Would someone be willing to help me build a prototype for my magnet motor idea?

[DMBoss]

Got any good WhipMag models handy? 

0c:

You mean sim models, or real bench models?  I've got both actually.

But while the principle is obvious once you do some diligent testing, there are so many interrelated variables with too many degrees of freedom as to make the odds of hitting the right combo of variables extremely high! (like the odds of winning 1st prize in a lottery)

It's on my back burner for that reason (it's a potential sink hole of time and resources because the "hunt" has to be with trial and error, meaning buying all manner of different bearings and testing each with all manner of variations of the obviously non concentric machining of the stator roller bodies by the inventor)

It works by a resonant coupling of various magnetic and mechanical aspects.  so mass, inertia, bearings - which were some no name cheapos meaning their parameters of angular, radial and axial tolerances are unknown, as are their resonant behavior under load and on and on and on.

Heck how you cut the rotor slots, makes an air pump shoving air towards the roller magnets - which has an effect.

That said I did get how and why it works,  and verified some sub parts to that by making some deliberate changes and monitoring the results...

But you can't even use simulation to get into the ballpark, because of the unknown imperfections - a dynamic motion sim makes "perfect" joints - which don't exist in reality. You can introduce simulated imperfections with great care, but it's very time consuming and tedious. However the bearing imperfections are completely unknown (do these single bearings have 2 degrees or 20 degrees of tilt when cantilevering the loads of the roller magnets?) (how much was the inventor's toy lathe off center?  It assuredly was as it was reputed to be a Sherline, and he admitted only some of them worked) (what magnetic orientation relative to the physical non circularity of these roller bodies was used) (etc.)

See attached for some images in my archives.  PDF with comments.

I was able to get some random spurts of acceleration, but there were not controllable, nor as yet engineerable due to too many variables, each with too many degrees of freedom.

it is a resonant thing for sure.  And with XY plots of my sense coils in quadrature, some amazing wobble patterns emerged when the bearings did not chatter - and several harmonic transitions occurred as these patterns moved and danced.

So if any one has the fortitude to go at this in a systematic trial and error fashion, I'd be glad to help if I can.  Otherwise, I will be posting  details on this project on my now under construction web site soon:

http://qdmechanic.com/

DMBoss

[0c]

0c:

You mean sim models, or real bench models?  I've got both actually.

I was referring to sim models, but that's quite a test rig you have there. How come I never saw it before? I'm in the process of trying to take WhipMag to the next level, not trying to fiddle with the Alsetalokin device, just get back to my original design and complete it the way I think it should work.

I'm well aware of most of the issues with the physical device, I have two copies right here that were contributed to me by a highly skilled machinist, and have spent many long hours, days, weeks, months unsuccessfully trying to replicate the apparent behavior Alsetalokin demonstrated in his video.

I have come to the conclusion that is unlikely to happen unless Alsetalokin decides to get back in the game. So I'm just trying to get back to the basic specifications: same magnet specs, 18" rotor with 8 rotor magnets, 13 stator magnets, latches, and pivots.

I have some other business to take care of first, so it may be a few months before I get started on the physical construction (it will be contracted out because I don't have the skills). It sure would be nice to have a fairly accurate simulation to work with to get an idea how things might behave.

There is a new and more efficient stator latching mechanism to consider, and the rotor magnet pivot action as well. I know these things will complicate the models, but I'd like very much to see what the sim does, and play with some of the parameters a bit.

What simulation software do you use? Or is it home grown?

But you can't even use simulation to get into the ballpark, because of the unknown imperfections - a dynamic motion sim makes "perfect" joints - which don't exist in reality.

My original design should not be as sensitive to minor imperfections as Alsetalokin's AGW anomaly.

If you might be interested in pursuing this further, we could take this discussion over to fizzx.com, or just move over to the whipmag thread here on overunity:

http://www.overunity.com/index.php?topic=3871.0

Thanks for the chat, insights, and file,
OC

[X00013]

@Kim, Post a YT vid of ur proto, name it whatever to avoid anti-fe-tubers, and post the link here in this forum and it will go thru five degrees of humanity insanity. Baking is a lot less cheaper then tooling. The incredible minds of this forum will do you Justice, You do not want to be a tool $$. Weather it b 2-3-4 or 5 axis, as TK said, many  have the toolz,  I look foward to ur YT vid, X13

[DMBoss]

I was referring to sim models, but that's quite a test rig you have there. How come I never saw it before? I'm in the process of trying to take WhipMag to the next level, not trying to fiddle with the Alsetalokin device, just get back to my original design and complete it the way I think it should work.

I'm well aware of most of the issues with the physical device, I have two copies right here that were contributed to me by a highly skilled machinist, and have spent many long hours, days, weeks, months unsuccessfully trying to replicate the apparent behavior Alsetalokin demonstrated in his video.

I have come to the conclusion that is unlikely to happen unless Alsetalokin decides to get back in the game. So I'm just trying to get back to the basic specifications: same magnet specs, 18" rotor with 8 rotor magnets, 13 stator magnets, latches, and pivots.

I have some other business to take care of first, so it may be a few months before I get started on the physical construction (it will be contracted out because I don't have the skills). It sure would be nice to have a fairly accurate simulation to work with to get an idea how things might behave.

There is a new and more efficient stator latching mechanism to consider, and the rotor magnet pivot action as well. I know these things will complicate the models, but I'd like very much to see what the sim does, and play with some of the parameters a bit.

What simulation software do you use? Or is it home grown?

My original design should not be as sensitive to minor imperfections as Alsetalokin's AGW anomaly.

If you might be interested in pursuing this further, we could take this discussion over to fizzx.com, or just move over to the whipmag thread here on overunity:

http://www.overunity.com/index.php?topic=3871.0

Thanks for the chat, insights, and file,
OC

OC:

Well, I'd very much like to have the opportunity to quote on your subcontract work  Part of the reason I'm setting up the website is to share insights and offer some free stuff, but to also custom machine parts and systems for this experimenter market,  I've been doing this full time for a decade, and amassed quite a lot of knowledge and equipment and software and experience using and applying it all.

Responding to these kinds of posts, is jumping the gun a bit as I haven't finished the website construction (writing mostly, the systems are in place and work fine).  But being able to do good simulations, including in 3D with both magnetics and dynamic motion, in an integrated manner as part of the design cycle and for refinements once you have bench data to feed back into the simulation loop - is VERY helpful to reducing or telescoping actual trial and error to lower amounts.

I have access to 3D magnetics software, which costs over $20k, and uses Boundary Element Method (so you don't have to mesh the air or non magnetic structures, saving huge amounts of solve time - solving time is proportional to the cube of the number of mesh nodes)

I sent a test model that I had actual bench torque data on as well to a company that sells 3D magnetic sim software using Finite Elements, and their engineers took a month to solve the complete cycle of 120 steps, whereas the BEM software solved it in just about 24 hours of run time.

so yes, it's real, professional sim software.  And I can [with some limitations depending on the kinds of motion] integrate the magnetic forces and torques from this, with a professional level, powerful 3D dynamic motion simulator, and then account for mass, inertia, spring constants, and gravity in an accurate manner - and feedback the dynamic motion/position data to the magnetic sim iteratively.

We should communicate privately and send me some drawings or sketches of your design, so I can whip up some sim models and get some analysis done.  I only vaguely recall your original design nikolabackwards was working on when he stumbled upon the whipmag effect.

But I will say that intermittent, or non linear motion is one "key" to unlocking some gain from the non linearity of magnetic forces!  So latching has gain potential in a general sense - but not any old latching.  One of Steorn's purported models is the epitome of latching motion - not unlike the escapement of a clock.

so yes, let's collaborate.  Watch my web site often in the next days and weeks, and contact me there, or here.  (there you can send me bigger attachments)

http://qdmechanic.com/

there's only a temporary page there now, but there is a contact email link at the bottom.

There'd be no charge for generating a sim.  That's something I have extensive experience with, but it also requires practice to stay good at it. 

I'm glad you've reduced the number of variables - that is a killer regards doing the "art of the possible".  Even though we may know precisely why such a such and thing worked, if there are too many unknowns or variables trial and error could take years or decades - and few have that kind of resources or time to spend chasing rainbows.

A careful simulation integrating 3D magnetics with 3D dynamic motion could reveal whether, and by how much it might or might not produce a non zero net output.  And reveal if there's enough gain to overcome inertial and frictional and magnetic losses of various kinds too.

that's why I was able to get some results in the first go with my FM replication detailed on the temporary web site page: because I spent a couple weeks doing many integrated 3D motion and magnetic sims first before deciding on the exact form to try.  Not listed yet on the website, but I've made some changes to the FM replication and got much closer to the goal. 

Not that it matters, but attached is a chart showing one very early simulation result, just using FEMM/lua scripting and some dynamic motion feedback, with some wobble parameters picked out of a hat.  You CAN get a gain sufficient to account for the whipmag self running!  But how to find the right physical parameters to match these hand chosen parameters - is as you indicated not easy!

DMBoss