magnets: to stack or not to stack

[brandnewb]

Putting them side by side one up one down will likely reduce the flux through the coil as much of the flux will take the shortcut to its neighbour, instead of crossing the coils to the other rotor.

But yes, getting the magnets in postion and keeping them there may well be... challenging  .

Yes I think you are correct.. I have been taking quite a bit of readings today and for what I have seen the field does not go deep but the field reversal could have been what made the difference.

I no longer take that as a possibility for a single stator dual rotor setup.

The jury is still out for the single rotor dual stator setup.

I have not started officially yet taking readings. As I want to take a more structured way of doing things going forward.;

[brandnewb]

Nope, stacked same pole, yes it is a pain in the... glue one magnet down with JB Weld and give it 24 hours to setup, then move next magnet in and glue and clamp in place. Helps to make a template to place the magnets that contain magnets sliding sideways.

Gives a wider pole on such a large machine so a stronger pulse from each coil. IF you follow the 4:3 rule...

That was just the challenge one needs on a sunday one had hoped sitting by the fireplace watching disney films with the kids and then this bomb drops.

But because I am blast proof I thought to actually try and understand the rest of what you had to say but I grinded to a halt.

4:3 rule ??

I did a quick search and I find only car related results. And on this forum the search comes up with no results.

Can you please elaborate a bit?

[Mary B]

I may have it wrong on the ratio... getting over a stomach bug so brain is a bit fuzzy today... ratio of coils to magnets... 12 magnets to 9 coils... divide by 3 so 4:3 ratio of magnets to coils...

[brandnewb]

I just am not sure how to respond. I hope you get to recover. You are a great mental support to me.

And I hope that this heartfelt statement is of help to you too.

[TechAdmin]

JB Weld? Does that thing even work? I had to use it for my bike exhaust and it came off before even installing the exhaust itself back -_- (Yes I did mix it and followed the procedure and all)

[MattM]

Just a heads up, JB Weld is magnetic.

[JW]

Haaaaaa!

I had this customer and the O2 sensor on the exhaust manifold was striped out. I decided to use the JB Weld (hi temp epoxy) and reinstall the O2 sensor, let it sit overnight went for a test drive in the morning, after driving for ten minutes it spit the O2 out. I had to replace the manifold to make it right.

[JW]

https://www.fieldlines.com/index.php?topic=138707.0

[Mary B]

Surface has to be perfectly clean and roughed up for JB weld to grab well. Friend cracked the block on his lawnmower... JB welded it back together and used it for 5 more years before the engine grenaded...

[brandnewb]

Ok so it turns out, again, that I am at fault. Not the TD8620 Tesla Meter that I have been using.
One of those classic RTFM moments at display again. Anyway if I press the Null button once a session starts I am able to produce consistent readings all the time.

But consistent in this context does not mean consistent. You see holding the probe precisely is difficult and that goes double for one with non stable hands But fear not I estimate an average of what I see coming across during a reading.

I will admit that I can't produce a deep dive yet and unless there is a demand for it I probably will not bother. You see I am cutting corners just to get to what I want to see. This is all far from scientific I know but I have been slaving away all day and not even done yet.

This is what I have so far.
Magnet 1 in isolation I N                  
id   field polarity   surface   measurement point   distance (mm)   milli tesla   remark
I   N   top   a   0   306   
I   N   top   b   0   324   
I   N   side   c   0   212   
I   S   side   d   0   220   
                  
I   N   top   a   10   54   
I   N   top   b   10   51



Magnet 2 in isolation II S                  
id   field polarity   surface   measurement point   distance (mm)   milli tesla   remark
II   S   top   a   0   304   
II   S   top   b   0   312   
II   S   side   c   0   235   
II   N   side   d   0   142   
                  
II   S   top   a   10   54   
II   S   top   b   10   53   
                  
                  
both magnets stacked in isolation                  
id   field polarity   surface   measurement point   distance (mm)   milli tesla   remark
I   N   top   a   0   412   
I   N   top   b   0   428   
I   N   side   c   0   325   
I   N   side   d   0   112   
II   S   side   c   0   316   
II   S   side   d   0   115   
                  
