Ken
I'm getting into WM. I always appreciate a new tool.
Have never been ashamed to pull over and ask for directions either...it's stupid to drive around lost, ya know.
Have to leave now to go work at the hospital tomorrow morning, 89 miles from my shop. It's my other life.
It Would Be More Noble To Clear J. Bessler's Name, than trying to solely profit
Moderator: scott
re: It Would Be More Noble To Clear J. Bessler's Name, than
Robert (The Carpenter's Boy)
There's never time to do it right the first time, but there's always time to do it over again.
There's never time to do it right the first time, but there's always time to do it over again.
re: It Would Be More Noble To Clear J. Bessler's Name, than
Have been getting deeper into WM. I haven't built a Gera sized wheel yet. Instead, I put an equivalent sized project, with weights and other things on WM, hooked up the motor for 60 rpm, and have been seeking inspiration that way.
I have only been partially successful in transferring my method from autocadLT as a .dxf file, to WM. It seems to behave differently each time, and either locks up, or crashes WM.
Single points created in AcLT are pinned to the background in WM no matter what. Get around this by just substituting .02" radius circles. But AcLT is great for positioning everything first for WM later.
I have been able to prove that my weight's movement balances out on both sides of the axle. So, just a small force is required to move them one way or the other, in order for the whole wheel will turn by itself. The perimeter weights could be any weight, and 1 pound (plus whatever is needed to overcome friction), could move them into the positions they need to move to. I have all the pins and pivots set for a coefficient of friction of .003. A decent ball bearing is less than that.
The program crashes when I try to run this setup. Is it possible that WM senses PM, and judges that it conflicts with some norm?
Should I begin building Wheeler?
I have only been partially successful in transferring my method from autocadLT as a .dxf file, to WM. It seems to behave differently each time, and either locks up, or crashes WM.
Single points created in AcLT are pinned to the background in WM no matter what. Get around this by just substituting .02" radius circles. But AcLT is great for positioning everything first for WM later.
I have been able to prove that my weight's movement balances out on both sides of the axle. So, just a small force is required to move them one way or the other, in order for the whole wheel will turn by itself. The perimeter weights could be any weight, and 1 pound (plus whatever is needed to overcome friction), could move them into the positions they need to move to. I have all the pins and pivots set for a coefficient of friction of .003. A decent ball bearing is less than that.
The program crashes when I try to run this setup. Is it possible that WM senses PM, and judges that it conflicts with some norm?
Should I begin building Wheeler?
Last edited by rks1878 on Sun Jul 31, 2005 3:19 pm, edited 3 times in total.
Robert (The Carpenter's Boy)
There's never time to do it right the first time, but there's always time to do it over again.
There's never time to do it right the first time, but there's always time to do it over again.
re: It Would Be More Noble To Clear J. Bessler's Name, than
I don't know Ken ,but that would be a disappointment if it was true.
Even though I do not use it regular, I do try to follow your experiences as
you post.
How are you feeling lately?
Even though I do not use it regular, I do try to follow your experiences as
you post.
How are you feeling lately?
JB Wheeler
it exists I think I found it.
it exists I think I found it.
re: It Would Be More Noble To Clear J. Bessler's Name, than
OK, got everything into WM pretty well this time suddenly....after a few days of trying.
Wheel begins to turn fine, the weights are held in position until it's time for them to change positions, then as the wheel accelerates, a message pops up saying physical instability has been detected. In other words..("Warning, YOU are a First Law Violator. YOU are a bad Person.)??????
Probably just my own ignorance of the program, but if I reset to where it locks up, it begins to turn again, then the same result after it accelerates.
Maybe I should do something about this......
Wheel begins to turn fine, the weights are held in position until it's time for them to change positions, then as the wheel accelerates, a message pops up saying physical instability has been detected. In other words..("Warning, YOU are a First Law Violator. YOU are a bad Person.)??????
Probably just my own ignorance of the program, but if I reset to where it locks up, it begins to turn again, then the same result after it accelerates.
Maybe I should do something about this......
Robert (The Carpenter's Boy)
There's never time to do it right the first time, but there's always time to do it over again.
There's never time to do it right the first time, but there's always time to do it over again.
