Page 1 of 1
physical simulation of a marble run
Posted: Mon Oct 18, 2021 4:43 pm
by Hansjoerg
Hello to all,
to get a little bit more into physical motion simulation, I was looking for a simple "project".
Well, what is fun for the little ones to play, is good for the big ones to practice.
https://www.youtube.com/watch?v=-yMGzGU9Ccs
But it is not as easy as it looks.
In my first attempt, the marble simply gets stuck, even though the hole in the starting block is large enough. (Marble run with Funnel.SLDASM)
Since I couldn't get the problem with the stuck marble under control, I built a start ramp (marble run with ramp.sldasm). But again I am not completely happy with the simulation. I have the impression that the marble loses disproportionately much speed already in the lower area of the ramp.
If you compare the simulation with the video, you can see that the marbles are much more lively in the same launch situation.
Am I missing any parameters here, or why does my result deviate so far from reality, or does it not work at all?
If you want to play a bit more, here is a simulator
https://www.cuboro-webkit.ch/
Re: physical simulation of a marble run
Posted: Mon Oct 18, 2021 4:52 pm
by mike miller
That's pretty cool! Unfortunately I'm not much help but I'm going to keep an eye on this because it's fascinating.
This is another fascinating project.
Backspin is important to your basketball free throw simulation study
Re: physical simulation of a marble run
Posted: Mon Oct 18, 2021 5:57 pm
by MJuric
Hansjoerg wrote: ↑Mon Oct 18, 2021 4:43 pm
Hello to all,
to get a little bit more into physical motion simulation, I was looking for a simple "project".
Well, what is fun for the little ones to play, is good for the big ones to practice.
https://www.youtube.com/watch?v=-yMGzGU9Ccs
But it is not as easy as it looks.
In my first attempt, the marble simply gets stuck, even though the hole in the starting block is large enough. (Marble run with Funnel.SLDASM)
Since I couldn't get the problem with the stuck marble under control, I built a start ramp (marble run with ramp.sldasm). But again I am not completely happy with the simulation. I have the impression that the marble loses disproportionately much speed already in the lower area of the ramp.
If you compare the simulation with the video, you can see that the marbles are much more lively in the same launch situation.
Am I missing any parameters here, or why does my result deviate so far from reality, or does it not work at all?
If you want to play a bit more, here is a simulator
https://www.cuboro-webkit.ch/
How is the marbel getting stuck. I didn't download what you've already done but dia really quick similar simulation and it seemed to work fine. Attached a quick video of it.
All I did was do a Motion analysis with a sphere and a swept path and it worked fine. I typically use "Use precise contact" and bump up the accuracy until parts aren't going thru each other. Beware though the higher the accuracy the longer it takes to calculate.
Re: physical simulation of a marble run
Posted: Sat Oct 23, 2021 7:01 am
by Hansjoerg
hey,
I've done a few additional tests.
I reduced the diameter of the marble. With the smaller marble, it is possible to get past the point where the marble with the original dimensions gets stuck.
But a little further on, the smaller marble also gets stuck. It looks like it is hanging on a silk thread from an sky hook.
I made a video of the simulation with the original dimension and the reduced dimension.
Re: physical simulation of a marble run
Posted: Sat Oct 23, 2021 12:21 pm
by zxys001
Hansjoerg wrote: ↑Mon Oct 18, 2021 4:43 pm
Hello to all,
to get a little bit more into physical motion simulation, I was looking for a simple "project".
Well, what is fun for the little ones to play, is good for the big ones to practice.
https://www.youtube.com/watch?v=-yMGzGU9Ccs
But it is not as easy as it looks.
In my first attempt, the marble simply gets stuck, even though the hole in the starting block is large enough. (Marble run with Funnel.SLDASM)
Since I couldn't get the problem with the stuck marble under control, I built a start ramp (marble run with ramp.sldasm). But again I am not completely happy with the simulation. I have the impression that the marble loses disproportionately much speed already in the lower area of the ramp.
If you compare the simulation with the video, you can see that the marbles are much more lively in the same launch situation.
Am I missing any parameters here, or why does my result deviate so far from reality, or does it not work at all?
If you want to play a bit more, here is a simulator
https://www.cuboro-webkit.ch/
..yeah, imho, something is not quite right with the assembly or the path curves?... so, to simplify this, I'd create another part or profiles to create your combined path and use that for the cutting for the others parts or use it as your main body for the simulation.. and it seems to work fine..(image)