I   N   top   a   10   86   
I   N   top   b   10   90   

both magnets stacked in isolation placed on  top of 10mm powder (1mm filament gap; mt on the back of the powder slab reads 26 S so not saturated yet)                  
id   field polarity   surface   measurement point   distance (mm)   milli tesla   remark
I   N   top   a   0   469   
I   N   top   b   0   471   
I   N   side   c   0   497   
I   N   side   d   0   161   
II   S   side   c   0      no space for probe
II   S   side   d   0   32   
                  
I   N   top   a   10      wow completely forgot what happened here. Maybe start over
I   N   top   b   10      wow completely forgot what happened here. Maybe start over



both magnets stacked in isolation placed on  top of 5mm steel plate         because I am not too interested in steel as it is rather expensive and I am looking to use the powder again I cut corners and did not measure the field on the back         
id   field polarity   surface   measurement point   distance (mm)   milli tesla   remark
I   N   top   a   0   512   
I   N   top   b   0   522   
I   N   side   c   0      not interested anymore
I   N   side   d   0      not interested anymore
II   S   side   c   0      not interested anymore
II   S   side   d   0      not interested anymore
                  
I   N   top   a   10      not interested anymore
I   N   top   b   10      not interested anymore



It is taking forever for the glue to dry holding the stacked magnets in place for my final test so please allow some time.
The final test will be the stacked magnets with only the south part covered with powder. Not sitting on top of anything but half way submerged

In the meantime if anyone can shoot holes in my approach then please do so. I can still make changes

[Bruce S]

I am curious.
How noisy is your VAWT?


Bruce S

[brandnewb]

You mean the previous heavy weight cups at 2.44m tall and 63cm wide?

I have not used a decibel meter yet as I have none but, and please understand that ones ability to hear sound is subjective (mine is 20/20 btw), I hear almost nothing. I am sure my neighbor would not mind.

Now for the Air Wheels I am in such an early phase that they have not spun yet. So no assessment could be made. Somehow I am confident I can make the decibels agreeable to all parties involved.

Or are you worried about the alternator becoming a discotheque?

[brandnewb]

A more in depth response is that the spinning of the cups it self is almost silent.

What one can hear a bit from 4 meters away standing in the direction of the wind is the mechanics and structure of the turbine cracking a bit.
I am aware of this issue yet this does not signal to me a potential noise problem with the neighbors but rather that the structure is under too much stress.
I am on the case and waiting for new material to arrive to give the structure the strength it deserves.

[brandnewb]

ok disaster averted.

I placed the 10mm powder slap on the stacked magnet while it was drying.
But this does mean that the reading is now on the south pole of the second magnet.

both magnets stacked in isolation placed on  top of 10mm powder
(1mm filament gap;)
EDIT" ( mt on the back of the powder slab reads 26 S so not saturated yet): this part should be discarded. I did not test and even if I would have it would have been the N pole                 
id   field polarity   surface   measurement point   distance (mm)   milli tesla   remark                 
II S   top   a   10      103
II   S   top   b   10      108

[brandnewb]

And one very important thing I forgot to mention is that in this case the magnets were somewhat equal in strength.

But before my unscheduled disassembly with earlier magnets I had readings on the side in where one magnet was partly taken over by the other one. So on the stronger magnet on the side surface it would read on point c and d an N pole and on the other twin magnet it would read and N and an S pole.

My prints already take this into account in where I aim for a submersion of only 2 mm. It is far too much work to fine tune each segment to each magnet pair.

[Bruce S]

You mean the previous heavy weight cups at 2.44m tall and 63cm wide?

I have not used a decibel meter yet as I have none but, and please understand that ones ability to hear sound is subjective (mine is 20/20 btw), I hear almost nothing. I am sure my neighbor would not mind.

Now for the Air Wheels I am in such an early phase that they have not spun yet. So no assessment could be made. Somehow I am confident I can make the decibels agreeable to all parties involved.

Or are you worried about the alternator becoming a discotheque?

My question based on the original set up and possibly any current tests.

Thanks for the update.

Bruce S

[Adriaan Kragten]

Rigid panels are better because as your panels open and close it requires drag energy to close that is working against your goals.

