0
\$\begingroup\$

I am starting to gather information for a project I am excited about; my project is essentially a barbell that can detect the weight loaded on it. I will quickly summarize the function and restraints. So normally the barbel is made of stainless steel and has two areas where you can load the weight, each side can hold about 7 plates without bending or warping so about 630lbs of safe load.

What I am trying to do is be able to detect the weight placed on both of the loading areas but I'm unsure of an elegant way to do this.

It has been proposed pressure sensors but I don't see this as elegant as it potentially could be. This is where I need your help; My question is - what sensors are able to detect weight in a robust reliable way that would make sense for my project?

Specifications/ info of embedded system: A men's Olympic bar is a metal bar that is 2.2 metres (7.2 ft) long and weighs 20 kilograms (44 lb). The outer ends are 50 millimetres (2.0 in) in diameter, while the grip section is 28 millimetres (1.1 in) in diameter, and 1.31 metres (4.3 ft) in length.

Electronics involved: stm32** chip with Bluetooth and accelerometers.

Requirments

  • able to detect weight from 5 lbs - 315 lbs
  • robust and easy calibration cycle
  • will not interfere with the loading of weights and de-loading process ( bar must be able to fit inside an Olympic barbell)

What we have tried As mentioned we have brainstormed a few ways to potentially embed this but most of our ideas involve a spring system. Although we have been exploring a fiber optic strain system but are unable to dig up info on this or how we could implement one or even if it would be reliable in our project.

\$\endgroup\$
5
  • \$\begingroup\$ It absolutely has to be in the bar and not in the rack (where the bar rests before the athlete lifts it)? \$\endgroup\$ – The Photon Mar 8 '20 at 17:25
  • \$\begingroup\$ Yes so real word example I load my barbell and pick it up and it detects the weight through the sensors encased within the barbell. Reason for this is because we may not always be working in a rack and the accelerometers hold very important info in the context of athletic performance and we can do all sorts of cool stuff with this. Otherwise, I would just put the scale on the resting arm that you're referring to. \$\endgroup\$ – Invic18 Mar 8 '20 at 17:28
  • 2
    \$\begingroup\$ alternatively, consider putting identification smarts into the plates? \$\endgroup\$ – vicatcu Mar 8 '20 at 17:56
  • \$\begingroup\$ Rather get a nice large weigh platform for the athlete to stand on and you get additional information about the force profile during the lift and the unload. You can track which athlete is doing the lift by their personal mass (unless you have two athletes). \$\endgroup\$ – KalleMP Mar 8 '20 at 20:29
  • \$\begingroup\$ Another problem: where would you place the Bluetooth antennas? Placing a 2.4GHz chip antenna inside what's nearly a Faraday's cage simply won't work. And then... would it be fine mechanically to drill a hole in such a bar? Sounds rather safety-critical... and then you'll probably not want an antenna peeking out somewhere anyhow. \$\endgroup\$ – Lundin Mar 10 '20 at 10:44
3
\$\begingroup\$

The normal way to do this is to use conventional strain gauges bonded to the bar. These are thin and can be epoxied to the bar. You will probably need a bridge arrangement to detect the strain, and it will be even better if you can put one on the bottom of the bar and the other on the top.

\$\endgroup\$
4
  • \$\begingroup\$ And possibly on both sides. \$\endgroup\$ – Solar Mike Mar 8 '20 at 17:40
  • \$\begingroup\$ This looks very interesting, I will read up on this more this afternoon and if appropriate ask some questions related to this topic, thank you. \$\endgroup\$ – Invic18 Mar 8 '20 at 17:40
  • \$\begingroup\$ @Elliot, that was my initial thought too but wouldn't the strain depend on grip position? Two hands close to centre of the bar would induce much more bending moment and strain than two hands further apart, would they not? Plus we've got all the problems due to rotation of the bar so the strain gauge could be neutral if at the side and in compression if underneath. \$\endgroup\$ – Transistor Mar 8 '20 at 17:49
  • \$\begingroup\$ @Transistor Yes, since I'm not a gym rat I guess those things didn't occur to me. It's not an easy problem to solve, and I agree that it would be much simpler to weigh the thing before it is lifted. \$\endgroup\$ – Elliot Alderson Mar 8 '20 at 18:58
0
\$\begingroup\$

Since the weight will not change during the lift - unless weights fall off - it makes much more sense to weigh before or after the lift and just monitor position and/or acceleration during the lift.

The strain gauge approach, I suspect, will vary too much with grip position? Two hands close to centre of the bar would induce much more bending moment and strain than two hands further apart, would they not?

Strain gauges effectively measure the elongation of one or more surfaces of the device being monitored. This is usually straightforward if the loading is from a consistent direction such as on a weigh scales. On your bar you've got all the problems due to rotation of the bar so the strain gauge designed to be in tension on top of the bar could be neutral if at the side and in compression if underneath. I suspect the project will get less and less exciting as you try to figure out how to solve these problems.

Why not just key in the weight by adding up the weights on the bar!

\$\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.