6

I am a Btech. student from India. We have been given a project in which we have to design a micro-torsion testing device.

We are planning to clamp the specimen and rotate one end by fixed angles using a stepper motor while measuring the torque for given deflection. The shear modulus and the angle/torque at which the specimen fails has to be found out.

The most accurate way to do this seems to be using torque sensors but since our budget is small ( Rs. 10k - arnd 150 dollars) we cant use this. Another way we had in mind was to measure the current drawn by the motor and find out the corresponding torque. But this seems to be prone to errors. Can someone help me out with this ?

nekomatic
  • 1,490
  • 11
  • 15
Atharva Kulkarni
  • 69
  • 4
  • 1
    What range of torques do you have to measure, and how accurately? – Bruce Abbott Sep 30 '19 at 19:52
  • 2
    Approximately 0-150Nm. Can tolerate an error upto 20%. – Atharva Kulkarni Sep 30 '19 at 20:12
  • 3
    150 Nm isn't exactly a micro torque! But I hope you succeed with your project. :-) – Catalyst Sep 30 '19 at 20:25
  • 3
    Why don't you use a torque wrench? –  Sep 30 '19 at 21:34
  • @Catalyst: micro torsion does not mean necessarily micro torque (namely if the DUT is very rigid) – Curd Sep 30 '19 at 22:22
  • I'm voting to close this question as off-topic because it's not related to electrical engineering. It could be on-topic on physics.SE. – Dmitry Grigoryev Oct 01 '19 at 08:51
  • 2
    @Dmitry Grigoryev: it is very well related to electrical engeneering if the measuring process should be automated (which is obviously the case; "by an electric or electronic method") – Curd Oct 01 '19 at 12:09
  • @Curd why do you assume it must be automated? many electric or electronic devices are manual... – Solar Mike Oct 01 '19 at 13:58
  • 1
    @Solar Mike: if a machine is electric then it is already automated to some extend: force/torque does not have to be provided by manually pushing/turning/placing weights etc. but by simple turning a switch. In addition if the displacement/angle etc. is measured by an electrical decoder and thus can be loggeg automatically instead of measuring manually thats another degree of automation. – Curd Oct 01 '19 at 14:23
  • @Curd so manual switch... I was expecting you drop the specimen into a hopper and get an email 30 seconds later with the result... – Solar Mike Oct 01 '19 at 14:26
  • @Solar Mike: or **you** write the E-Mail which causes the machine to fetch the specimen... – Curd Oct 01 '19 at 14:36
  • @Curd, you're correct in the strict sense, but I strongly suspect that the test specimens a prudent instructor will supply won't be extremely stiff, both because of cost -- and safety when tested to and through yield -- especially by students. – Catalyst Oct 03 '19 at 14:59

5 Answers5

11

Fix a lever of a precise known length to the specimen and add calibrated masses. Measure the deflection with a dial gauge.

Perpendicularity might be a concern and could be addressed with geometry... However, the error that introduces is probably small compared to other sources. The sources of errors may well be worth checking. I know that for the dynamometers we used on engines, pumps etc there was no correction but the beam was usually corrected back to a given calibrated zero point.

Solar Mike
  • 6,319
  • 1
  • 12
  • 28
  • 1
    And most importantly - no need to have any motor or electrical engineering whatsoever.. – Eugene Sh. Sep 30 '19 at 19:52
  • How do I make sure that the force applied by the weights always remains perpendicular to the lever? – Atharva Kulkarni Sep 30 '19 at 20:11
  • 2
    @AtharvaKulkarni A pulley that carefully winds the mass up so the string doesn't overlap on the shaft? Do you need continuous rotation? – DKNguyen Sep 30 '19 at 20:15
  • What do you think of building a torque sensor using two load cells placed on the opposite sides of the diameter on a disk and a thin bar between them which will be connected to the specimen? The motor will drive the disk which through the load cells will rotate the bar and thus the specimen. The specimen will then exert torque on the bar which in turn will exert force on the two load cells and since we know the radius, we can find the torque applied by the specimen. – Atharva Kulkarni Oct 03 '19 at 13:25
4

I think measuring the motor current is not a bad idea.
I wouldn't use, however, a strong stepper motor directly but a weaker one in combination with a reduction gear. That way you get many rotations at the motor axis even if the angle of torsion at the specimen is only very small. So the current measurement can be averaged over several turns.
I assume that accuracy will then be by far enough for your application (<<20%).

Also measuring the torsion angle will be much easier if done at the motor axis (before the reductio gear); e.g. using an incremental angle encoder, or, if a stepper motor is used, simply by counting the steps.

Curd
  • 16,043
  • 34
  • 43
  • Note that, if using a stepper, you can't use motor current to measure torque without encoder feedback. – jms Oct 02 '19 at 17:52
2

Make a torsional pendulum and measure its frequency against a known intertia.

Orient the test piece vertically, clamp the top end, add a known disk inertia to the bottom end. Now you have a torsional pendulum. Twist the bottom disk a small angle and measure the frequency of torsional oscillation. Optical measurement is easy, and the sensors are inexpensive. Just add a contrasting dot to the disk if using a reflective sensor. Or a pair of holes near the disk's edge (for symmetry) if using a slotted optical sensor. Hope this helps!

Catalyst
  • 256
  • 2
  • 5
  • We had something similar in mind but this just gives the torsional constant and thus shear modulus of the material. We also need the angle/torque at which the specimen fails. – Atharva Kulkarni Sep 30 '19 at 20:29
  • In that case, you should really so state in your question. The lowest cost way to impose a pure torque is with equal tangential forces. If you keep the test piece vertical, use a disk as suggested above, add cord (or thin wire rope if available) and pulleys to each cord: then you can add equal weights to each cord, hanging down, outboard of each pulley to get a wide range of pure torques. You still need to measure the angle, but calipers will do in a pinch. (So I'm leaving you some homework if you want to measure that electronically.) Hope this keeps the cost down. – Catalyst Sep 30 '19 at 20:36
  • 1
    If you need to measure the angle, then I suggest replacing the disk with a bar, attached to the test piece at the bar's center. Then, with something to mark/remember the neutral position, the displacements at the bar's ends are easy to measure, and will give you the angle as you add weights to load the specimen more and more in torsion. Be careful -- 150 Nm is nontrivial when it lets go! – Catalyst Sep 30 '19 at 21:00
0

Fixing your sample to the end of a rod that will plasticaly deform within the range of torques you are interested in measuring, then affixing strain guages to the rod, would allow the torque to be measured electronically. The resistance of the strain guage(s) would have to be measured and calibrated against some known torques (which could be applied as suggested in other answers, with a beam and a known mass or force).

sugseg
  • 1
0

I think it might be easier to apply a known torque and measure the deflection. Applying a known torque can be as simple as using your stepper motor to control the position of a weight on a lever, and you can use, for example, optical interferometry to measure the deflection accurately.

Matt Timmermans
  • 1,358
  • 6
  • 7