That isn't true. In my report KD 416 I have proven that the cup which moves in the direction of the wind supplies maximum power if it moves with 1/3 * V. However, the opposite cup which moves against the wind, moves with 4/3 * V for that condition. I have used half hollow spheres because this geometry has the largest difference in drag coefficient depending on the wind direction. It appears that the cup which moves against the wind consumes more power than supplied by the cup which moves in the direction of the wind if the speed of this last cup is 1/3 * V. So even if both cups have a large difference in drag coefficient, this difference is overruled by the effect of the difference in relative speed. To really get the maximum power for two moving cups, I found that the cup which moves in the direction of the wind must move with a speed of about 0.15 * V. So it is certainly useful to make that the cup which moves against the wind has less drag and the negative force needed to close the cups will be much lower that the negative force of cups which arn't closed.

[brandnewb]

I'll just keep dumping here my readings until I am all done and then perhaps if one is interested I can share the spreadsheet.

I am taking readings as best as I know how so I trust that the readings are accurate. But it is because that I trust the readings that I get confused when I see the following

a new pair of magnets stacked in isolation                  
id   field polarity   surface   measurement point   distance (mm)   milli tesla   remark
III   N   top   a   0   421   
III   N   top   b   0   434   
                  
                  
a new pair of magnets stacked in isolation on top of 13mm steel (on the back of the steel beam reads 0.6 S so I guess saturated enough?)                  
id   field polarity   surface   measurement point   distance (mm)   milli tesla   remark
III   N   top   a   0   515   
III   N   top   b   0   535   



I wonder why did the field not increase between magnets stacked in isolation on 5mm steel and another pair of magnets stacked in isolation on 13mm steel?
This time around I also measured the back of the steel and I think 0.6 mt means that it is close enough to being saturated?

maybe this steel beam is rather dirty? so more gaps between magnet and steel?

Any pointers one would have of course I would really like, and might even benefit all. You see I realised I can take readings while progressing with the build. So the cutting of corners is not strictly needed to avoid loosing time.

[brandnewb]

I must be doing something wrong.

I am getting no further than that stacking these particular magnets horizontally increases the field by around 30%

I really am going to try and stack vertically.

brace for impact!!!

[brandnewb]

BTW!! the second pair of magnets (the one with ID III) displays the same overtaking of the other magnet again on the sides I described earlier with the pair that decided against being subjects in my experiments.

The C measurement point has an N pole. D = S, E = S, F = S

This would make fine tuning the powder cups a lot of work and most like not a real return on investment.

So I am sticking to my gut and not going any deeper than 2mm with the submersion.

[Adriaan Kragten]

When you have two iron disks with magnets glued to the inside of the disks, like you have in a real generator, the magnetic flux in between two opposite magnets is almost homogeneous. So the magnetic flux lines are almost straight and you will measure about the same flux density independent of the distance to one of the magnets. But if you mount a magnet only on one sheet, the magnetic flux lines will be strongly curved and so some lines will be much longer than others which means that the flux density will be less as the flux lines are longer. This means that the flux density which you measure will depend very much on the position. So it is much better to create a magnet and steel sheet configuration similar to what you have in the generator and then measure the flux density in between the magnets.

If the thickness of the steel sheets is already that large that the iron isn't saturated, you will measure almost no difference if you make the iron sheets thicker.

[brandnewb]

Rgr sir, understood and thanks!

I'll press on!!

[brandnewb]

I respect that not everyone measured the field they use. Either by just not feeling the urge and or not having access to a meter.

But still I would like to turn the whole thing on its head by asking. What would be a good flux?

I know this is opening pandora's box because there are many factors at play.

But then is it really? if the most popular configuration is a dual rotor then I would like to know the flux.

I have seen countless designs on various sources and I always feel unsatisfied because I am left wondering. What was the flux.

Also what I am going against in this whole endeavor is that coils aren't stacked in the designs I have seen until now. Axial speaking that is.

So I am basically asking for a target flux that I would reach at the center of the dual rotor setup (magnets on opposite side of the coil).

I already think I have to reduce the coil height (else get no where with this setup powdered or not) and luckily I know I can while keeping 300 winds/600 turns. I think I can reduce by 20mm. so going from 38 to 18mm.

That would bump up the field enormously yet also create all kinds of engineering challenges I am not sure I am ready yet for.

[brandnewb]

So this time I have decided to dive a little less shallow than I did before.

I have added to the readings of the magnets III and IV measured in isolation and stacked. This time around I do not bother with steel. And for the time being also not yet with iron powder.

And unless I am making a huge mistake somewhere or things start improving in the next phase I might even decide to actually not stack vertically. You see the data I have do not show me yet an increase that is enough to warrant the investment of getting another 192 of those magnets.

