Rubik's Cube, A 3-D mechanical puzzle....
do u ever tried mechanism for it in Creo.. if not try it see attechment and hv fun ^^)
Message was edited by: DalbeerSinghSohal
Geez Dalbeer... You're trying to give people heart palpitations, aren't you?
I'm not sure I have the brain power to figure out another mind-bender. I'm still working on the pattern relations slides!
hahhahaha no thats is not my intenitons clearly.., it just came into my mind. i'm still trying to do it but no success so far. if there is a way to do it we will figure it out..
Yeah I thought about this one during my drive home (40 minutes) and I have a few ideas. I'm sure I can get one face to turn... the problem is when cubes need to switch faces. The best I could come up with was something tying the mechanism to coordinate system (possibly a spherical one if that's possible)... I just don't know.
I'll play with it and share the results!
Hopefully you've seen enough Rubik's cubes in your life to understand how they work. Building the center section with the center faces attached is easy enough. If you insert the EDGE pieces next using a ball joint (point to point centered on the cube), you start to get the right motion. You'll need to put a point in the EDGE part to represent the center of the cube to make this work.
Finally, you can do the same thing with the corners. If you put a point in the corner piece and assemble it as a ball joint, you can put the cube together.
Once the cube is together, if you start dragging bodies... everything starts going haywire. Turn on enable_advance_collision in your config.pro. Go to Tools>Assembly Settings>Collision Detection Settings and set global_collision_detection with the optional Push Objects on Collision button.
The calculations cause the model to perform sluggishly but it STARTS to work... a little. Any luck on your end??
trying some basic random stuff.... never knew there is somthing like push type feature on creo. going to start with somthing new tonight.. hoping for the good luck...
The more I play with this, the more I'm convinced it's beyond our current capabilities. You can simulate the motions of the cube by careful selection of the bodies and snapshotting the movements for animation. However, I don't think you can define the location of all the cubes and allow the mechanism to simply work through collision detection. It starts to work and then breaks down.
I've tried using ball connections, 6DOF connections with constraints, and a few radical ideas, too. Nothing really gives acceptable results. I even tried creating points on the edge and corner cubes then tying them to an invisible control sphere within the cube. I also tried adding 3D contacts to help maintain control of the various pieces. I'm totally able to constrain the model and make it look pretty. I can animate it (like Vladimir Palffy did in one of his videos). I sure can't make it move dynamically without bringing performance to its knees. At some point the components collide and "jam"... and then they don't move anymore.
The only thing you MIGHT be able to do... is create some sort of relation which would allow you to manipulate the cube mathematically. Maybe using matrices? It still doesn't give you the ability to click and drag the cube but it probably allows you to generate a realistic image of the cube in any possible configuration.
Hmm... not sure what to do with this!
i also use 6DOF connections, 3D conecction and collision that u tell and nothing comes out perfect. still running horeses utill they fed up..... lol . Vadimer animate it? i knew it! but not sure... but love the way he done animation 4 promotion...
Matrix? calculation? certainly not my cup of tea i hv this mathmanic-phobia
Here is my animation of Rubik's Cube
Have a nice day.
PS: For quick animation you can use many planar constraints and snapshots
Do you need create same assembly like org. Rubik's Cube (the real one) - create main body with many Pin constraints combine with planar constraints.
We were able to create the main body and place all the cubes. We were also able to turn on collision detection and get some of the cubes to turn as we "moved" them around. However, we were never able to mimic the complex motions of the cubes as they begin to move independently of each other. Think of how each cube can rotate and translate along the cube independent of the other pieces... BUT each cube also helps constrain all others at the same time.
This is tremendously difficult to achieve. I did come up with a mathematical way to represent the cube using coordinate systems. I'm sure this could be fed into Mathcad to generate all of the different possible configurations but this is beyond my capabilities.
Do you have any ideas we haven't tried? I think Dalbeer and I gave up on trying to go any further with this.
It is really impresive what you already obtain - congratulation!
I was design also Rubik's Cube (10 mm size) in Pro/E and few few days after I made it by hand.
By the way I solved the problem how to connect three axis like XYZ, but having each in one piece.
My real metal Rubik's Cube doesn't work so good, so I wish to make a new design in future, but first I wish to do it in Creo...
What do you think to use Pro/Program to parametrise a little our puzzle?
Love the Metal Rubik's Cube... if you could get that design working, I'm sure you're find plenty of geeks who'd line up to buy them!
How cool would it be to have a mini metal rubik's cube made out of, say, anodized aluminum?
My hand made version is still under construction, but I start finking to go and use CNC milling machine to manufacture really smooth working mini-cube.
I have few ideas how to finish & colour the sites 😉
Your idea is pretty nice to use anodized aluminium, but I think it is really hard to make different colours on each site in real live...
Yesterday I haven't got access to Creo, so here is my design:
I was design and make prototype of the first cube in the world using real bearings - you can see it on my YT channel: