What MT 138 Really IS!

[TheVisitorV]

Hi guys, had a good sim but ...shit happened, and for 2 days now my computer is acting up and the software as well, maybe 2 many parts, i'll try to reduce the number....i feel like messing the hole shit up.

How i came to find this, ...well is something that bothers me for a while and that is weight distribution on a horizontal axis, they way i did it it was always a balanced, cause i was using the centers...so i decided to try something a bit different when i realized that i could use a linkage to fix it in a different way, a bit different.

anyway, i will let you know what is the thing. Maybe some of yo can start to replicate if you have parts at your disposal. Or if you can try and simulate in other applications.

it's not a wheel it's a square, on each side of the squares there are 2 linkages, 8 in total. each of them is connected to the corners (through) so if one opens the other one closes, 90 degrees angles connection, and they either push each other or pull each other off the center. While the "wheel" spins, the weight is redistributed constantly.

The linkages are the weight, each link, X, is connected to the rod (in my sim) by a small sliding weight.

They can extend beyond the middle, center of the rod, so they are not only pushing pulling in 90 degrees but they do the same in 180 degrees as well. They keep constantly fighting over reaching their balancing points at every turn, every angle.

And because there are a number of weights, that expand or contract the weight distribution is constantly different and not symmetrical.

Its quite amazing....there might be a bit more to it, i'm trying desperately to make the sim but i don't know why it doesn't let me anymore....i will prevail in the end!!!!


tonight i have a number of hours at my disposal, but in the next few days i'll be a bit busy...

If you can please try and replicate this, lets see what we can come up with.

[TheVisitorV]

also i'm trying to figure out something else about linkages, they are special, they can do things that pulleys by themselves can't do, but i think they are complementary at some point.

anyway, i hope i can do a video later tonight with something worth showing. I hope it's just the numbers that i use that is messing everything up, many many links, lots of bearings ...

[silent]

.

[TheVisitorV]

Sure, this is what i'm struggling with right now....it all got messed up in strange ways, tried to redo all several times...gonna get back at it....the thing is that the more pieces there are the better weight redistribution it has...but i'll try do get away with 3 links on each side/corner now cause i don't know what else to do and i have no idea why is not working anymore.

https://www.youtube.com/watch?v=FIcLUioVvhU

[TheVisitorV]

just found a really cool video.

https://www.youtube.com/watch?v=7Fn-26Jmi5E

[silent]

.

[TheVisitorV]

i don't know what is going on here,i've been doing links in this software for weeks now and now not even the simplest ones work....i'm loosing my mind....

Your welcome silent, hopefully i'll get to the bottom of this and make it work again...

i think there is one more thing to that design, i'll try a different approach as well once i get it to work again.

[silent]

.

[ME]

just found a really cool video.
https://www.youtube.com/watch?v=7Fn-26Jmi5E

And we can create a speed-doubler :-)
http://www.besslerwheel.com/forum/viewt ... 220#163220

i don't know what is going on here,i've been doing links in this software for weeks now and now not even the simplest ones work....i'm loosing my mind....

Don't know that specific program, but by the looks of it your design needs to spend a lot of resources on collisions (Not sure, just a guess). Each element 'out of N' in potential contact with the rest means a there are (N²-N) pairs to check... with 3D-complex-volumes.
Perhaps you could separate some groups of collisions, cause even two groups will save you ½N² checks (monitor the memory usage if that gets reduced).
Hope that helps a bit. Good luck.

[TheVisitorV]

Hey, it's not a memory issue, the only good thing about my comp is that i have 16gb of ram, not computing issue, the cpu is really old and laggy but it processes ,slow but it does. The thing is that they behave strangely, they are all stiff and that, i've found one of the issues, being the way i grouped the objects into the tree. For some reason the sim plugin won't read the comp tree properly and won't recognize the objects, but there is a lot more then that.

I'm currently redoing it for the millionth time now, the first part seems ok, hopefully it will go smooth until the end.

i'm using less parts now and maybe i'll try to avoid direct collision by using some links at 180, seems the software works better with that sort of thing then direct collision.

There is a lot i wanna try with this concept but i got stuck in all sorts of stupid things, bugs, cpu power etc.

This reconfiguration geometry, there is a lot about it, being able to use the weights against each other and keep them shifting their center of gravity constantly....sorry i can't show you more right now, it all got messed up. Hope i will get it going tonight and be able to post a video of it, there is something magical about it, at least to me :)

working wheel or not, it would make some really nice kinetic sculptures :))

[TheVisitorV]

managed to get them move the way they should but now i got collision issues, so everything gets messed up fast, i know it can run flawlessly so i will push forward...

[silent]

.

[TheVisitorV]

Silent ty,
yes, it was great, to bad i messed that one up.

Anyways, managed to make stuff work again, did a few simulations, the last one from the video is very interesting, i will do more in the following days. There is something about this concept, i know i was so excted, maybe it was a fluke or a bug...i don't know anymore...

watch the video guys and let me know what you think.

https://www.youtube.com/watch?v=op1k0rcGQkw

I'll go to sleep now, been a long day :)

[Fletcher]

Hi Denver .. FYI you could probably make your sims a lot faster, more simple, and more reliable, by having less parts and interactions to compute.

Try some experiments with SB's. Also called Scissors, and Student Forceps etc. See if what I say is true for you ? It may save you some time and some angst with the program in future.

The many segmented Scissor mech like this ... xxxxxxxx ... only must be used when space or volume is a consideration. Luckily sims don't need to worry too much about that if parts are made to not collide.

That means you can replace many x's (segments) with just one (1) X or <x> segment etc. The caveat is that the the X angles remain the same as the angles on the original multi-x segmented SB.

IOW's there is no discernible difference in the function or performance between the two types however the lesser segmented SB has less parts, and less nodes, and less sliding pivots, and less just about everything to be kept track of.

All The Best.

[TheVisitorV]

Hi Fletcher, i did so, the initial models had many more links, but i had to increase the resolution of the calculus from the default 300hz to 5000hz so the software will properly calculate the collisions values. as you can see it plays a big part in working properly or messing everything up.


Here is the video with 3000hz,10x higher then default values
https://www.youtube.com/watch?v=sGtQTuY54os


Anyway, sorry guys, i've jumped the fence to early i guess :(

But i still think there is a lot more to this, having this kind of setup it allows the displacement of weights and reconfiguration of the entire system by having fixed and dynamic axes at the same time. The weights don't just hang there, they can be used to displace other parts through the entire system.

Next i'll have to work on adding between them some folding levers to use all that pressure from SB to displace inward and outward other weights


also if you have any ideas let me know