top

WhipMag JavaScript applet

Post new topic Reply to topic FizzX.org Forum Index | WhipMag Discussion/Development Goto page Previous  1, 2, 3, 4, 5  Next   Page 4 of 5

Mon Nov 24, 2008 8:41 pm PostPost subject:
korkskrew
Regular Contributor
Regular Contributor


Joined: 25 Jan 2008
Posts: 106
Location: Longmont, CO

Reply with quote

Wow...totally missed that thread. Shocked

So you came up with 0.00000032 Kg m and I estimated 0.00000025

Nope...not a factor of ten. But at least the difference is in the right direction. (edit) I noticed that you made some assumptions in that calculation that would tend to give you a result that is a little larger than reality though.
Back to top
View user's profile Send private message Send e-mail
 
Mon Nov 24, 2008 9:02 pm PostPost subject:
overconfident
Major Contributor
Major Contributor


Joined: 10 Feb 2007
Posts: 1121

Reply with quote

Based on what we were saying before about lighter stators being better, wouldn't Harvey's calculation be a step in the wrong direction? At least as far as OU is concerned?
Back to top
View user's profile Send private message
 
Mon Nov 24, 2008 9:24 pm PostPost subject:
korkskrew
Regular Contributor
Regular Contributor


Joined: 25 Jan 2008
Posts: 106
Location: Longmont, CO

Reply with quote

It isn't lighter stators that is the trick. It is a very large difference in MOI. You could add a flywheel to the stator and it would still go as long as you added a much larger flywheel to the rotor (and didn't increase friction losses), it would just go slower.

I haven't played with it enough to know the exact nature of the MOI ratio to residual torque relationship...but it seems like 1000 : 1 is a pretty reasonable ratio.
Back to top
View user's profile Send private message Send e-mail
 
Tue Nov 25, 2008 12:14 pm PostPost subject:
Harvey
Major Contributor
Major Contributor


Joined: 16 Jan 2008
Posts: 1927

Reply with quote

korkscrew wrote:

... Wow...totally missed that thread.


No worries. I've got stuff all over this forum that I doubt anyones ever read.
http://www.fizzx.com/viewtopic.php?p=7170#7170

I would be interested in the answer to that one. (I dropped that in when things were real slow here)
Back to top
View user's profile Send private message
 
Tue Nov 25, 2008 2:19 pm PostPost subject:
korkskrew
Regular Contributor
Regular Contributor


Joined: 25 Jan 2008
Posts: 106
Location: Longmont, CO

Reply with quote

I was messing with the javascript model last night and found something I totally didn't expect.

I increased the MOI of the stator to the 0.00000032 number and it didn't change the behavior that much. But then, I just boosted the stator friction by about 20X and voila! It maintains AGW sync down to 600RPM!

It still accellerates though without a large load on the rotor. Now I'm starting to get a theory going about what some of the mystery components in Al's rig are doing. If the stator lags the rotor too much then the residual torque on the rotor diminishes quickly, so too much stator friction is a bad thing. But when the stator friction gets too low, the AGW sync becomes unstable. The MKJD's stabilize the rotational behavior of the stator without adding friction. They just sort of limit peak velocity.

The GW stators are a little more of a mystery, but I have an idea about those too. They load the rotor so that it can't accelerate while the AGW sync is stabilizing. Once stabilized, the load can be removed and the system will stay stable as it accelerates.

The MKJD's aslo act to "flatten out" the non-linear motion of the stator as the speed increases. This reduces the net residual torque until it matches the friction and drag losses and the system stabilizes at it's top speed.

The replications don't accelerate because they all have too much stator friction, causing the stator to lag the rotor too much to produce the residual torque. Of course an excessive MOI relative to the rotor MOI would be very detrimental also. This would be a very deceptive flaw though, because the excessive stator friction encourages AGW sync, which makes the system seem to behave more like it "should".

Plausable theory? It all makes pretty good sense except for the residual torque part. That's the part that makes it a perpetual motion machine. It can't be right...but it looks right. How could FEMM get it so wrong? I can't believe that FEMM messed up that bad, but I can't believe the result either.

