Inventing wheels.. with built-in suspension =p
Forum rules
Be nice to others! Respect the FreeCAD code of conduct!
Be nice to others! Respect the FreeCAD code of conduct!
Re: Inventing wheels.. with built-in suspension =p
There seems to be deformations in the place it broke and as the material itself isn't best suited for such task this could be contributing to the end result.
And next time don't try it directly on asphalt. Try it out first on for example soil like ground.
And next time don't try it directly on asphalt. Try it out first on for example soil like ground.
- DeepSOIC
- Veteran
- Posts: 7896
- Joined: Fri Aug 29, 2014 12:45 am
- Location: used to be Saint-Petersburg, Russia
Re: Inventing wheels.. with built-in suspension =p
Measured the load-vs-deflection curve for a slice of the same wheel (re-printed, of course). Scaled up the results for the whole wheel.
The deflection is measured on the wrong side of the wheel. This was simply easier to make. But from how much did I turn the screw, I can tell that the rim probably bends about as much as spokes do.
Conclusions.
1. the wheel was supposed to carry half of my weight during normal operation. That point is a little bit too close to "hard stop". Increasing the number of spokes a little bit should fix this.
2. The wheel did not return to its original shape after removal of load. This means that fatigue is the most likely reason of failure. This has to be addressed.
3. The spokes straightening provide a kind of hard stop by themselves. But that hard stop is sustained by a handful of spokes that have been straightened out completely. This isn't a great hard stop in terms of reliability, especially if there is a sideways component in the load (but side loads shouldn't be significant on kick scooter, so that shouldn't be a big deal).
4. I grossly overestimated the max deflection to be 8mm. The graph shows it is a bit below 4.5mm, actually.
Measurement setup:
I wanted to stress the wheel to the point of breakage. But the setup cannot do it - that thin plywood lever starts to bend. So the breakage point is still unknown.The deflection is measured on the wrong side of the wheel. This was simply easier to make. But from how much did I turn the screw, I can tell that the rim probably bends about as much as spokes do.
Conclusions.
1. the wheel was supposed to carry half of my weight during normal operation. That point is a little bit too close to "hard stop". Increasing the number of spokes a little bit should fix this.
2. The wheel did not return to its original shape after removal of load. This means that fatigue is the most likely reason of failure. This has to be addressed.
3. The spokes straightening provide a kind of hard stop by themselves. But that hard stop is sustained by a handful of spokes that have been straightened out completely. This isn't a great hard stop in terms of reliability, especially if there is a sideways component in the load (but side loads shouldn't be significant on kick scooter, so that shouldn't be a big deal).
4. I grossly overestimated the max deflection to be 8mm. The graph shows it is a bit below 4.5mm, actually.
- DeepSOIC
- Veteran
- Posts: 7896
- Joined: Fri Aug 29, 2014 12:45 am
- Location: used to be Saint-Petersburg, Russia
Re: Inventing wheels.. with built-in suspension =p
Crushed.
The rim has failed. The spokes did a good job! I like this outcome, to be honest!
I had set up video recording of weight scale reading, to not distract myself from observing the actual collapse. Then I checked the video to see the value, to end up disappointed that the weight scale has been overloaded, so no actual value. All I can tell for sure is that it took more than 180 kg-f (scaled to whole wheel).The rim has failed. The spokes did a good job! I like this outcome, to be honest!
Re: Inventing wheels.. with built-in suspension =p
IMO you have the result for hit in the sharp edge, and thats good to know how wheel acts on sharp edge. As I see force acts in two points, and force cut the rim closer to the side where wheel is blocked by pivot point, under the direction of the spokes are bending . To simulate a flat surface, road, you was supposed to put an end of the clamp something flat like board. Perhaps such test is the better way to see the behavior of spokes.
- DeepSOIC
- Veteran
- Posts: 7896
- Joined: Fri Aug 29, 2014 12:45 am
- Location: used to be Saint-Petersburg, Russia
Re: Inventing wheels.. with built-in suspension =p
I feel lazy about fatigue testing, as I'm quite sure that it will be successful in destroying the wheel, and the result will likely be the same as with me riding the scooter (one small failure will avalanche into a mess before I stop the rig). So, I'm currently exploring the subject with the goal to tweak wheel parameters to avoid fatigue.
So far I found no data on fatigue characteristics of PLA, and even searching for its yield strength didn't give satisfactory results.
I found this page:
https://plastics.ulprospector.com/gener ... processing
It is in these funny American units, but quite a lot of data. The important value for me was yield strain, which I can easily calculate for the spoke.
Spokes have radii of curvature equal to 8.44 mm, and thickness of 0.45 mm. From this, I calculated the max strain in a fully straightened spoke to be 2.7%.
But when comparing this to the data on PLA (see link above), I feel lost. One method shows yield strain to be 9.8% to 10%. That is, I'm well below yield strain. The other method reports a vast range of 1.0% to 8.5%, which I am exceeding by a large margin if my PLA is worst-case.
