Besslers prime mover and its enabler.

[Fletcher]

Here is the WM sim - I increased the mass factors by 10 x (all relative) so that the metrics would show - Numbers & Units set to show 3 decimal places while accuracy set to 7 significant digits .. Motor time On Zero to start ..

Replaced parallelograms (which bind usually) with gearing parallelograms so they work smoothly ..

You had the parallelograms exactly vertical in your sim therefore there was no torque (no masses could lose PE) - so I cranked them over by 1 degree to allow them to lose PE and rotate ..

..............



..............

[Roxaway59]

Thanks for taking the time Fletcher and I'm fairly sure I know what the problem is with Algodoo doing what it did.

Most people don't simulate at the size scale I very often simulate at they usually go for something bigger which is what you have to do in Algodoo to get more accurate results.

I use my simulations as a blue print for building the real thing and I usually go for something small and manageable but you cant build small accurate models in Algodoo. Bigger is better in Algodoo.

I was aware of this problem in Algodoo but I didn't realize that it was coming into play at the scale I was working at but its obvious now that it does.

So just to recap. Algodoo can be useful as a simulator but bigger is better it gets worse the smaller you go.
Graham

[Roxaway59]

This image seems to me like you understand what should happen. The left and right are equivalent, and this demonstrates what often happens with multiple pivots. It also demonstrates in a way how to simulate a 3d object in a 2d space...

Over the years Tarsier79 I have looked at and simulated all kinds of mechanisms and a lot of that work is now lost. I only wish I could go over my old WM2D files. When I looked at what Algodoo was doing I thought that cant be right, how is it doing that? So I knew I had to check it with WM2D.

Although this hasn't worked out as planned I will continue to focus on these types of overbalancing mechanisms to see if something significant has been missed.

Graham

[Fletcher]

Hey Graham .. glad you identified the problem with scale you were having ..

I have never found any practical differences with scale used in terms of concepts sim tested - so I always build in the 1 to 2 meter radius range for a wheel and its internals - then if something looks good I stress test the concept by changing the scale i.e does it perform the same larger or smaller, heavier or lighter - as long as the ratios are consistent they always do ..

Therefore to use Algodoo effectively I'd build bigger and when happy with the concept reduce size accordingly to reflect some bench-top size you might actually build ..

[Roxaway59]

Hi Fletcher yes I'm glad I don't have to give up on Algodoo because I find it easier to put things together and my old version of WM2D is far from perfect. Its never been glitch free and I kind of got use to it crashing on me.

My main focus at the moment is still on overbalancing my mechanisms and so far the most promising thing appears to be ones that overbalance successfully but then shift their weight so that the yellow weight can then overbalance the blue weight.

I can do that but getting it to then go round the wheel and fully reset is the age old problem. It doesn't seem impossible though just tricky.

Graham

[Fletcher]

To reduce instances of WM crashing it needs to be in a more stable location - I suggest you move it from your programs file etc into 'downloads' and leave it there - then make a short cut to open on your desktop etc .. that's how I manage mine and it made a huge difference to its stabiliy issues on my old laptop ..

[Roxaway59]

At the moment it does seem fairly stable but if it starts becoming an issue then I will do that. I know that some programs take up so much memory and processing power that you some times have to instruct the pc to make that program a priority. I had to do that with my recording software.

Graham

[Roxaway59]

Just an update on Algodoo.

I decided to go back to basics with my wheel and simulated it in Algodoo on a larger scale. I ended up with something that was very promising so without hesitation this time put it into WM2D.

Basically once again WM2D disagreed with Algodoo and it didn't work as hoped.

So the moral of the story is that if like me you use Algodoo to give you a good idea if a design works or not you are going to waste an awful lot of time and you will have to still check it in WM2D or similar once you have wasted all that time.

As others have said it hasn't been updated for a long time so instead of wasting time with the monkey its best to go straight to the organ grinder when it comes to gravity wheels.
Graham

[Robinhood46]

I think that is a bit like saying, because a racing car goes much faster and handles better, don't waste your time driving your nan to the market, in a Ford fiesta.
Basically Algodoo is very good for driving nan to the market, but don't bother taking it around the race track.

