realistic numbers

rocket sok
Wildman Rocketry
SilicdyneTRF
The Rocketry Forum

Help Support The Rocketry Forum:

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

watermelonman

Well-Known Member
Joined
Sep 22, 2014
Messages
2,597
Reaction score
10
Yesterday I launched my Punisher on a Cesaroni L935, and my goodness, that was simply the most awesome flight I have ever put together.

I have plenty more to say about it and have footage to share, but I am a little curious about the numbers coming back from my Eggtimer.

Open Rocket said to expect 14100 feet and about 1300mph. My readings from the flight? 9800 feet and 3000mph. First off, this seems extremely low for this motor. I hear about people busting 15000 feet on lessor motors with this rocket. It did have a serious wiggle while trans sonic, and it took a fairly significant arc so I am sure I lost altitude with this. I tend to get about 75% of simulated altitude on this rocket so perhaps with that arc the altitude is believable.

3000mph, though? What the heck is that. I mean, wow, awesome! But, is that even realistic? Maybe I need to take the two numbers together and figure that going far faster than expected meant more drag. Or maybe that number is simply an errant reading.
 
watermelonman said:
3000mph, though? What the heck is that. I mean, wow, awesome! But, is that even realistic? Maybe I need to take the two numbers together and figure that going far faster than expected meant more drag. Or maybe that number is simply an errant reading.
Click to expand...

3000 mph = ~mach 4

Not happening. Definitely an errant reading.
 
Yes, that is what I think. I wonder what range it was in, in reality. The altitude still seems quite low, but at least I have a plausible explanation.
 
What's the size and configuration of the avionics bay vent holes ??

Teddy
 
Great point; I bet the vents are messing with readings. Sadly there is not much choice on this model! I have 3x 3/16" holes spaced at thirds, exactly like CJ in his original Punisher build. Vertically I try to keep them between the fins, though with the drogue separation being friction it is not an exact science.
 
watermelonman said:
Yesterday I launched my Punisher on a Cesaroni L935, and my goodness, that was simply the most awesome flight I have ever put together.

I have plenty more to say about it and have footage to share, but I am a little curious about the numbers coming back from my Eggtimer.

Open Rocket said to expect 14100 feet and about 1300mph. My readings from the flight? 9800 feet and 3000mph. First off, this seems extremely low for this motor. I hear about people busting 15000 feet on lessor motors with this rocket. It did have a serious wiggle while trans sonic, and it took a fairly significant arc so I am sure I lost altitude with this. I tend to get about 75% of simulated altitude on this rocket so perhaps with that arc the altitude is believable.

3000mph, though? What the heck is that. I mean, wow, awesome! But, is that even realistic? Maybe I need to take the two numbers together and figure that going far faster than expected meant more drag. Or maybe that number is simply an errant reading.
Click to expand...


Those readings seem suspect.. I launched my bird right before you did on a CTI K500 redline (a 4 grain motor) and both of my altimeters said it went ~6500ft.... that is a 4" rocket which was much larger and heavier than yours.

it is suspect that in the same day/same conditions your rocket with the L935 went only 9800ft.. when i saw your rocket in the air it certainly Looked like it went more than 3k feet higher... sometimes hard to tell though... I think you said you recovered it a mile down the road correct? in the same wind conditions mine (under drogue) also landed in the same field about 1000 feet away. My descent was ~75fps under drogue. Just thinking that if your descent rate was roughly close - given how much further you drifted - could also be an indicator that you were much much higher...

just a few thoughts...
 
watermelonman said:
Great point; I bet the vents are messing with readings. Sadly there is not much choice on this model! I have 3x 3/16" holes spaced at thirds, exactly like CJ in his original Punisher build. Vertically I try to keep them between the fins, though with the drogue separation being friction it is not an exact science.
Click to expand...