If I am to reduce strain down to 1.0%, then I'm certainly screwed.
So I want to try to determine maximum bend radius of my PLA by experiment....
So far I found no data on fatigue characteristics of PLA, and even searching for its yield strength didn't give satisfactory results.
I found this page:
https://plastics.ulprospector.com/gener ... processing
It is in these funny American units, but quite a lot of data. The important value for me was yield strain, which I can easily calculate for the spoke.
Spokes have radii of curvature equal to 8.44 mm, and thickness of 0.45 mm. From this, I calculated the max strain in a fully straightened spoke to be 2.7%.
But when comparing this to the data on PLA (see link above), I feel lost. One method shows yield strain to be 9.8% to 10%. That is, I'm well below yield strain. The other method reports a vast range of 1.0% to 8.5%, which I am exceeding by a large margin if my PLA is worst-case.
If I am to reduce strain down to 1.0%, then I'm certainly screwed.
So I want to try to determine maximum bend radius of my PLA by experiment....
Last edited by DeepSOIC on Sat Sep 10, 2016 10:19 pm, edited 1 time in total.
Reason: fixed wrong radius value
Reason: fixed wrong radius value
Re: Inventing wheels.. with built-in suspension =p
What can be learned from the snaps design guide, is that you have to avoid to concentrate the stress and or strain in the material into one point. This seems to be the case with the current design.DeepSOIC wrote:So, I'm currently exploring the subject with the goal to tweak wheel parameters to avoid fatigue.
The point with the highest stress is where the spokes are connected to the rim and the hub. The spokes are bend against these points. So I am thinking, it would be better to make these points stronger and the spokes in the middle weaker. Such a design will distribute the stress and strain over the whole spoke and effectively lowering the maximal stress.
The curve from post posting.php?f=24&mode=reply&t=11913&sid ... 7#pr136668 shows a linear section. Above this section you will have yield effects. The yielding has to be avoided at minimum. But fatigue can start also in the linear section.
Making different designs and see if you get only linear deflection curves should give you an experimental start. With the FEM you could look at one spoke and see if the stress is distributed even over the whole spoke. You may need to do tension and bending tests with the FEM.
Ulrich
- DeepSOIC
- Veteran
- Posts: 7896
- Joined: Fri Aug 29, 2014 12:45 am
- Location: used to be Saint-Petersburg, Russia
Re: Inventing wheels.. with built-in suspension =p
No. This curve is non-linear due to design of the wheel, and not due to yielding. Yielding would have caused stress (aka load) to saturate. On the graph, the opposite is happening - strain (aka deflection) saturates.ulrich1a wrote:The curve from post posting.php?f=24&mode=reply&t=11913&sid=def853cbada572d2c3a03e69af073887#pr136668 shows a linear section. Above this section you will have yield effects.
This however doesn't mean there is no yielding on that graph - there is, as proven by the wheel not recovering from deformation. But the onset of yield is not related to the onset of nonlinearity.
Re: Inventing wheels.. with built-in suspension =p
I stumbled upon this searching for something else:
https://en.wikipedia.org/wiki/Airless_tire
http://www.bridgestonetire.com/tread-an ... cept-tires
https://en.wikipedia.org/wiki/Tweel
And if you look at the video of Tweel (especially around the 4:05) it does look like material you use isn't suitable for such task as the concept itself does work:
https://www.youtube.com/watch?v=UYZWL1ngS5E
https://en.wikipedia.org/wiki/Airless_tire
http://www.bridgestonetire.com/tread-an ... cept-tires
https://en.wikipedia.org/wiki/Tweel
And if you look at the video of Tweel (especially around the 4:05) it does look like material you use isn't suitable for such task as the concept itself does work:
https://www.youtube.com/watch?v=UYZWL1ngS5E
Re: Inventing wheels.. with built-in suspension =p
Some time ago I came across this video from Tomas Salander where he bakes 3d printed parts to release stresses built up during the printing process. I would guess that layer adhesion could be improved as well. Keep in mind the shrinkage witch is also a result of this process.
Hope this can help you even out the yield strain of your wheel, and bring them closer to spec
Hope this can help you even out the yield strain of your wheel, and bring them closer to spec
Need help? Feel free to ask, but please read the guidelines first
Re: Inventing wheels.. with built-in suspension =p
still working on the project?
I ran into this today, Bridgestone airless bike tyre, spikes - polyurethane.
The same principle was on my mind, a set of three parts where in the middle is flexible one ("the air").
ОК, your idea is a compact part, but...
I ran into this today, Bridgestone airless bike tyre, spikes - polyurethane.
The same principle was on my mind, a set of three parts where in the middle is flexible one ("the air").
ОК, your idea is a compact part, but...
- Attachments
-
- t1.jpg (30.47 KiB) Viewed 2913 times
-
- b1.jpg (33.96 KiB) Viewed 2913 times