Algodoo

thanks, it still confuses me but it provides a interactive explanation!

Steps are listed below. Report any steps you don't understand. Anyone who understands that step can respond.

What was done:
  1. Turn off clouds.
  2. Turn off gravity.
  3. Read entire article.
  4. Set motorTorque for all axles to +inf.
  5. Add mechanism to pull lowest link.
  6. Select all links and turn off all collision layers.
  7. Set bendConstant for thinner sections to 10.
  8. Set sim.frequency to 240.
  9. Set air multiplyer to 100 (for damping).
  10. Add code to puller hinge to track force and deflection.
  11. Add display box to show force and deflection.
  12. Add circle to track force and deflection.
  13. Add plot to track circle.

What still needs to be done:
  14. Tune each of the (3) mechanisms individually by adjusting the bendConstants so that the force deflection curves look like those in the research article.
  15. Modify the bendConstants in the assembly to the tuned values from the previous step.

Here is a video by Veritasium that describes this mechanism very well...

This mechanism shrinks when pulled

Thanks Xray. That helps a lot.

You're very welcome.


My observation: While experimenting with this mechanism I discovered that it has four distinct snap points. Two occur when pulling the mechanism down and two while reducing the force allowing the mechanism to move up. The snap points are as follows:

When pulling DOWN, the first snap occurs at a force of 56 (+or- 1.0). The second snap occurs at a force of 121 (+or- 1.0).
When reducing the force UP, the first unsnap occurs at a force of 97 (+or- 1.0). The second unsnap occurs at a force of 21 (+or- 1.0).
The hysteresis of the upper snap points is 35, and the hysteresis of the lower snap points is 24. Hysteresis is the difference between the snap force while pulling down and the unsnap force while releasing the tension.

The forces that you cite in the comment above I will assume are correct for that scene, but the scene does not match the behavior of the mechanism in the video. I suspect you understand how the real mechanism is supposed to work as well or better than anybody else here. You may be able to get the scene to work correctly by adjusting bend constants. You (or somebody else) may be able to adjust the bend constants for each part of the assembly by multiplying all the bend constants for that component by the same value. Whoever gets the mechanism to work like the one in the video can post it as a response to ploters scene.

As I said in my prior comments, that was simply my observation of the scene, on my particular device (an older PC running Windows-10). I wasn't trying to prove anything, nor was I considering editing the scene to make it operate like the device in the video. You are orders of magnitude more talented than I am concerning mechanical devices and assemblies. There is no possible way in Algoland that I can make any mechanical scene operate better or more correctly or more accurately than you can. I always consider your scenes as learning experiences for me and I appreciate having you here on Algobox. Thanks!

My point was that I'm stuck and willing to accept help from anyone. The last upload is my best shot, but I'm not sure that it functions the same as the one in the video.

Understood. I think it would be very difficult to simulate the actual mechanism in the video without knowing some important details such as the densities, frictions, and elastic properties of each component. Given plenty of time for "trial and error" engineering maybe someone can come very close to the actual behavior.

Density doesn't matter since gravity is off. The components are frictionless. The elastic properties are shown in the force deflection curves in the paper. I added scenes that display the force deflection curves for the present system as a response to this scene. Modify bend constants to tweak the curves. You are running out of excuses (other than the "trial and error" thing).

Excuses?






Okay, yeah.