[Roxaway59]

Oh I'm still going to use it for taking nan to the market but I wont be kidding myself this time that I can overtake anyone and I will be sure to take spare oil and water.

The mistake I made with Algodoo was believing some of the write up about it like for example -

Algodoo is based on the latest technologies, from Algoryx Simulation AB, for interactive multiphysics simulation, including variational mechanical integrators and high performance numerical methods.

It makes it sound like its capable of simulating fairly basic things accurately but it can't even get basic things right that's the problem.

If you have a healthy attitude towards the program Robinhood46 then that's all well and good but personally its led me down the garden path more times than I care to mention.

There is also a very real danger that it could end up putting people off because its very disheartening to find that the project you’ve been working on for weeks is nothing more than a computer bug.

Bessler never had these problems so every time he had an idea he tried it out for real. I don’t think he would have even succeeded if he had been using Algodoo.

I know that you and others do real world experiments and if anything is going to lead us to the truth its going to be doing that. I will be sticking to using Algodoo for pictures from now on.

Graham

[Robinhood46]

I know that you and others do real world experiments and if anything is going to lead us to the truth its going to be doing that. I will be sticking to using Algodoo for pictures from now on.

I think you are being a bit unfair, when you say that real world experiments is going to lead us to the answer, because Algodoo can lead us to the real world experiments.
You are right that i have a positive attitude to Algodoo. I do think it is an excellent tool for sharing our thoughts. I certainly wouldn't say it is only good for sharing pictures, because a pencil and paper is far easier than Algodoo for making pictures. Algodoo allows us to share our thoughts of movement, far more accurately than drawings. I'm not referring to the accuracy of the movements themselves, more to the accuracy of the thoughts. One of the big problems we have is communication. Sam's thread is a good example of communication being a problem.
I think if Sam had the possibility to show a simulation with the movements he is describing, things would have been totally different than explaining them with only words. Pictures would have made it easier but an Algodoo simulation would have been the best. There wouldn't be any questions as to what Sam is trying to do, because we would see it. We could then focus on discussing what is actually going on with his attempt, as apposed to what exactly is he trying to do.
Sharing our thoughts is where we are at present, and Algodoo does this very well.

I only used WM2D for a very short period, and i found it more complicated than Algodoo. I think the term is "user friendly", and we didn't get on, we weren't mates. I must be honest and admit that i don't always get on with Algodoo either, so it would be more accurate to say that i find Algodoo less of a bloody annoying programme than WM2D.

[Roxaway59]

We agree a lot more than you may think RH46 I know exactly what you mean. Seeing an animation if you will is better than a picture and I too find Algodoo user friend. I think that's my problem though I find it too friendly and its like putting on a nice comfortable set of slippers.

WM2D on the other hand is like wearing hobnail boots. I will probably still use it for doing animations as well as pictures but I wont be posting on the forum as anything that I believe might work I will only be using my hobnail boots for that.
Graham

[Roxaway59]

Give me a minute and I will try to clarify why I don't think Algodoo should be taken seriously.

Graham

[Roxaway59]

Below is the latest simulation I did in Algodoo to see how accurate it would behave with my wheel design.

If I have this right what should happen is the wheel should gradually accelerate until just before 3 o'clock. At that point it begins to lose speed and just before 6 o'clock it should stop.

What happens in Algodoo is it accelerates past 3 o'clock then begins to lose speed and eventually stops past 7 o'clock.

If this was real my wheel would work.

Notice that all Algodoo is dealing with is a wheel with 3 weights on it but it doesn't even get this right and the scale is a lot larger than I usually use. If I have got this wrong then I'm sure that others can check it and correct me.
Graham

[Fletcher]

Graham .. you have air resistance ON with a wind blowing form left to right .. in WM if air resistance is on it is the objects velocity relative to still air conditions i.e. no wind blowing ..

Your sim is just a simple pendulum with the 27 gram green weight near the top making it top and slightly right heavy - if you turn off the wind factor it swings down CW like any pendulum would and the green weight will rise again to just about the same vertical height it started from but left of tdc, then reverse swing down etc .. this is what WM would show with air resistance off and no further friction losses ..

ETA .. basically Algodoo appears in this example to be doing what I would expect it to do ..