Measuring the force of a collision

Hey! I've been trying to implement a couple of sensors on a curling stone using The Things Network to send data. The ones I've been thinking of attaching to the stone as of now are:

  • 3-axis gyroscope and accelerometer (for friction against ice)
  • Temperature and humidity sensor (for surface data)
  • Capacitive sensor (to detect the release)

I'm currently clueless as of how to implement some sort of collision detector as well as measuring how hard two stones collided. Is it possible using the accelerometer, and if so, how?

Thank you! :slight_smile:

Kolission is simply detected my acceleration - as is the force with which 2 devices collide.

3-axis gyroscope and accelerometer (for friction against ice)

How do you get the friction against the ice by these sensors?

I'm currently clueless as of how to implement some sort of collision detector as well as measuring how hard two stones collided.

Both can be done by the accelerometer. A collision is a negative acceleration (almost immediate slow down). During the sliding face the acceleration will be almost zero but as soon as you hit another stone you get a quite high acceleration. The value of that acceleration shows you how hard you hit the stone.

Measuring a collision force requires an accelerometer that has high frequency response* and you will need a high sample rate in order to hope to catch peaks.

  • accelerometers made for measuring shocks, usually piezoelectric accelerometers.

pylon:
How do you get the friction against the ice by these sensors?

It's not exactly accurate, but the idea is that friction against the stone will show up as acceleration in the Z-axis. By using the Serial Monitoring and Plotting I'd get an idea of the roughness and friction between the stone and the ice sheet.

EDIT: In the case of using The Things Uno, it doesn't seem like it'll benefit me in this situation as I'll need a high sample rate as @groundFungus said. I'll be hoping to see how the Arduino Uno WIFI will work.

but the idea is that friction against the stone will show up as acceleration in the Z-axis

Is that an acceleration into the ice?
If so then no.

Remember acceleration is the rate of change of speed, not speed.

By measuring the acceleration (change in speed) parallel to the ice surface, you can estimate the friction coefficient of the ice.

Consumer grade accelerometers are noisy and not very sensitive, so it isn't clear whether you could reliably measure such small accelerations as in gliding over ice.

Certainly, the ones used to measure acceleration during collisions (tens or hundreds times g) will not work for measuring friction coefficients. Let us know what your experiments reveal.

Hi,
I think you need to realise that the curling stone does not necessarily remain in the same attitude as it moves down the ice.
The stone will up to a point slowly rotate, this means your accel sensor will see a rotating accel/decel force,
A sudden force of impact will be from any direction parallel to the course surface.

Tom.... :slight_smile:

I think you are missing the fact that you will need to do similar measurements on both stones and add the forces together to get the real force. The inertia of the moving stone will be used to move the second stone.

This assumes the mass of both stones is identical.

Paul

F = m * a
Don't forget the mass..

add the forces together to get the real force

?
Newton would disagree.

You could also mount a video camera (controlled via RPi, not Arduino) on the ceiling above each end of the sheet, and then post-process the images to have fun with calculations related to impulse-momentum, conservation of energy, and coefficient of restitution.

I had a similar requirement and instead of fine tuning the software code with an accelerometer, I used this sensor specifically designed to detect collisions. Grove Collision Detection Sensor for Arduino RaspberryPi 3.3-5V High Sensitivity | eBay

If you put a piezo disk in the top center of the stone you can measure the shock of the impact.

For those who measure acceleration, don't forget to factor in the elasticity of your colliding objects.

Back in the day we used time lapsed photos taken with strobe lights to see where billiard balls went and how fast. A fully elastic collision will be over very fast and hope you catch the big change as it happens with accelerometer.

If you could put a clicker on top of the stone and speakers around, moving the stone will doppler the click speed at the speakers and find result velocity which IMO is a more reliable indicator than momentary acceleration reads.

PS, just checked, wind would affect a sound-based sensing.