of the motors performance. The K530 is a small K, firing it in a canted fashion will yeild a J performance out of it.

 

[jraice]

I believe manf's can post the best results though, or the most optimal ones...

And also I believe its ok to be above or below (its usually below though) the certified impulse by 20 percent...

 

[shockwaveriderz]

I believe manf's can post the best results though, or the most optimal ones...

And also I believe its ok to be above or below (its usually below though) the certified impulse by 20 percent...

close but no cigar!

from nfp 1125:

(a) Standard deviation of the total impulse data shall be
no greater than 6.7 percent of the mean measured
value

(c) Average thrust shall be within 20 percent (or 1 N,
whichever is greater) of the average thrust that is
computed by dividing the mean total impulse measured
during propellant burn time by the mean propellant
burn time.

terry dean
nar 16158

 

[Rocketjunkie]

Don't forget that since the Duece's motors are canted, you are going to loose some of the motors performance. The K530 is a small K, firing it in a canted fashion will yeild a J performance out of it.

His Deuce has the motors canted 9 degrees. This only reduces the effective thrust 1.3%. A stock Deuce has a 15 degree cant reducing thrust 3.4%.

 

[bobkrech]

I was just recently performing some similar calculations on my Deuce, and it appears that the K530's that it had were significantly underperforming too. I can't put too much faith in this data, but I will recalculate the coefficient of drag as well as reweigh it at the next flight and see if the data still supports a lower-than-advertised performance. Assuming that the measurements were correct, the data seems to support the K530's being 82% J461's instead of the advertised 1412Ns K529's (10% K's)

I'll have to make more measurements and such before this is certain though...
[/img][/url]

Don't forget to compensate for gravity as well.

F(t) = 32*m(t)*(a(t)+1)+1/2*Cd(v)*rho(h)*A*v(t)^2 where a is in units of G.

Bob

 

[cjl]

Don't forget that since the Duece's motors are canted, you are going to loose some of the motors performance. The K530 is a small K, firing it in a canted fashion will yeild a J performance out of it.

I took that into effect (and there is only roughly a 1% loss due to the cant on that rocket - the angle isn't that great)

I'll keep looking for a cause though...

EDIT: I reran the numbers with a slightly more accurate weight and a few corrections, and I'm now getting a 92% J485. A little closer, but still significantly low...