That's it then.........
The anomaly you described would be well explained by a pressure wave from a bump of some sort
on the airframe above and inline ( or close ) to one of the vent holes..
That pressure wave will change shape and location it comes back in contact with the airframe as velocity increases...
Is there anyway there was a screw head,, a rivet head, or a shear pin head above and inline with a vent hole ??

Teddy
 
You should get Krech to answer this...
He would give you a definitive answer that you can work with and make a change to the rocket with...

Teddy
 
onebadhawk said:
You should get Krech to answer this...
He would give you a definitive answer that you can work with and make a change to the rocket with...

Teddy
Click to expand...

Or he will explain it using mathematical formulas and physics and the thread will die...:D
 
rharshberger said:
Or he will explain it using mathematical formulas and physics and the thread will die...:D
Click to expand...

Lol,,, lol,,, dat's pretty funny,,,,,
Nah,,, I can always screen what I need to know from his answers....
He just likes to speak from science,, as opposed to opinion...
I torture him all the time,,, I always get the answer I need to physically do what I have to do
even if it means reading some stuff I don't completely understand...
It's like talkin to Enstien,,,, lol.... Most didn't understand him either,,, lol...

Teddy
 
Hey,,,
Now my Avatar's broke.....
Get Krech to fix it...lol...

Teddy
 
Sometimes, altimeters just screw up.

I once had an accelerometer tell me an M1419 took a 45#, 6" airframe over 98K. Somehow, I don't think so.

If it fires the charges at the right times, that's what counts.
 
Last edited:
watermelonman said:
Yesterday I launched my Punisher on a Cesaroni L935, and my goodness, that was simply the most awesome flight I have ever put together.

I have plenty more to say about it and have footage to share, but I am a little curious about the numbers coming back from my Eggtimer.

Open Rocket said to expect 14100 feet and about 1300mph. My readings from the flight? 9800 feet and 3000mph. First off, this seems extremely low for this motor. I hear about people busting 15000 feet on lessor motors with this rocket. It did have a serious wiggle while trans sonic, and it took a fairly significant arc so I am sure I lost altitude with this. I tend to get about 75% of simulated altitude on this rocket so perhaps with that arc the altitude is believable.

3000mph, though? What the heck is that. I mean, wow, awesome! But, is that even realistic? Maybe I need to take the two numbers together and figure that going far faster than expected meant more drag. Or maybe that number is simply an errant reading.
Click to expand...

Could you attach a picture of the altitude and velocity graph from the eggtimer? It sounds like pressure spikes/drops around mach could have fooled the altimeter or something - if you look at the altitude plot it's very easy to tell if the 9,800' and 3,000 mph is completely anomalous or if it actually makes sense with a believable altitude plot.
 
watermelonman said:
Yesterday I launched my Punisher on a Cesaroni L935, and my goodness, that was simply the most awesome flight I have ever put together.

I have plenty more to say about it and have footage to share, but I am a little curious about the numbers coming back from my Eggtimer.

Open Rocket said to expect 14100 feet and about 1300mph. My readings from the flight? 9800 feet and 3000mph. First off, this seems extremely low for this motor. I hear about people busting 15000 feet on lessor motors with this rocket. It did have a serious wiggle while trans sonic, and it took a fairly significant arc so I am sure I lost altitude with this. I tend to get about 75% of simulated altitude on this rocket so perhaps with that arc the altitude is believable.

3000mph, though? What the heck is that. I mean, wow, awesome! But, is that even realistic? Maybe I need to take the two numbers together and figure that going far faster than expected meant more drag. Or maybe that number is simply an errant reading.
Click to expand...

Out of curiosity, did you weigh your rocket before launch to see if it compared with the weight used in the simulation?
A lot of folks don't.

-LarryC
 
Here is my Eggtimer data. The altitude looks reasonable, whereas I have not made heads or tails of the velocity yet.
View attachment tP2.txt

For kicks, and maybe a little analysis, here is a video of the flight -
[video=vimeo;131591813]https://vimeo.com/131591813[/video]
 
3000 mph. This is likely caused by a pressure transient on the barometric sensor.