-
ken_behrendt
- Addict
- Posts: 3487
- Joined: Wed Mar 02, 2005 7:45 am
- Location: new jersey, usa
- Contact:
re: It Would Be More Noble To Clear J. Bessler's Name, than
Robert...
We have to have "faith" that WM2D will show PM if it is present otherwise the use of the program (and perhaps all other CAD/CAE programs as well!) would be a complete waste of time. I believe that WM2D can be used to find PM because the program does allow PM motion to be displayed, say, in a swinging pendulum when the air resistance is set to zero.
You seem to be making excellent progress with WM2D. Yes, it can act quirky at times, but I find that I always find some way to, at least, approximate to an acceptable degree what I want to make with it.
ken
We have to have "faith" that WM2D will show PM if it is present otherwise the use of the program (and perhaps all other CAD/CAE programs as well!) would be a complete waste of time. I believe that WM2D can be used to find PM because the program does allow PM motion to be displayed, say, in a swinging pendulum when the air resistance is set to zero.
You seem to be making excellent progress with WM2D. Yes, it can act quirky at times, but I find that I always find some way to, at least, approximate to an acceptable degree what I want to make with it.
ken
On 7/6/06, I found, in any overbalanced gravity wheel with rotation rate, ω, axle to CG distance d, and CG dip angle φ, the average vertical velocity of its drive weights is downward and given by:
Vaver = -2(√2)πdωcosφ
Vaver = -2(√2)πdωcosφ
re: It Would Be More Noble To Clear J. Bessler's Name, than
When you all run your WM simulations, does your cpu get hot? I have a 3.2 ghz. PIV and I think it's wise to blow a fan on it while I have my pmm attempting to go on it. The cooling fan comes on when I start the simulation. It's up to frame 338 now after about 45 min and the wheel has turned in the right direction about 3 inches.....it is continuing to compute the frames.....
Robert (The Carpenter's Boy)
There's never time to do it right the first time, but there's always time to do it over again.
There's never time to do it right the first time, but there's always time to do it over again.
re: It Would Be More Noble To Clear J. Bessler's Name, than
My CPU fan runs continuously.
A few years ago I lost a CPU because the fan failed during hot weather. About a year ago I could hear the new fan starting to fail again (noisy bearing), so I replaced it. My CPU activity meter hits and stays at maximum when I run a WM simulation.
A few years ago I lost a CPU because the fan failed during hot weather. About a year ago I could hear the new fan starting to fail again (noisy bearing), so I replaced it. My CPU activity meter hits and stays at maximum when I run a WM simulation.
re: It Would Be More Noble To Clear J. Bessler's Name, than
Yeah, I have a regular table fan blowing on it now. It's in a laptop.
Just went past frame 883 in the simulation. Still headed in the right direction. Can reset and begin again, then it goes quickly back to computing the frames....just watching it go pixel by maddening pixel until it finishes. About ten frames, and it moves one pixel.
Just went past frame 883 in the simulation. Still headed in the right direction. Can reset and begin again, then it goes quickly back to computing the frames....just watching it go pixel by maddening pixel until it finishes. About ten frames, and it moves one pixel.
Robert (The Carpenter's Boy)
There's never time to do it right the first time, but there's always time to do it over again.
There's never time to do it right the first time, but there's always time to do it over again.
-
ken_behrendt
- Addict
- Posts: 3487
- Joined: Wed Mar 02, 2005 7:45 am
- Location: new jersey, usa
- Contact:
re: It Would Be More Noble To Clear J. Bessler's Name, than
I haven't noticed any sort of excess heat coming out of my PC while running WM2D simulations even when I run one that has bogged down and seems to be taking forever to grind out the individual frames. My PC has a 2.0 GigaHertz processor, but I noticed that my tower has two cooling fans in it. I also use the PC in a room where it is located about 5 feet from an air conditioner that I always run during hot days...that probably helps a lot.
ken
ken
On 7/6/06, I found, in any overbalanced gravity wheel with rotation rate, ω, axle to CG distance d, and CG dip angle φ, the average vertical velocity of its drive weights is downward and given by:
Vaver = -2(√2)πdωcosφ
Vaver = -2(√2)πdωcosφ
re: It Would Be More Noble To Clear J. Bessler's Name, than
Well, just "blew" through frame 1278.