An addition to the readings is now the angle at they were taken. I am not sure if it is of use, as I was already confident with my data earlier, but at least I tried. I am also preparing a test piece with the ability to measure inside of the structure. So even when filled with powder we can still take readings.


yeah I know. Not much to see there while weights are blocking the view ;( It is because I am aiming for a perfect flat finish during the drying of the glue holding in the magnet on the left.
And for the 3D print interested parties. The uneven tube going up with layers bulging at times is becuase I am cutting corners. I print functional parts not those meant to be pleasing for the beholder
One could change the nozzle size from this 1mm to 0.4mm and reduce the speed and things will look muchhhh better. But I just do not care at this point in time.


This is for the phase I am going in now. In where magnets are no longer in isolation but in a group of 3.

I start with a non stacked configuration and go through the motions again.

After that will be stacked without powder.

After that will be stacked with powder.

After that will be the addition of a counter pair of the same setup so that we can see the flux change as those counter pairs come closer and closer.

For those willing to see if they can spot obvious blunders here is the data I have.

the remark (taking these reading is not reliable because the slightest misalignment results in a reading of the opposite polarity) basically demonstrates that my earlier post about magnets being taken over is not demonstrated at all yet. So I now think it was poor experience taking measurements rather than a magnet being taken over.

a new single magnet in isolation                     
id   field polarity   surface   measurement point   distance (mm)   milli tesla   sensor angle   remark
III   N   top   a   0   310   0   0 = parallel, 90 = perpendicular to the surface
III   N   top   b   0   319   0   
III   S   top   a   0   308   0   
III   S   top   b   0   323   0   
III   N   top   a   10   58   0   
III   N   top   b   10   55   0   
III   S   top   a   10   58   0   
III   S   top   b   10   54   0   
III   N   side   c   0   213   0   
III   N   side   c   0   290   45   
III   N   side   c   0   235   90   
III   S   side   d   0   182   0   
III   S   side   d   0   277   45   
III   S   side   d   0   253   90   
                     
                     
a new single magnet in isolation                     
id   field polarity   surface   measurement point   distance (mm)   milli tesla   sensor angle   remark
IV   N   top   a   0   306   0   0 = parallel, 90 = perpendicular to the surface
IV   N   top   b   0   312   0   
IV   S   top   a   0   312   0   
IV   S   top   b   0   313   0   
IV   N   top   a   10   50   0   
IV   N   top   b   10   49   0   
IV   S   top   a   10   59   0   
IV   S   top   b   10   57   0   
IV   N   side   e   0   360   0   
IV   N   side   e   0   300   45   
IV   N   side   e   0   230   90   
IV   S   side   f   0   316   0   
IV   S   side   f   0   250   45   
IV   S   side   f   0   260   90   
                     
                     
a new pair of magnets stacked in isolation                     
id   field polarity   surface   measurement point   distance (mm)   milli tesla   sensor angle   Increase %
III   N   top   a   0   421   0   35.8064516129032
III   N   top   b   0   434   0   36.0501567398119
III   N   top   a   10   91   0   56.8965517241379
III   N   top   b   10   83   0   50.9090909090909
IV   S   top   a   0   431   0   38.1410256410256
IV   S   top   b   0   443   0   41.5335463258786
IV   S   top   a   10   93   0   57.6271186440678
IV   S   top   b   10   91   0   59.6491228070175
                     
III   N   side   c   0   350   0   64.3192488262911
III   N   side   c   0   371   45   27.9310344827586
III   N   side   c   0   284   90   20.8510638297872
III   N   side   d   0      0   taking these reading is not reliable because the slightest misalignment results in a reading of the opposite polarity
III   N   side   d   0   310   45   11.913357400722
III   N   side   d   0   307   90   21.3438735177866
                     
IV   S   side   e   0      0   taking these reading is not reliable because the slightest misalignment results in a reading of the opposite polarity
IV   S   side   e   0   333   45   11
IV   S   side   e   0   307   90   33.4782608695652
IV   S   side   f   0   334   0   5.69620253164557
IV   S   side   f   0   348   45   39.2
IV   S   side   f   0   301   90   15.7692307692308

[MattM]

So if you did three layers of magnets separated by iron discs between layers, would that show any difference?

[brandnewb]

I am not sure I follow Matt,

There a quite a few things to unpack there as I am now in a rather detailed frame of mind.