I still gotta build a test rig to verify the model.

Would anyone be interested in making some detailed torque measurements to verify the FEMM simulation? I could provide a set of key angular positions and only relative torque need be measured. It's just the relative shape of the surface that I need to know. Not the absolute magnitude.


Last edited by korkskrew on Tue Nov 25, 2008 2:57 pm; edited 1 time in total
Back to top
View user's profile Send private message Send e-mail
 
Tue Nov 25, 2008 2:46 pm PostPost subject:
korkskrew
Regular Contributor
Regular Contributor


Joined: 25 Jan 2008
Posts: 106
Location: Longmont, CO

Reply with quote

An interesting life lesson here...

I had given up on this model. I can't remember exactly why I decided to try again using excel, I think maybe it was becaus I was looking for a way to fill down time while I was in Israel. But it wasn't until after I gave up on it that it bore the fruit that I expected it to bear when I wrote it.

Feels kinda zen...like "to truely have something you must release it" or something.

Never mind, now I'm starting to sound like a fruitloop. Rolling Eyes (I mean aside from calling it a perpetual motion machine)
Back to top
View user's profile Send private message Send e-mail
 
Tue Nov 25, 2008 3:55 pm PostPost subject:
overconfident
Major Contributor
Major Contributor


Joined: 10 Feb 2007
Posts: 1121

Reply with quote

Sounds kinda like my dream. I had been trying so hard to understand what Orbo was all about. Then after the failed demo, I went into a slump for a few days, got all cynical end everything. I pretty much gave up on Steorn.

Then one morning I had a dream of my own about something else.
Back to top
View user's profile Send private message
 
Tue Nov 25, 2008 4:32 pm PostPost subject:
lostcauses
Major Contributor
Major Contributor


Joined: 02 May 2008
Posts: 871
Location: NM

Reply with quote

korkskrew
Interesting results.

On them Gw idler, in motion, turning; they act up on the mess differently than when stopped. If the timing is set up properly, they reduce the magnetic drag during the rotor magnets they encounter, When stoped they use that same magnetic drag at the rotor magnets reducing the magnetic push pull on the area away from GW.

It seem to me they shift the area they cause drag on this thing. They when moving may even add a slight gain to the thing, were as stopped they seem to be acting on rotor to stabilize the center of magnet field crossover of the AGW.

Them Idlers are as much fun to try and figure out as the AGW thing, LOL
Back to top
View user's profile Send private message
 
Tue Nov 25, 2008 6:36 pm PostPost subject:
MADPROF
Major Contributor
Major Contributor


Joined: 31 Jan 2008
Posts: 545

Reply with quote

@Korkskrew

With your model, is it possible to increase the strength of the stator magnets?

From my tests I think that the field which balances the stators between each other is weak.
_________________
"simplicity is the ultimate sophistication"
Back to top
View user's profile Send private message
 
Tue Nov 25, 2008 6:56 pm PostPost subject:
korkskrew
Regular Contributor
Regular Contributor


Joined: 25 Jan 2008
Posts: 106
Location: Longmont, CO

Reply with quote

Not exactly. The torque tables take about a week to generate, and I'm not gonna. (but you could if you like...my FEMM model is public...see the links near the start of this thread)

Stronger or weaker magnets can be aproximated by decreasing or increasing, respectively, the MOI and friction parameters in proportion to the change in the force of the magnets. F=MA is a linear equation so an increase in F is the same as a proportional decrease in M.

