0
$\begingroup$

I am currently attempting to solve the following problem. Say that I have an object that can expand and contract, given some external force. What I want to know is how much it has expanded, or contracted, over time. ( the diameter difference ) In addition, while this expansion or contraction is occurring, the object is free to move around. ( Example, object rolling around on the floor while wearing these sensors on itself )

I am considering of solving this problem by using two sensors, where one lies at the front of the object and one on its complete opposite ( back of the object ). An example of this can be seen below:

enter image description here

When the object expands, here is the size difference:

enter image description here

Obviously, the diameter of the second object is substantially bigger, and it is something I'd like to record as it increases / decreases.

Are there some good sensors that I can use to record this change over time? ( Note, sensors that can attached to a Raspberry pi board, or wirelessly connected to, that is ). I'd also like to stay away from sensors that would easily be lost if movement of the object was introduced, or require cords. ( i.e Capacitance belts, conductive yarns, etc. as they can easily give false readings of "contraction" and "expansion" given that the movement is too large. ex: Capacitative sensor gets stuck on the side of the object, and gives enormous diameter readings )

$\endgroup$

2 Answers 2

1
$\begingroup$

Sounds like you're looking for a strain gauge. They are quite common when analyzing stress on materials in specific directions.

It sticks to a surface and can record the change in length in one or more axis.

Since you're only concerned about diameter change, you would only need one.

Might take some playing to configure it with a Raspberry Pi, but it should be relatively easy to find one for your application.

$\endgroup$
4
  • $\begingroup$ can you think of any other sensors that are wire free? Alike the picture I have shown, where one component of the sensor is on the front, and the other on the back? $\endgroup$
    – angryip
    Aug 1, 2017 at 19:28
  • $\begingroup$ @angryip Do the wires actually hinder the operation if they are well routed? When talking wireless sensors, it's getting a bit out of my knowledge base; but I assume besides a really high additional cost, whatever is being used as a transmitter may be pretty large/cumbersome. You would also need to make sure the signal isn't distorted. There seem to exist passive RFID based sensors; but the associated measuring devices could be pretty annoying to couple with a Raspberry Pi. This also all depends on the accuracy and cost that you are looking for. $\endgroup$
    – JMac
    Aug 1, 2017 at 20:59
  • $\begingroup$ Wires would not cause issue, as long as they are used primarily for data transfer. If they are part of the reading, somewhat like a capacitative sensor, they will cause an issue. && I'd need something small, but the price itself does not matter. I'd like a precision of millimeters, but am willing to settle for centimeters if must be. $\endgroup$
    – angryip
    Aug 1, 2017 at 21:12
  • $\begingroup$ Keep in mind that depending on the requisite precision and the object's stiffness (product of its cross-sectional moment of inertia and the material's Young's Modulus), the deformation due to the object's own self-weight when rolling around may need to be taken into consideration. $\endgroup$
    – Wasabi
    Aug 2, 2017 at 3:48
0
$\begingroup$

A strain gauge is the simplest answer. You affix it to the object, and when it deforms, there is a small change in the resistance of the strain gauge which can be read using a circuit such as a wheatstone bridge. If this thing is rolling around, I suppose using computer vision to try to see the deformation is out of the question.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.