This same thing could mess with altitude determination. Do you have an altimeter other than the engineer you could try on another flight? Curious to see how they would correlate.
 
Did you model the drag from the camera in your sim?
 
Cris Erving was also curious and took a look at my data. He came up with mach 1.6 after dropping outliers and smoothing. Apparently the trick was in mach transition occurring near motor burnout, making a double spike and making the data tough to interpret.

Thanks as usual, Cris!
 
watermelonman said:
Cris Erving was also curious and took a look at my data. He came up with mach 1.6 after dropping outliers and smoothing. Apparently the trick was in mach transition occurring near motor burnout, making a double spike and making the data tough to interpret.

Thanks as usual, Cris!
Click to expand...

Did that change the altitude determination as well?
 
spdntckt said:
Those readings seem suspect.. I launched my bird right before you did on a CTI K500 redline (a 4 grain motor) and both of my altimeters said it went ~6500ft.... that is a 4" rocket which was much larger and heavier than yours.

it is suspect that in the same day/same conditions your rocket with the L935 went only 9800ft.. when i saw your rocket in the air it certainly Looked like it went more than 3k feet higher... sometimes hard to tell though... I think you said you recovered it a mile down the road correct? in the same wind conditions mine (under drogue) also landed in the same field about 1000 feet away. My descent was ~75fps under drogue. Just thinking that if your descent rate was roughly close - given how much further you drifted - could also be an indicator that you were much much higher...

just a few thoughts...
Click to expand...

Definitely good thoughts. I think mine wound up so far away because of the arc. It never got all that slow, throughout the bulk of the flight. I think the altitude is realistic.

markkoelsch said:
Did that change the altitude determination as well?
Click to expand...

Sadly no.

timbucktoo said:
hey melonman, out of curiosity - how did you sim the camera drag?
Click to expand...

A simple big, goofy, single fin from a freeform set with extra drag. Not exactly precise, but I figured it was better than nothing. Of course I compared to the unaltered simulation as well.
 
cherokeej said:
Sometimes, altimeters just screw up.
Click to expand...

Right. Your simulation of velocity is probably more accurate than the altimeter.

I looked at your data. The velocity problem is simply the limitation of trying to differentiate barometric altitude into speed. It doesn't work very well.

In the plot, the velocity curve is simply just the numerical derivative of the altitude it is given. The altitude is all squirrely before t=5000 (real or not?), thus the derived velocity is even more crappy, and hence the strange peak values. The other columns in the file, FALT and FVELOC, are probably some kind of filtering, which help smooth the velocity spikes. However, the altitude needs little filtering after t=5000, so the max altitude of 9800 ft around t=22500 does not change.

The moral of the story is do not trust the peak values from the altimeter. You need to look at the time history of the data, massage it, fudge it, and interpret it correctly. The second moral of the story is that barometric altimeters are not numerically precise enough for velocity measurement. You need an accel unit for that. A $50 baro altimeter is really only good enough to measure apogee altitude and fire ejection charges within a couple seconds of when you expect them to do so. For those purposes, the baro altimeters work just fine.

Capture.png
 
Last edited:
Buckeye said:
Right. Your simulation of velocity is probably more accurate than the altimeter.

I looked at your data. The velocity problem is simply the limitation of trying to differentiate barometric altitude into speed. It doesn't work very well.

In the plot, the velocity curve is simply just the numerical derivative of the altitude it is given. The altitude is all squirrely before t=5000 (real or not?), thus the derived velocity is even more crappy, and hence the strange peak values. The other columns in the file, FALT and FVELOC, are probably some kind of filtering, which help smooth the velocity spikes. However, the altitude needs little filtering after t=5000, so the max altitude of 9800 ft around t=22500 does not change.

