HCSR-04

Can we use the ultrasonic sensor to measure acceleration?

Yes*

* under some circumstances

Probably possible but not easy/accurate/universally applicable. Use an accellerometer instead.

Thanks for your replies
Actually I need a sensor which can measure acceleration without being installed on the moving object. If some body can suggest something better it will be helpful.

Draw a picture how you want to place the sensor and how your moving object will move. You probably want some kind of a radar that the police use to measure speed of cars. If the thing is fast enough, you can calculate the acceleration from the consecutive measurements.

...and if the thing is too fast, you may only be able to make one or two measurements - these things don't have a great range, particularly if the target is small.

Depends on whether "the thing" refers to the moving object, or the device that takes the measurements.

Ultrasound has a range of 3-5 meters, and can do a couple dozen measurements in a second (depending on a.o. the actual distance covered). If those specs are good enough for the project, then it may work. Three measurements are the bare minimum to get an estimation of acceleration of the approaching object.

Johan_Ha:
Draw a picture how you want to place the sensor and how your moving object will move. You probably want some kind of a radar that the police use to measure speed of cars. If the thing is fast enough, you can calculate the acceleration from the consecutive measurements.

I want to measure acceleration of linearly moving aluminium disc with area of 20.26 cm^2 . The disc will move a maximum of 2cm distance in less than 1 second.

If your test area is not larger than 2 cm and you know exactly the path of the object and the object is not rotating or thrembling chaotically, it should be easy to set up a bunch of tiny light sensors of some kind to register the object passing by.

Johan_Ha:
If your test area is not larger than 2 cm and you know exactly the path of the object and the object is not rotating or thrembling chaotically, it should be easy to set up a bunch of tiny light sensors of some kind to register the object passing by.

Please suggest some light sensor models and a little detail about their arrangement.

Just google and read their datasheets. And use your creative mind. You are at the point, where you have to decide whether you want someone to do the job for you or you should do it yourself. With good luck someone has done something similar. But don't depend on it.

Ahmad-Yar:
I want to measure acceleration of linearly moving aluminium disc with area of 20.26 cm^2 . The disc will move a maximum of 2cm distance in less than 1 second.

So you want to measure the acceleration over a distance of 2 cm?

Then ultrasound is out as 2 cm is pretty much within error margins. So you won't be able to measure anything.

For light sensors: the TSL2561 is pretty tiny, you can build a PCB (yes, talking about a custom PCB here) with about 5 of them over a distance of 2 cm. Or if you stagger them, 10-15. That should give you a neat resolution.

There are also IR proximity sensors that can give fast and accurate distance within such a range. A.o. Adafruit offers such sensors. Sparkfun probably as well.

But a sketch and better description of the object and what's actually going on there would be very helpful.

If an object movs 20 mm and it starts from zero velocity, it really is a trivial thing. One measures the time it takes from 0 mm to 20 mm. Perhaps it's Arduino which triggers the event. Then you need one sensor for the 20 mm point.

But something says me you haven't described everything for us.

I would add a second sensor that's very close to the start to help measure the starting time of the motion. This works only if the acceleration is constant.

wvmarle:
So you want to measure the acceleration over a distance of 2 cm?

Then ultrasound is out as 2 cm is pretty much within error margins. So you won't be able to measure anything.

For light sensors: the TSL2561 is pretty tiny, you can build a PCB (yes, talking about a custom PCB here) with about 5 of them over a distance of 2 cm. Or if you stagger them, 10-15. That should give you a neat resolution.

There are also IR proximity sensors that can give fast and accurate distance within such a range. A.o. Adafruit offers such sensors. Sparkfun probably as well.

But a sketch and better description of the object and what's actually going on there would be very helpful.

I am making High Voltage Electrostatic Voltmeter. Electrostatic forces are actually very weak but at high voltages become significant. I am using two discs (One being fixed and other movable) to measure high voltages above 20kv. When high voltages are applied the movable disc is attracted towards the fixed disc with speed that 2cm distance is covered in less then a second. As i increase voltage levels the speed increases decreasing the attraction time. So, i need a system to get precise readings from the assembly.
Acceleration will not remain constant according to my calculations, it will increase.
I hope this explanation will work.

Sounds like optical is the way to go. It's got the precision and speed you need. Two sensors (one right at the beginning to record the start of the movement) and one at the end is all you need. The sensors would be a bit closer together than 2 cm of course, but that shouldn't affect the method itself.