I can show that 800+ mt is doable



But not at all applicable and also please mind there is a similar array on the back that one can't see.

But please elaborate with more specifics and I am willing to give it a try if I have the means readily available.

[brandnewb]

So I have available the data in an array for single and stacked both without powder.





And the data that goes with it

field in between single magnets axially configured, magnet III being the middle one with the north pole facing down.                     
N/A   S   N/A   N/A   0   21.64   0   towards the test magnet that is in the middle
N/A   S   N/A   N/A   10   27.97   0   
N/A   S   N/A   N/A   20   26.19   0   
N/A   S   N/A   N/A   30   14.14   0   
N/A   S   N/A   N/A   40   6.75   0   
                     
field in between stacked magnets axially configured, magnet III being coupled in the middle by magnet IV.                     
N/A   S   N/A   N/A   0   24   0   towards the test magnet that is in the middle
N/A   S   N/A   N/A   10   36   0   
N/A   S   N/A   N/A   20   58   0   
N/A   S   N/A   N/A   30   34   0   
N/A   S   N/A   N/A   40   15   0   
                     
reading on stacked pair while in array                     
III   N   top   a   0   429   0   
III   N   top   b   0   449   0   
III   N   top   a   10   96   0   
III   N   top   b   10   92   0   

[brandnewb]

I guess that is serendipitous timing as this is the time of year it should snow over in my neck of the woods So I say let it snow!! Iron Powder

[brandnewb]

Ok now I go have a rest.

This is with powder. it holds up to what 13 mm steel does but I am still confused.



data
field in between single magnets axially configured, magnet III being the middle one with the north pole facing down.                     
N/A   S   N/A   N/A   0   21.64   0   towards the test magnet that is in the middle
N/A   S   N/A   N/A   10   27.97   0   
N/A   S   N/A   N/A   20   26.19   0   
N/A   S   N/A   N/A   30   14.14   0   
N/A   S   N/A   N/A   40   6.75   0   
                     
field in between stacked magnets axially configured, magnet III being coupled in the middle by magnet IV.                     
N/A   S   N/A   N/A   0   24   0   towards the test magnet that is in the middle
N/A   S   N/A   N/A   10   36   0   
N/A   S   N/A   N/A   20   58   0   
N/A   S   N/A   N/A   30   34   0   
N/A   S   N/A   N/A   40   15   0   
                     
reading on stacked pair while in array                     
III   N   top   a   0   429   0   
III   N   top   b   0   449   0   
III   N   top   a   10   96   0   
III   N   top   b   10   92   0   
                     
                     
reading on stacked pair while in array with compacted (At hand pressed force using wood ) powder that seem close to saturation..                      
III   N   top   a   0   516   0   
III   N   top   b   0   516   0   ? why no difference between a and b?
                     
N/A   N   N/A   N/A   0   45     0   N pole here?
N/A   S   N/A   N/A   10   16   0   confused by these readings
N/A   S   N/A   N/A   20   12   0   '’
N/A   S   N/A   N/A   30   8   0   '’
N/A   S   N/A   N/A   40   6   0   '’

If there is someone the can help me out of my misery Ed(MagentJuice) comes to mind.

I might just cancel the whole vertical stacking thing when looking at this

[brandnewb]

ahhh of course. while looking at some random stuff on tv I had this sudden stroke.

The reason the in between readings while with iron powder are so low is because there is now plenty of pathways shorter and with less resistance than where the probe is at.

Sometimes things need to settle a little and then uncoinsoisly it works all out.
That is of course only if I am not making a huge error in judgement here.

I am open to critique.

[brandnewb]

perhaps making both sides of the rotor hallbach configured would spice up thins enough to make things worthwhile?

[MattM]

I am not sure I follow Matt,

There a quite a few things to unpack there as I am now in a rather detailed frame of mind.

I can show that 800+ mt is doable

(Attachment Link)

But not at all applicable and also please mind there is a similar array on the back that one can't see.

But please elaborate with more specifics and I am willing to give it a try if I have the means readily available.

I was just figuring if alternating polarity of three layers (with iron laid between each magnet) increases or decreases the combined strength.

[brandnewb]

I have 2 sheets of 5mm steel.

so;
single non stacked magnet
--steel plate--
single non stacked magnet
--steel plate--
single non stacked magnet

if that is what you meant I can do the test later today.