The moral of the story is do not trust the peak values from the altimeter. You need to look at the time history of the data, massage it, fudge it, and interpret it correctly. The second moral of the story is that barometric altimeters are not numerically precise enough for velocity measurement. You need an accel unit for that. A $50 baro altimeter is really only good enough to measure apogee altitude and fire ejection charges within a couple seconds of when you expect them to do so. For those purposes, the baro altimeters work just fine.
]
Click to expand...

And for those interested in even more math: To get velocity data from position data you have to use an integral function. The problem is it magnifies errors that exist in the original data.
However, obtaining velocity data from acceleration data requires you to do a differential function, which is the opposite of integrating. Therefore, this tends to reduce problems in the original data. All other things equal, this makes accelerometer data the preferred measurement.

But we also need to discuss timing issues. If the clock chip driving this data collection is significantly off, then we have velocities that will appear faster or slower than they really were. For example, a clock that runs 5% too fast will result in velocities that appear to be about 5% slower than they really were. Worse than that is a clock that drives data collection at inconsistent intervals. Instead of data being collected at 0.1, 0.2, 0.3, etc., data was collected at 0.1, 0.19, 0.33, etc. instead. While this should have no real discernible effect on altitude data, it royally screws up velocity data no matter what type of sensor you use.
 
Zeus-cat said:
And for those interested in even more math: To get velocity data from position data you have to use an integral function. The problem is it magnifies errors that exist in the original data.
However, obtaining velocity data from acceleration data requires you to do a differential function, which is the opposite of integrating. Therefore, this tends to reduce problems in the original data. All other things equal, this makes accelerometer data the preferred measurement.

But we also need to discuss timing issues. If the clock chip driving this data collection is significantly off, then we have velocities that will appear faster or slower than they really were. For example, a clock that runs 5% too fast will result in velocities that appear to be about 5% slower than they really were. Worse than that is a clock that drives data collection at inconsistent intervals. Instead of data being collected at 0.1, 0.2, 0.3, etc., data was collected at 0.1, 0.19, 0.33, etc. instead. While this should have no real discernible effect on altitude data, it royally screws up velocity data no matter what type of sensor you use.
Click to expand...

Nope, you have it backwards. Velocity is the derivative of position, and numerical differentiation is very inaccurate. Velocity is the integral of acceleration, and numerical integration is much more robust. So for speed purposes, accelerometers = integration = good.

I have never seen data from a baro altimeter with inconsistent intervals. I hope no altimeter would do that.
 
Buckeye said:
Right. Your simulation of velocity is probably more accurate than the altimeter.

I looked at your data. The velocity problem is simply the limitation of trying to differentiate barometric altitude into speed. It doesn't work very well.
Click to expand...

Thanks for taking the time!
 
Buckeye said:
Nope, you have it backwards. Velocity is the derivative of position, and numerical differentiation is very inaccurate. Velocity is the integral of acceleration, and numerical integration is much more robust. So for speed purposes, accelerometers = integration = good.

I have never seen data from a baro altimeter with inconsistent intervals. I hope no altimeter would do that.
Click to expand...

Oops, this always happens when I do rocket stuff when I'm really sick. You're right, v = dx/dt

Of course the last time I did rocket stuff when really sick was much worse; I lawndarted a high power rocket. I screwed up the ejection charges. OUCH.
 
Last edited:

Similar threads

Ron Soto
help with OR..?
Replies
4
Views
284
bobbyg23
bobbyg23
K
My Level 2 Attempt - a Diurnal Flight of "Nocturnal Missions"
Replies
5
Views
612
kjhambrick
K
B
Securing a payload bay?
Replies
3
Views
321
shanejohnson2002
shanejohnson2002
Adrian A
120 G flight yesterday
Replies
27
Views
2K
Spacedog49Krell
Spacedog49Krell
caheaton
quick backyard launch
Replies
1
Views
545
PDawg
P

Latest posts

Back
Top