Hypersonic velocity measurement

The Rocketry Forum

Help Support The Rocketry Forum:

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

GabeG

New Member
Joined
Sep 2, 2024
Messages
4
Reaction score
2
Location
Los Angles
Hello all,
I am working with a college team to build a (hopefully) hypersonic rocket (M5+).
Our goal is to go fast, so we need to be able to prove it.

What are some ways we can measure velocity at these speeds?
We’re tossing around ideas of building a ‘Rayleigh’ pitot tube, and/or using the Taylor-McCall shock come equations to find differential pressure to calculate speed from that.

Has anyone accomplished velocity measurement at high speed? Anything above Mach 2 is helpful to us.
thanks!
 
Hello all,
I am working with a college team to build a (hopefully) hypersonic rocket (M5+).
Our goal is to go fast, so we need to be able to prove it.

What are some ways we can measure velocity at these speeds?
We’re tossing around ideas of building a ‘Rayleigh’ pitot tube, and/or using the Taylor-McCall shock come equations to find differential pressure to calculate speed from that.

Has anyone accomplished velocity measurement at high speed? Anything above Mach 2 is helpful to us.
thanks!

I'm no Aerospace engineer, but modern accelerometers are probably going to be more accurate in determining actual speed than using barometric differential pressure measurements.

Also, mach 5+ is very hard to achieve with commercial motors and metal hardware. It's a mass fraction problem.

I hope for the best with your project, I'm working on a project for next year that is a small 54mm to 38mm two-stage rocket using the fastest burning commercial motors that will fit, and I'll be lucky to break mach 3 with it.
 
I tend to agree with Brian. John Krell might have some high fidelity suggestions for accelerometer integration, but for my 2c: I've been using the ADXL357 accelerometer lately and have been pretty damn impressed with my integration results from it. You don't need to sample it at crazy speeds because it internally does all this decimation magic for you. I only sample mine at something like 40Hz IIRC and even my insanely (intentionally) unstable hybrids provide a very close displacement agreement (to apogee) with baro and GNSS.

TP
 
Last edited:
I'm no Aerospace engineer, but modern accelerometers are probably going to be more accurate in determining actual speed than using barometric differential pressure measurements.

Also, mach 5+ is very hard to achieve with commercial motors and metal hardware. It's a mass fraction problem.

I hope for the best with your project, I'm working on a project for next year that is a small 54mm to 38mm two-stage rocket using the fastest burning commercial motors that will fit, and I'll be lucky to break mach 3 with it.
Hey thanks for the input. Good point about the accelerometers, that’s definitely going to be one of our measurement systems.

and yes, you’re very much correct about the difficulty of reaching such speeds with commercial motors… so we’re making our own very high energy propellant 🙃

Best of luck with your project as well, sounds exciting!
 
I tend to agree with Brian. John Krell might have some high fidelity suggestions for accelerometer integration, but for my 2c: I've been using the ADXL357 accelerometer lately and have been pretty damn impressed with my integration results from it. You don't need to sample it at crazy speeds because it internally does all this decimation magic for you. I only sample mine at something like 40Hz IIRC and even my insanely (intentionally) unstable hybrids provide a very close displacement agreement (to apogee) with baro and GNSS.

TP
That’s good to know! I’ll look into the Krell accelerometers. I was looking at the ADXL357 today, seems like a good idea to have one onboard, considering how small and light they are.

Thanks!
 
Transmit a fixed frequency and measure the Doppler shift? SDR might be useful for the receive end.
how much shift would there be? you would need some very precise and therefore expensive instruments for that.
so we’re making our own very high energy propellant 🙃
ooohhh looks like fun!
 
Sounds like a great project.
Please keep us updated on your success.

Where are you planning to launch? FAR?
 
how much shift would there be
Doppler equation

https://byjus.com/physics/doppler-effect/#doppler-effect-formula

A licensed amateur could transmit a marker tone on say 1296.303 MHz. use 300000km/sec for Vp.

With a GPS disciplined oscillator the receiver frequency would be within 10**-9 sec, 1 ppb, 1hz at 1296MHz. So, more than good enough.

The transmitter frequency stability is a different problem. Most xtal oscillators don't enjoy high G forces, and I doubt the GPS DO receiver will keep lock at mach 1 and faster.

Maybe someone has a better idea for the xmitter.
 
Doppler equation

https://byjus.com/physics/doppler-effect/#doppler-effect-formula

A licensed amateur could transmit a marker tone on say 1296.303 MHz. use 300000km/sec for Vp.

With a GPS disciplined oscillator the receiver frequency would be within 10**-9 sec, 1 ppb, 1hz at 1296MHz. So, more than good enough.

The transmitter frequency stability is a different problem. Most xtal oscillators don't enjoy high G forces, and I doubt the GPS DO receiver will keep lock at mach 1 and faster.

Maybe someone has a better idea for the xmitter.

I think you would need an ITAR controlled device to TX hold over Mach 2, certainly 3
 
The oscillator will experience drift under boost, due to acceleration. After burnout the forces should be significantly reduced. Try using a mems-based oscillator or other suitable clock that is more stable.
 
Last edited:
The oscillator will experience drift under boost, due to acceleration. After burnout the forces should be significantly reduced. Try using a mems-based oscillator or other suitable clock that is more stable.
Yes, that's what we said

How about this:

Instead of carrying a stable oscillator on board, use the one on the ground.

Fly a wide transponder that echos the time coded pulses from the ground station. 2 way trip for the reference signal, 2x the Doppler effect. I think?

Time to go scribble some arithmetic.
 