Wheel has turned about 113 degrees ccw.
I have four 5 lb weights that move in and out from the center. All are evenly spaced on a 6 foot diameter. Thought I'd just try 4 instead of more. Everything connected to each weight that moves them, is the same for all of them.
The follow-on weight has gone past where I started the simulation. It appears the wheel is starting over, the weights are cycling.
Will have to check this more closely. Must have done something wrong in the set-up.
At the rate we're going it won't be finished til tomorrow morning.....
Wheel has turned about 113 degrees ccw.
I have four 5 lb weights that move in and out from the center. All are evenly spaced on a 6 foot diameter. Thought I'd just try 4 instead of more. Everything connected to each weight that moves them, is the same for all of them.
The follow-on weight has gone past where I started the simulation. It appears the wheel is starting over, the weights are cycling.
Will have to check this more closely. Must have done something wrong in the set-up.
At the rate we're going it won't be finished til tomorrow morning.....
Robert (The Carpenter's Boy)
There's never time to do it right the first time, but there's always time to do it over again.
There's never time to do it right the first time, but there's always time to do it over again.
re: It Would Be More Noble To Clear J. Bessler's Name, than
Not convinced yet about WM.
After running all night, the simulation got through over 4500 frames. When I ran it this morning, the wheel stopped at various places. When begun again at different places, same thing. It's like it isn't sure of what to do. Each weight has only four parts, including the weight, and still, all this massive calculation over many hours....
Am trying again with 8 weights, with the four, as it approached a shift point (at about TDC), the wheel noticeably slowed, then speeded up after passing it.
Maybe 8, or more, will do it.
One reason I passed on WM at first is that it is a lot more difficult to position things, but after discovering that my autocad can do all that first to thousanths of an inch, I'm on to WM now, except for the above. Even when I put equal weights around the wheel precisely, it still falls one way or the other...
After running all night, the simulation got through over 4500 frames. When I ran it this morning, the wheel stopped at various places. When begun again at different places, same thing. It's like it isn't sure of what to do. Each weight has only four parts, including the weight, and still, all this massive calculation over many hours....
Am trying again with 8 weights, with the four, as it approached a shift point (at about TDC), the wheel noticeably slowed, then speeded up after passing it.
Maybe 8, or more, will do it.
One reason I passed on WM at first is that it is a lot more difficult to position things, but after discovering that my autocad can do all that first to thousanths of an inch, I'm on to WM now, except for the above. Even when I put equal weights around the wheel precisely, it still falls one way or the other...
Robert (The Carpenter's Boy)
There's never time to do it right the first time, but there's always time to do it over again.
There's never time to do it right the first time, but there's always time to do it over again.
re: It Would Be More Noble To Clear J. Bessler's Name, than
Robert,
Something isn't right. It should not take that long to run! When you import from a cad program, what type of objects do you get? Polygons maybe? Polygons seem to slow Working Models down to a crawl. How many objects (bodies, contraints, points, etc.) are listed in your properties window?
You say it is a lot more difficult to position things? Why is that? I find it a lot easier than a cad program. WM objects 'snap' to even incremental positions based on the viewing scale. And if they are not located quite where you want them just change their X & Y values at the bottom of the screen. If their size is not right just change that also. If it is not rotated right just change its angle.
With a CAD program you enter coordinates for both end points of a line. Then to make a square you must make four lines. Eight numbers total.
With WM you just click two corners to make a rectangle about where you want it. Then enter the width, the height, the center X & Y coordinate, and an angle if you want it tilted. Just four to five numbers. Many times no key strokes are even needed.
I don't have Autocad (I run BobCad) so I don't know what is happening to slow WM so much? But I suspect it has something to do with importing the Autocad file. You say each weight has only four parts so it can't be to very complex.
Something isn't right. It should not take that long to run! When you import from a cad program, what type of objects do you get? Polygons maybe? Polygons seem to slow Working Models down to a crawl. How many objects (bodies, contraints, points, etc.) are listed in your properties window?
You say it is a lot more difficult to position things? Why is that? I find it a lot easier than a cad program. WM objects 'snap' to even incremental positions based on the viewing scale. And if they are not located quite where you want them just change their X & Y values at the bottom of the screen. If their size is not right just change that also. If it is not rotated right just change its angle.