A change in distance is more problematic, becasue that changes the proportions (and to a lesser extent the shape) of the torque tables.
Back to top
View user's profile Send private message Send e-mail
 
Tue Nov 25, 2008 7:28 pm PostPost subject:
lostcauses
Major Contributor
Major Contributor


Joined: 02 May 2008
Posts: 871
Location: NM

Reply with quote

korkskrew
Think of the Gw in motion as a attraction against the (AGw attraction to loss.) The things then go into a balance that is essentially friction, and extremely small magnetic drag.
It does fit what you say on the friction thing.
Back to top
View user's profile Send private message
 
Tue Nov 25, 2008 7:54 pm PostPost subject:
korkskrew
Regular Contributor
Regular Contributor


Joined: 25 Jan 2008
Posts: 106
Location: Longmont, CO

Reply with quote

lostcauses,
I don't know...Al said that the GW stators make it easier to get AGW sync. Then he stops them after the first acceleration and it accelerates more. I just don't see them as much more than a load when they are spinning and nearly non-existant when they're stationary.

That's just my perception. I'm not saying you're wrong. You've studied it a lot more than I have, that's for sure.
Back to top
View user's profile Send private message Send e-mail
 
Tue Nov 25, 2008 8:49 pm PostPost subject:
MADPROF
Major Contributor
Major Contributor


Joined: 31 Jan 2008
Posts: 545

Reply with quote

@korkskrew

Thanks but no thanks, I don't stand a chance of doing that Laughing

Bigger diameter stator mags would work, but the extra mass knackers the bearing friction. May be OC's floating stators?
_________________
"simplicity is the ultimate sophistication"
Back to top
View user's profile Send private message
 
Tue Nov 25, 2008 8:58 pm PostPost subject:
korkskrew
Regular Contributor
Regular Contributor


Joined: 25 Jan 2008
Posts: 106
Location: Longmont, CO

Reply with quote

Wait'll you see the test rig I'm cooking up. Cool
Back to top
View user's profile Send private message Send e-mail
 
Tue Nov 25, 2008 9:23 pm PostPost subject:
MADPROF
Major Contributor
Major Contributor


Joined: 31 Jan 2008
Posts: 545

Reply with quote

@Korkskrew

I read some of your ideas a few post back.. Sounds good Cool Please keep me posted.

My next project will be a floating spindle, no bearings....Plenty of wobble Mr. Green
_________________
"simplicity is the ultimate sophistication"
Back to top
View user's profile Send private message
 
Tue Nov 25, 2008 11:31 pm PostPost subject:
lostcauses
Major Contributor
Major Contributor


Joined: 02 May 2008
Posts: 871
Location: NM

Reply with quote

korkskrew wrote:
Wait'll you see the test rig I'm cooking up. Cool


Will be waiting to see this...
Back to top
View user's profile Send private message
 
Tue Nov 25, 2008 11:49 pm PostPost subject:
Harvey
Major Contributor
Major Contributor


Joined: 16 Jan 2008
Posts: 1927

Reply with quote

korkscrew wrote:

Would anyone be interested in making some detailed torque measurements to verify the FEMM simulation? I could provide a set of key angular positions and only relative torque need be measured. It's just the relative shape of the surface that I need to know. Not the absolute magnitude.


It could be done, but would require a custom jig and accurate calibration to be useful. I have a few ideas in mind, all of which would take about a day to fabricate and calibrate. Perhaps you have something in mind?

Personally, I would like to confirm the tables mathematically. It seems all of the programs (FEMM & Vizimag etc) utilize the electromagnetic equivalency to produce the forces. IOW, they convert the PM into a solenoid and then use ampere turn calculations to derive B. This comes from the 'understanding' that the electrical force and magnetic force are inexorably linked. That is, if electric current is flowing there must be a magnetic field, and inversely, if a magnetic field exists, electric current must be flowing. The potential energy between two charges (whether particle or entire surfaces) is Voltage. When the charges are allowed to balance or stabilize through a conductor, current flows and a magnetic field is created. Electronics 101. So a charged battery has potential across its poles and it just sits there right? But there is an electric field that surrounds those poles but no magnetic field. This leads some to view the PM as a type of magnetic battery. It has a magnetic field surrounding its poles but no electric field. But is this really possible? To have a magnetic field where no current is flowing? This is where the math typically used breaks down. Some try to rationalize that the current is flowing, in every atom in the magnetic material and thus the combined fields of each electron is what results in a magnetic field. Sounds plausible.