Pulse-doppler is your answer for those velocities if expecting any type of actual resolution.
 
Another major issue you will have to work through is heat. Objects travelling at high Mach numbers get very hot very quickly. You will need to build your rocket.

Also, on the topic of heat, your rocket will likely be surrounded by a plasma sheath, which will inhibit transmission from any onboard telemetry devices. To get around this, you should consider a ground-based radar. With a little persistence, a DoD facility like Edwards AFM or Pt. Mugu test center may assist on a non-interference basis. It never hurts to ask.
 
Another major issue you will have to work through is heat. Objects travelling at high Mach numbers get very hot very quickly. You will need to build your rocket.

Also, on the topic of heat, your rocket will likely be surrounded by a plasma sheath, which will inhibit transmission from any onboard telemetry devices. To get around this, you should consider a ground-based radar. With a little persistence, a DoD facility like Edwards AFM or Pt. Mugu test center may assist on a non-interference basis. It never hurts to ask.
what if you had a drone with a high speed camera at the expected altitude (if its within 1 km it should possible to reach with a drone) you need a FFA waver anyway.
 
I tend to agree with Brian. John Krell might have some high fidelity suggestions for accelerometer integration, but for my 2c: I've been using the ADXL357 accelerometer lately and have been pretty damn impressed with my integration results from it. You don't need to sample it at crazy speeds because it internally does all this decimation magic for you. I only sample mine at something like 40Hz IIRC and even my insanely (intentionally) unstable hybrids provide a very close displacement agreement (to apogee) with baro and GNSS.

TP
The ADXL357 is an excellent accelerometer. Very low noise with full range clock frequencies of 3.4 MHz for I2C and 10MHz for SPI. The Interpolation filter is a nice feature for higher data resolution.(Edit: at slower sampling speeds.) I've not tested this feature since I collect data in the 500Hz - 1000Hz range.

Single integration of high speed accelerometer data for velocity determination would produce better accuracy and be easier than a pitot tube or doppler shift. The time available at peak velocity will be in single digit milliseconds. The higher the sampling speed the better the accuracy and the greater the chance of capturing the peak velocity event.

Air friction heating will be an issue at velocities >Mach 3. I have yet to break Mach 2. I was the propellant insulation advisor for the USCRPL students on Traveler 3 & 4. Traveler 4 hit Mach 5.1 and came back looking like it had spent an hour in a fire.
 
Last edited:
The ADXL357 is an excellent accelerometer. Very low noise with full range clock frequencies of 3.4 MHz for I2C and 10MHz for SPI. The Interpolation filter is a nice feature for higher data resolution.(Edit: at slower sampling speeds.) I've not tested this feature since I collect data in the 500Hz - 1000Hz range.

Single integration of high speed accelerometer data for velocity determination would produce better accuracy and be easier than a pitot tube or doppler shift. The time available at peak velocity will be in single digit milliseconds. The higher the sampling speed the better the accuracy and the greater the chance of capturing the peak velocity event.

Air friction heating will be an issue at velocities >Mach 3. I have yet to break Mach 2. I was the propellant insulation advisor for the USCRPL students on Traveler 3 & 4. Traveler 4 hit Mach 5.1 and came back looking like it had spent an hour in a fire.
I tend to agree with Brian. John Krell might have some high fidelity suggestions for accelerometer integration, but for my 2c: I've been using the ADXL357 accelerometer lately and have been pretty damn impressed with my integration results from it. You don't need to sample it at crazy speeds because it internally does all this decimation magic for you. I only sample mine at something like 40Hz IIRC and even my insanely (intentionally) unstable hybrids provide a very close displacement agreement (to apogee) with baro and GNSS.

TP

I have not studied the sims for a Mach 5 hypersonic flight but would the +/- 40G max acceleration range of the ADXL357 be sufficient for such a flight ?

Analog Devices > ADXL357 Specs

Is there a version that can read higher ranges with the same precision and speed ?

EDIT: Maybe the Analog Devices > ADXL375 ?

-- kjh
 
I have not studied the sims for a Mach 5 hypersonic flight but would the +/- 40G max acceleration range of the ADXL357 be sufficient for such a flight ?

Analog Devices > ADXL357 Specs

Is there a version that can read higher ranges with the same precision and speed ?

-- kjh
Yeah, I forgot to mention that limitation. That could be a showstopper. Depends on flight profile. The optimum profile for altitude is generally a longer burn with lower thrust, but if your goal was to simply measure hypersonic velocity, then yeah, it's probably likely the rocket is getting one mighty kick up the beeeehind translating into a quite uncomfortable ride for the occupants.
Don't know if there's a comparable sensor with a higher G rating - I simply haven't looked.

TP
 
I have not studied the sims for a Mach 5 hypersonic flight but would the +/- 40G max acceleration range of the ADXL357 be sufficient for such a flight ?

Analog Devices > ADXL357 Specs

Is there a version that can read higher ranges with the same precision and speed ?

EDIT: Maybe the Analog Devices > ADXL375 ?

-- kjh
Mach 3 at 40g's in 2.5 sec.
Mach 5 at 40 g's in 4.2 sec.
It can be done at 40g acceleration. The ADXL 357 hard range limits are 10.24g's, 20.48g's, & 40.96g's.

Mach 3 at 100g's in 1.3 sec.
Mach 5 at 100g's in 1.9 sec.

I've added 250 milliseconds to the theoretical 100g times. A 100g acceleration does not happen instantaneously, unless you are firing a projectile out of a cannon. My instrumented micrograin rockets take approximately 250 milliseconds to overcome inertia and air drag to reach 100g acceleration.
 

Latest posts

Back
Top