With a CAD program you enter coordinates for both end points of a line. Then to make a square you must make four lines. Eight numbers total.
With WM you just click two corners to make a rectangle about where you want it. Then enter the width, the height, the center X & Y coordinate, and an angle if you want it tilted. Just four to five numbers. Many times no key strokes are even needed.
I don't have Autocad (I run BobCad) so I don't know what is happening to slow WM so much? But I suspect it has something to do with importing the Autocad file. You say each weight has only four parts so it can't be to very complex.
re: It Would Be More Noble To Clear J. Bessler's Name, than
Jim,
I had the arc and circle smoothness set at 600. Moved it back to 100. Circles look a little rough now, but that may be what was hogging everything.
Positioning? In AutoCadLT (I don't have AutoCad Release 12), you can do a radial copy. Draw something once, set the degrees covered, number in the array, the center, click, done. All precisely spaced, etc.
Then there are the little things. I can pan and zoom with the wheel/ button mouse. Click the mouse button to repeat last command,... undo every previous step, not just one,...etc,.etc.
Have made changes to the AutoCad .dxf and imported to WM, ready for tomorrow and beyond....trying 9 weights.....
Good night all.
I had the arc and circle smoothness set at 600. Moved it back to 100. Circles look a little rough now, but that may be what was hogging everything.
Positioning? In AutoCadLT (I don't have AutoCad Release 12), you can do a radial copy. Draw something once, set the degrees covered, number in the array, the center, click, done. All precisely spaced, etc.
Then there are the little things. I can pan and zoom with the wheel/ button mouse. Click the mouse button to repeat last command,... undo every previous step, not just one,...etc,.etc.
Have made changes to the AutoCad .dxf and imported to WM, ready for tomorrow and beyond....trying 9 weights.....
Good night all.
Robert (The Carpenter's Boy)
There's never time to do it right the first time, but there's always time to do it over again.
There's never time to do it right the first time, but there's always time to do it over again.
re: It Would Be More Noble To Clear J. Bessler's Name, than
Robert,
I going to guess... autocad arc and circle smoothness set at 600 creates circles made from say 600 line segments? When you import into WM I'm again guessing that it turns each of those autocad circles into a multi-segment polygon. If you have any polygons look at their geometry (Ctrl-K) to see how many line segments each is made from.
You can copy from one WM window into another WM window. They will locate into the same location, unlike copying from a same window. You can rotate a wheel then re-copy and paste again.
Yes with WM positioning numerous weights can be tricky, but the faster simulation time by using real circles intead of polygons will more than make up for the extra effort. Besides, polygons don't collide properly.
I going to guess... autocad arc and circle smoothness set at 600 creates circles made from say 600 line segments? When you import into WM I'm again guessing that it turns each of those autocad circles into a multi-segment polygon. If you have any polygons look at their geometry (Ctrl-K) to see how many line segments each is made from.
You can copy from one WM window into another WM window. They will locate into the same location, unlike copying from a same window. You can rotate a wheel then re-copy and paste again.
Yes with WM positioning numerous weights can be tricky, but the faster simulation time by using real circles intead of polygons will more than make up for the extra effort. Besides, polygons don't collide properly.
-
ken_behrendt
- Addict
- Posts: 3487
- Joined: Wed Mar 02, 2005 7:45 am
- Location: new jersey, usa
- Contact:
re: It Would Be More Noble To Clear J. Bessler's Name, than
Robert...
I find that when my simulations get bogged down, that there is a simple quick fix that often gets things moving. Just click World>Accuracy>Fast. Doing that may help the program grind out those frames at an acceptable pace...
ken
I find that when my simulations get bogged down, that there is a simple quick fix that often gets things moving. Just click World>Accuracy>Fast. Doing that may help the program grind out those frames at an acceptable pace...
ken
On 7/6/06, I found, in any overbalanced gravity wheel with rotation rate, ω, axle to CG distance d, and CG dip angle φ, the average vertical velocity of its drive weights is downward and given by:
Vaver = -2(√2)πdωcosφ
Vaver = -2(√2)πdωcosφ