Does an electric spark in a vacuum produce a magnetic field? Or does there have to be a conductor for the field to develop?

It is currently impossible to treat magnetism mathematically without linking it to current flow. But I postulate that there are other particles and methods of producing magnetism where electrons do not exist. (can you think of any matter void of electrons?) When these are fully publicized it will rush in a whole new approach to magnetism.

In the mean time, we need some algorithms to apply to the WhiPMag that allow the insertion of variables for experimentation. We have equations for pole to pole calculations, but when we introduce an incident angle it throws the equations for a loop. Why? Because B is a vector and shoots straight out the center of the PM, but the real life force 'felt' at the pole is spherical and produces nearly the same force throughout the entire sphere. So, there are two spheres, one for each pole and they intersect at the equator. EXPERIMENT: Take some quarter inch neos, say 10 and stack them into two stacks of 5 each. Now holding one stack in each hand, place them in repulsion. Now, carefully rotate the incident angle always keeping the center of the pole of the one facing the center of the pole of the other. So you are moving one stack around and the other is held firm. Imagine the firmly held stack is transparent so you can point at the center of its pole face even from behind near the center of the stack. Keep an even pressure throughout the excercise, noting the average distance the poles are separated at that distance. Now choose that distance on the fixed polar plane (perpendicular to fixed B) such that the mover B is pointing at the fixed B and the two are 90 from each other. Keeping the two poles at this distance and using the pole as a pivot point, rotate the moving stack so that both stacks are parallel with B vectors pointing in the same direction. Now pivot it back to where you started and continue on for another 90 so that the B vector has rotated through 180.

From that experiment you should have learned that the force separating the magnets is the same in all directions (spherical) and is independant of the B vector. The only mitigating factor is that the poles are of a particular polarity. Like poles have a repel force, opposite poles have an attraction force.

From this can begin new mathematical models that treat the poles as separate single spherical point sources of force of a magnetude* that follows the inverse cube rule. The subsequent interactions between these spheres becomes as simple as modeling the interaction of jelly balls Mr. Green (Ok, its a bit more complex than that, but you get the drift).

And from that, we can derive tables to compare to the FEMM force tables. Wink

Cool

*this is a typo that results in a new punny word that exemplifies the measured attitude of a magnet Mr. Green
Back to top
View user's profile Send private message
 
Wed Nov 26, 2008 12:01 am PostPost subject:
overconfident
Major Contributor
Major Contributor


Joined: 10 Feb 2007
Posts: 1121

Reply with quote

I tend to think of those fields as whirling (vortex?) soap bubbles.
Back to top
View user's profile Send private message
 
Wed Nov 26, 2008 12:27 am PostPost subject:
korkskrew
Regular Contributor
Regular Contributor


Joined: 25 Jan 2008
Posts: 106
Location: Longmont, CO

Reply with quote

Harvey wrote:
korkscrew wrote:

Would anyone be interested in making some detailed torque measurements to verify the FEMM simulation? I could provide a set of key angular positions and only relative torque need be measured. It's just the relative shape of the surface that I need to know. Not the absolute magnitude.


It could be done, but would require a custom jig and accurate calibration to be useful. I have a few ideas in mind, all of which would take about a day to fabricate and calibrate. Perhaps you have something in mind?

Actually yeah. I don't agree about the calibration thing. As long as the measurements have decent resolution and are repeatable, I don't need to know the absolute value, just the relative value. This should make calibration moot. A protractor, and two straight, parallel, springy things (piano wire or coffee stirs or something).

The reason I'm concerned with only the relative measurements is that it's just the shape of the torque space that I'm trying to confirm. If the shape is right then the tables are probably accurate enough to trust. It's the shape, the locations of the peaks and valleys, that interact with the velocity changes to produce the residual torque.

But to be honest, if the tables are accurate I'm gonna start wigging out a little. It's the last possible error source that I can figure out. If they are right then it looks like Al accidently figured out a way to keep going down a hill without having to go all the way back up it. That's just preposterous!

