3
\$\begingroup\$

I want to make my own load cell; how specific do the materials need to be?

For example, is it necessary to buy all the cleaning and gluing equipment shown in strain gauge installation videos? Or can I just scrub the surface, rinse & dry it clean, and superglue it on?

Additionally, instead of a solid metal bar, do strain gauges also work with 3D printed objects or a pipe?

And why do some of the commercial small straight bar strain gauges have weird holes in them? For example, sparkfun load cell has that hole in the middle.

\$\endgroup\$

3 Answers 3

9
\$\begingroup\$

I want to make my own load cell; how specific do the materials need to be?

How closely and reliably do you want your results to match the prototype, or from one instance to the next?

For example, is it necessary to buy all the cleaning and gluing equipment shown in strain gauge installation videos? Or can I just scrub the surface, rinse & dry it clean, and superglue it on?

See above. If you just want quick results, and you don't mind if your gizmo doesn't work for long or will be exactly the same next time -- go for it.

I would try to use the same type of glue that is shown (or is recommended by the strain gauge manufacturer). I.e., epoxy if they use epoxy, superglue if they use superglue, etc.. I would also bear in mind that consumer- or hobby-grade glues are usually designed to be inexpensive, to last a long time in poor storage conditions, and to stand up well to poor handling. Strength is compromised as a result.

If you want a commercial-grade load cell, you probably want to go through all the folderol. Personally, if I just wanted to try something, I'd do what you're doing -- but if I wanted it to last, then once I knew it worked I'd make another one "for real".

Additionally, instead of a solid metal bar, do strain gauges also work with 3D printed objects or a pipe?

Yes and no.

The "strain" that strain gauges measure is as defined in mechanical engineering texts -- it's the amount of distortion that the underlying bar, pipe, 3D printed thing or whatever undergoes in response to the stress put on it.

If you put a strain gauge onto a metal pipe it'll measure the strain on that metal pipe -- you'll need to translate that into whatever you're trying to actually measure.

(I'm assuming here you mean 3D printed plastic, not fancy pro 3D printed metal -- although, some of the things I mention may still apply to sintered metal vs. machined alloy).

If you put a strain gauge onto something made out of plastic, then you run into a bunch of potential problems. I don't know what all of them are, but I can think of three things right off the top:

  1. Plastic has a lot more hysteresis in its strain response than metal. Meaning, if you pull on it and let go, it'll be a lot less likely to spring back to where it was. This will cause you no end of grief if you need a reliable "return to zero" function
  2. Plastic has a much higher coefficient of thermal expansion, so you'll see more drift with temperature. Closely related to this, if the strain gauge is compensated for the coefficient of thermal expansion of a particular alloy, then you'll lose that compensation entirely by going to something as wildly dissimilar as plastic.
  3. Plastic -- at least the stuff one normally 3D prints with -- is a lot more elastic. If it's more elastic than the gauge itself, then instead of you pulling on the carrier and the gauge going along for the ride (and reporting on it), the gauge will be a structural member. This will both mess up your calculations and it may vastly increase the chances of your glue joint, or some glue joint on the gauge, breaking.

And why do some of the commercial small straight bar strain gauges have weird holes in them? For example,

Because the gauge measures strain. By making the members under the gauge thinner (because of the holes) the amount of stress per unit force is increased; this means that the strain per unit force in increased, and, hence, the whole gauge is more sensitive, without needing to go to an insanely small strain element.

\$\endgroup\$
5
\$\begingroup\$

The load cell pictured is a variation on the so-called "binocular" design. The holes provide areas of high strain to improve sensitivity. The strain gauge(s) themselves are mounted in such areas.

enter image description here

The COMSOL-generated image is from this paper which briefly touches on the math and shows how they were able to simulate the response accurately with finite-element analysis and a 3D model.

The overlapping hole design can be fabricated in aluminum with a saw, file and drill press so it's not very demanding. I would think something like PLA would be a very unsatisfactory material to use with conventional strain gauges if you're trying to actually measure something to high accuracy because of bad spring properties and other reasons. But for indication of impending failure or very rough measurements it might be useful.

There are special strain gauges that are intended for use on plastics. Here is a video that shows competent application of a gauge to a 3D printed beam using high-grade cyanoacrylate adhesive and catalyst. You can see when he checks it out, the pronounced creep characteristics.

\$\endgroup\$
4
\$\begingroup\$

There are other reasons for the odd holes and cut-outs in a typical bar load cell. While the thinned areas do concentrate the strain, the double hole pattern (bone shape) allows the bar to bend in a very specific way. With a mass placed at one end of the bar the inner areas bend but the opposite ends remain nearly parallel. A solid shape would just bend in a simple curve with the far end angled downward, this would result in all of the stretching on the top surface and all of the compression at the bottom surface. The special configuration creates 4 locations along the top & bottom of the bar that give opposing positive and negative strain. Placing strain gauges at these 4 points easily creates a Wheatstone bridge system. Even a half Wheatstone bridge could be made on a single surface if needed. See the link showing an exaggerated view of the bending. https://diyodemag.com/education/weight_measurement_using_load_cells_arduino

The reason for using the recommended glue is that the material needs to match up with the characteristics of the strain gauge. If the glue is too stiff it may hinder the deformation of the gauge material, if the glue is too weak it may tear under the stress and detached from the gauge or the bar material. Even the long term fatigue characteristics of the glue are important as the gauge and glue may be stressed and relaxed hundreds or thousands of times. If the attachment changes over time or frequent use the measuring system will go out of calibration, be non-repeatable, or have a drifting output.

In some cases even the placement of the wires connecting the gauges can affect the measured output of the load cell system.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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