measuring G force

[Gary Liming]

With the availability of low cost strain gauges and MEMS type vibration sensors now, has anyone seen any DIY devices that can be used to measure the amount of G shock that occurs on a component? For instance, if I want to know how many Gs a certain wiring technique can sustain before failure, I would make a test circuit, attach it to a sled like it would be in an avbay, and apply an increasing force until failure, etc. How can I measure the G shock that was applied each time? I have done a few projects that reads a strain gauges and reads an 6 DOF accelerometer, but I am not sure those would be accurate in the 6-15 G range. Anyone else tackled this?

 

[OverTheTop]

There are accelerometers out there that can read high accelerations. It shouldn't be too difficult to log the output with something like a Raspberry Pi.

An alternative is to use height to give a velocity and then decelerate it over a known distance to get an estimate of the decel.

 

[Gary Liming]

Besides the ADXL377, do you know of any solid state high g accelerometers? Seems no one has built one?

 

[heada]

Do you need more than the +/- 200G that the ADXL377 can handle?

 

[UhClem]

You are mixing two different things: force and shock. If you want to measure shock you need an appropriate sensor, signal conditioning, and data acquisition. I collected some acceleration data using a MPU9250 but its 4KSPS was inadequate for analyzing shock.

Endevco makes some really nice accels and the 2250 is one of my favorites.

 

[OverTheTop]

Shock is just deceleration of a known velocity over a controlled distance. Know the velocity and distance of deceleration and the decel value is calculable. This is how we shock test our spectrometers for regulatory compliance.

I forgot to mention in my earlier post that the industrial accelerometers are very expensive!

 

[Gary Liming]

I think shock and jerk are being confused here, probably my fault for referring to G shock in the OP instead of jerk. Jerk is the change in acceleration per unit time. Shock, although not well defined, is usually a discontinuous event, like an energetic impact with the ground. I don't see how shock would be otherwise applicable to rocket flight.
David, Thanks for the reference to the 2250.

 

[Gary Liming]

Aaron, its not the g range of the 377, its the 1KHz data rate.

 

[Nytrunner]

I don't see how shock would be otherwise applicable to rocket flight.

Deployment events can be quite shocking (I refuse to apologize)

 

[UhClem]

Shock is usually simulated in the lab using a half sine acceleration curve. Designed to produce a particular shock response spectrum.

But the only way to accurately produce a pyrotechnic shock response spectrum is with explosives.

 

[Gary Liming]

Well, I know that one ISO standard (1413 Horology) for testing shock on watches is defined as a 3 kg hammer suspended as a pendulum striking the watch at 9:00 and on its face, moving at 4.43 m/s, imparting about 30 joules, and still maintaining the watch's its accuracy. So at least one standard defines shock as a quantity of kinetic energy, which kind of makes sense.
My original post was really about seeing if anyone had ever rigged a test stand to measure the impact of mounting failures, etc.
Anyone? Bueller? Anyone?

 

[Nytrunner]

It would be interesting to see if a recovery ground test type event could be instrumented to measure the acceleration (from which you could perhaps get shock with some math)