I keep thinking I made another mistake, but I didn't. dt is constant. M is constant. F comes from the table. It's really easy math...

Harvey wrote:
Personally, I would like to confirm the tables mathematically. It seems all of the programs (FEMM & Vizimag etc) utilize the electromagnetic equivalency to produce the forces....{giant snip}...And from that, we can derive tables to compare to the FEMM force tables.

*whew* Um...yeah. I agree it would be cool to derive some calculus that would describe the whipmag system or even some approximation of it. But really I think finite element analysis, though it is just an approximation, when done at adaquate resolution is a pretty darn good approximation...and it doesn't require the whole math genius thing.

I think experimental confirmation is about as good as we're going to get. *shrug*

(edit) ...but making the magnets idealized point source vectors is a pretty cool idea. Maybe if I find myself with some more extra time...
Back to top
View user's profile Send private message Send e-mail
 
Wed Nov 26, 2008 3:27 am PostPost subject:
lostcauses
Major Contributor
Major Contributor


Joined: 02 May 2008
Posts: 871
Location: NM

Reply with quote

Can you convert your sim to a gw direction and show what the forces should be, a curve of the action across the rotor magnet might alow folks to see if your sim works in that form.

It also could also work with the AGW at some rpm, if you could draw the curve and then some one can go with a Oscope, just possibilities..
Back to top
View user's profile Send private message
 
Wed Nov 26, 2008 3:33 am PostPost subject:
overconfident
Major Contributor
Major Contributor


Joined: 10 Feb 2007
Posts: 1121

Reply with quote

Phase angles would be completely different. Would probably require complete regeneration of the FEMM tables, a week's work. It would be nice to compare. Don't know whether Korkskrew has the time or inclination to do it.
Back to top
View user's profile Send private message
 
Wed Nov 26, 2008 3:34 am PostPost subject:
lostcauses
Major Contributor
Major Contributor


Joined: 02 May 2008
Posts: 871
Location: NM

Reply with quote

OC true.
Back to top
View user's profile Send private message
 
Wed Nov 26, 2008 4:24 am PostPost subject:
korkskrew
Regular Contributor
Regular Contributor


Joined: 25 Jan 2008
Posts: 106
Location: Longmont, CO

Reply with quote

YOU can do a GW sim. Just make one RPM positive and the other negative. Easy!

Or were you talking about the excel model?

The torque tables cover all possible angular orientations...that's why they take so damn long to make. The things that require re-running the simulations is changing the distance between magnets or the size or something like that.
Back to top
View user's profile Send private message Send e-mail
 
Wed Nov 26, 2008 5:45 am PostPost subject:
lostcauses
Major Contributor
Major Contributor


Joined: 02 May 2008
Posts: 871
Location: NM

Reply with quote

korkskrew
What is the range of torque we are dealing with?? What kind of torque gauge range will we need to get??
Back to top
View user's profile Send private message
 
Wed Nov 26, 2008 1:53 pm PostPost subject:
korkskrew
Regular Contributor
Regular Contributor


Joined: 25 Jan 2008
Posts: 106
Location: Longmont, CO

Reply with quote

I only know what is in the torque tables...

The peak rotor torque in the table is about 0.63Nm. The peak stator torque is 0.1Nm.

As we all know, magnets vary...but that should get you in the ball park.

BTW, I never answered your concern about error from using discrete time steps. It is true that the discrete approach introduces errors. This is the reason I made the new hi-rez torque charts.

I just want to assure you that I was not ignoring this as an error source. One of the javascript revisions was made just to prevent this kind of error. I made the time slicing automatic to prevent someone from chosing a bad time scale that caused clearly incorrect behavior.

In the Excel model, I could actually measure the error produced by using different length time slices. The time slices that I ended up using produced errors below my ability to measure.

