Rocket acceleration?
Mike15
Posts: 109
I have completed a rocket payload project and have now launched and collected data.
Attached is the data from the two flights and data from a rocket simulation program.
The first tab shows the plotted data of each flight. The filtered reading is exactly that a "digitally filtered" reading of the acceleration.·The at rest reading is that of the output of the accelrometer when the rocket is at rest on the launch pad.
The second tab shows the raw data stored in·an external·eeprom. The first column is a reading from·the real time clock. the next is the accelrometer reading. The third the filtered reading and the at rest reading after that.
The·last tab shows the·raw flight data and raw simulator data compared and times matched. the plot shows how close the raeding of the accelrometer and the simulated acceleration are.
·I·ran a higher weight on the simulator program·Then what the rocket actully ended up weighing. Thats why I think the differance is there.
Now for the question with this data how can altitude be figured out?
Is it easiest to do it row by row? But then all the error incured by each row adds up and.
Is there a diffrent approach to calculating altitude?
Any feedback would be greatly appreciated.
If there is any intrest I will post code and schematic.
Attached is the data from the two flights and data from a rocket simulation program.
The first tab shows the plotted data of each flight. The filtered reading is exactly that a "digitally filtered" reading of the acceleration.·The at rest reading is that of the output of the accelrometer when the rocket is at rest on the launch pad.
The second tab shows the raw data stored in·an external·eeprom. The first column is a reading from·the real time clock. the next is the accelrometer reading. The third the filtered reading and the at rest reading after that.
The·last tab shows the·raw flight data and raw simulator data compared and times matched. the plot shows how close the raeding of the accelrometer and the simulated acceleration are.
·I·ran a higher weight on the simulator program·Then what the rocket actully ended up weighing. Thats why I think the differance is there.
Now for the question with this data how can altitude be figured out?
Is it easiest to do it row by row? But then all the error incured by each row adds up and.
Is there a diffrent approach to calculating altitude?
Any feedback would be greatly appreciated.
If there is any intrest I will post code and schematic.
Comments
I would be interested in seeing what you have , Mike. I am working on a similiar project, and while it is almost complete I am always interested in seeing how someone else approched similar problems.
Either way , thanks for sharing the above data!
Alohas,
Robert
1. What ADXL are you using? The ADXL150? I can't see where you've converted the filtered acceleration data into actual acceleration, and you either need to calibrate the sensor (best) or have the datasheet to convert the ADC reading into acceleration. Or did I miss it?
2. What kind of filter are you using? Kalman filtering?
3. I see in your program that you've got the capability to get pressure data too. But it looks like you just zeroed it out? If you have pressure AND acceleration data, you can estimate your altitude to accuracies exceeding the resolution of either sensor alone.
4. What exactly are you doing here:
sValue = (adc */ P50) + (sValue */ P50)
?
If you can atleast give me the PN for sensor (or calibration data), I'll calculate altitude in your spreadsheet and repost it.
Dave
as for the "filtering" thats this line of code
sValue = (adc */ P50) + (sValue */ P50)
it takes a percentige of the last reading and multiplies it with a percentage of the next reading
I used a G- 104 from aerotech rocketry
Airfarame 24.5 oz
Motor hard. 2.2 oz
As for the pressure sensor I am incorperating it and GPS to the next payload.
That's good info. But you also need to tell me the accelerometer's PN. A change of 1mv for one accelerometer might be equivalent to a 1g change, while for another accelerometer it might be a .001g change. Some accelerometers measure +/- 500g's, while others only measure +/- 2g's. So it's not a simple relationship (like 1mv = 1g for ALL accelerometers ever made). They vary from PN to PN and application to application. I see an ADXL on your schematic, but no info on the rest of the PN...
Also, if you're using a 12-bit ADC, I think 1bit is closer to 1.22mV.
If you can provide the PN, I can provide the rest.
(also what was your rocket diameter?)
Dave
Post Edited (achilles03) : 8/25/2006 2:15:23 PM GMT
The accelrometter is the ADXL78 from analog devices. If I remeber right 1g is equal to 27 counts?
Thanks for the attention to this matter
In a nutshell, it looks like you hit 325 to 330 meters for each flight.
It looks like your sensor was for AD22279, so it's 55mV/g. It also looks like you may have underguessed your drag coefficient a little. I redid your simulator, so you can change things in the spreadsheet and see how it affects your flight. I didn't worry about the filtered data, as your filter is mainly just smoothing.
It looks like you have a very good correlation between simulation and actual data. Good job. Feel free to ask any questions.
Dave
Post Edited (Mike15) : 8/30/2006 11:27:23 PM GMT