I am quite sure that discrete time slices, as I implemented it, could not account for such a large positive residual torque.
Back to top
View user's profile Send private message Send e-mail
 
Wed Nov 26, 2008 7:40 pm PostPost subject:
lostcauses
Major Contributor
Major Contributor


Joined: 02 May 2008
Posts: 871
Location: NM

Reply with quote

I have put out a few requests for torque gauges, May get luck and borrow one.
As for the residual torque, I believe it is there with proper timing. I take it you went with 1/2 inch magnets. I will also have to set up a large enough degree wheel to get to at least a good measurement of degrees, LOL

The HSF vids, show a travel of the stator in proper sync to shift the oposing pole to pole position in a tight area, This is a decrees in the pull to and push against area that happens due to not at 180 degrees,LOL my latch.
This will cause a shift, small in the break strength across the magnets. This of course reduces the amount of momentum needed to get it across this point. If timing is off here, the mess hits the strong wall causing the thing to slow down. This is what is suspect is going on with this in the replications.

The flip into the repulsion, Gain, then attraction against, Loss: is were this gets it gain. Strangely enough this is were them idlers come into play by counteracting the attraction to loss if set up properly. Again magnet distance from each other is the key to the idlers hellping, or hurting.

I do not doubt the gain I see in the HSF vids. Yet to keep it there with variations in magnet strength etc,, well.... that is going to be the trick. I do suspect you sim is correct in most things. using a stable and consistent magnet length and strength in a sim is not what you get in real life.

I still say this is the magnet pendulum, and you ideas and tables are correct in a gain in torque. If it can be passed to the next one with some gain, it will continue to go as long as it can continue to stay in sync.

Yea I know that they say it is impossible to do, yet great minds have also been saying it can be done just as long. LOL

I wounder if I should get into the aspect of gravity and the similarity to Bessler's wheel with this thing. Hmm make me wounder....
Back to top
View user's profile Send private message
 
Wed Nov 26, 2008 8:08 pm PostPost subject:
overconfident
Major Contributor
Major Contributor


Joined: 10 Feb 2007
Posts: 1121

Reply with quote

lostcauses wrote:
I still say this is the magnet pendulum, and you ideas and tables are correct in a gain in torque. If it can be passed to the next one with some gain, it will continue to go as long as it can continue to stay in sync.


I agree with the pendulum concept. It's just a pendulum laid on its side and reacting to dipolar magnetic forces instead of gravitic forces.

The first partial test ever done of the WhipMag concept was a pendulum constructed by Axle in July or August 2007. It was pretty crude and inefficient and did not show any gain at the time. I tried to interest Dr. Mike in doing some additional pendulum experiments, but he declined.

I still think it may be possible to demonstrate a gain with a pendulum. I'm just not sure how to design a practical experiment that is likely to succeed.
Back to top
View user's profile Send private message
 
Thu Nov 27, 2008 5:02 am PostPost subject:
jcmax
Newbie


Joined: 26 Nov 2008
Posts: 3

Reply with quote

Harvey wrote:

From this can begin new mathematical models that treat the poles as separate single spherical point sources of force of a magnetude* that follows the inverse cube rule. The subsequent interactions between these spheres becomes as simple as modeling the interaction of jelly balls


I might as well throw this one over the transom, then:
http://jcmax.pbwiki.com/SpinningDipoles
Back to top
View user's profile Send private message
 
Thu Nov 27, 2008 5:27 am PostPost subject:
overconfident
Major Contributor
Major Contributor


Joined: 10 Feb 2007
Posts: 1121

Reply with quote

@jcmax, where you been hiding all this time?

(90% of the math goes right over my head. But what you said makes a lot of sense.)
Back to top
View user's profile Send private message
 
Thu Nov 27, 2008 6:01 am PostPost subject:
lostcauses
Major Contributor
Major Contributor


Joined: 02 May 2008
Posts: 871
Location: NM

Reply with quote

A good read, looks correct in the variables given. You do show the AGW is a natural thing which confirms the replications so far.

deleted three post to this so far so I will leave it at this... LOL
Back to top
View user's profile Send private message
 
Post new topic Reply to topic FizzX.org Forum Index | WhipMag Discussion/Development
View previous topic
View next topic
Display posts from previous:   




You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum