Regarding magnets on oil filters

[Hatsuwr]

When using multiple magnets on an oil filter, there common orientation is a N-S-N-S-N pattern. The magnets don't repel each other that way, and the magnetic field strength is somewhat higher than if you managed to install them with the poles facing the same direction. But there is an interesting arrangement called a Halbach array that can do even better. The Wikipedia page has plenty of information and illustrations: https://en.wikipedia.org/wiki/Halbach_array

I made a quick model of my filter using five .5"x.5"x1" N42SH magnets I have lying around.

Here is the legend for the plots below:

Here is the magnetic field strength with poles oriented the same direction:

Here is the plot for them installed with alternating poles:

Here is the plot for a Halbach array:

The magnetic field strength with the Halbach array is significantly stronger using the same materials, although installation is a bit trickier.

 

[ZeeOSix]

How to turn your magnetized oil filter into a black hole.

 

[UncleDave]

Never assembled a halbach with anything but bar or cube magnets, and I've never put a magnet like that on a filter.

I guess you could do it with disks but I've always preferred to try to space out multiple mags vs line them up together.

If I recall NSNS performed even better than halbach. (?) K&J used to have a good halbach page.

 

[Hatsuwr]

Never assembled a halbach with anything but bar or cube magnets, and I've never put a magnet like that on a filter.

I guess you could do it with disks but I've always preferred to try to space out multiple mags vs line them up together.

If I recall NSNS performed even better than halbach. (?) K&J used to have a good halbach page.

This is the closest comparison I could find relevant to this application: https://www.duramag.com/techtalk/halbach-arrays/benefits-and-drawbacks-to-using-halbach-arrays/

Magnet size is about right with their .5" cubes. Halbach array had a bigger advantage when the gap between magnets was larger. Their largest tested gap was 0.06". In my example the magnets would be touching on the edges closest to the filter, but about 0.15" apart at the outer edges. Seems like it would roughly average out similar to the 0.06" test.

 

[Mewbs]

When using multiple magnets on an oil filter, there common orientation is a N-S-N-S-N pattern. The magnets don't repel each other that way, and the magnetic field strength is somewhat higher than if you managed to install them with the poles facing the same direction. But there is an interesting arrangement called a Halbach array that can do even better. The Wikipedia page has plenty of information and illustrations: https://en.wikipedia.org/wiki/Halbach_array

I made a quick model of my filter using five .5"x.5"x1" N42SH magnets I have lying around.

Here is the legend for the plots below:

View attachment 247932

Here is the magnetic field strength with poles oriented the same direction:

View attachment 247933

Here is the plot for them installed with alternating poles:

View attachment 247934

Here is the plot for a Halbach array:

View attachment 247935

The magnetic field strength with the Halbach array is significantly stronger using the same materials, although installation is a bit trickier.

Kinda funny I was running a 360 degree magnetic coverage on an FE3600 I posted a week ago before my car got totaled.

Good news is I am now redoing this experiment again hopefully to see how well it holds up with some particle counts in future oil sample reports on my replacement vehicle.

Thread 'FRAM FE3600 - 3260 miles - RIP C&P'

Oct 25, 2024

This is the final oil filter that was on my Toyota Yaris 100 mile daily driver that got totaled in the rear by a large semi truck in the fast lane earlier this month.
Its quite unfortunate I wasn't be able to run this filter long enough to get the oil sampled for analysis as I had some running experiments with the magnets. Needless to say the magnets did its job quite well visually. The oil it was running was HPL SuperCar 5W-40, nothing else to say, everything looks clean on the inside...

...

• Mewbs
• Replies: 16
• Forum: Engine Oil Filters

• 

Post in thread 'Post your latest oil change'

Nov 1, 2024

2011 Toyota Yaris 3 door hatchback with 124k miles on the clock. This will be my new 100 mile daily driver, replacing my former totaled 2012 Yaris sedan that happened earlier this month. Did a double BG EPR engine flush with the old unknown jobber oil, followed by a Delvac ESP on the second flush. I know this is absolutely overkill though since its a used vehicle so I wanted to take the upmost precaution to get it highway ready for the months ahead without worrying.

What went out: Parts Plus Jobber filter PH2840

What went in: Purolator Boss PBL20195 as a flush filter for 2 rounds...

• Mewbs

• 

 

[Hatsuwr]

Kinda funny I was running a 360 degree magnetic coverage on an FE3600 I posted a week ago before my car got totaled.

Good news is I am now redoing this experiment again hopefully to see how well it holds up with some particle counts in future oil sample reports on my replacement vehicle.

Nice! What grade are those magnets by the way?

 

[Mewbs]

N42SH bars, 38 of them will fit on a 3" O.D. spin-on canister.

 

[Geekhead]

When using multiple magnets on an oil filter, there common orientation is a N-S-N-S-N pattern. The magnets don't repel each other that way, and the magnetic field strength is somewhat higher than if you managed to install them with the poles facing the same direction. But there is an interesting arrangement called a Halbach array that can do even better. The Wikipedia page has plenty of information and illustrations: https://en.wikipedia.org/wiki/Halbach_array

I made a quick model of my filter using five .5"x.5"x1" N42SH magnets I have lying around.

Here is the legend for the plots below:

View attachment 247932

Here is the magnetic field strength with poles oriented the same direction:

View attachment 247933

Here is the plot for them installed with alternating poles:

View attachment 247934

Here is the plot for a Halbach array:

View attachment 247935

The magnetic field strength with the Halbach array is significantly stronger using the same materials, although installation is a bit trickier.

What is the end result of this?

 

[BobsArmory]

I'm guessing the FilterMag would be considered a Halbach?

 

[UncleDave]

I'm guessing the FilterMag would be considered a Halbach?


View attachment 247991

I was curious about that as well, They don't state the arrangement they use.

 

[UncleDave]

You cant really make a circular halbach array without some sort of standoff device, but on the K&J site they have a great article

This is super interesting - could be 3d printed if someone were so inclined.


https://www.kjmagnetics.com/blog.as...xTrvfp3AwEqzXbXD4xWRF2FRZ7ixjbZZb9QjgqeIdNcA2

 

[jcy]

Please turn this post into a youtube video

 

[Hatsuwr]

What is the end result of this?

I'll be making a simple retainer for the magnets, and I suppose I could also do some pull force testing once it's done.

I'm guessing the FilterMag would be considered a Halbach?

Looks like alternating to me, but should be pretty easy for someone who owns one to test.

You cant really make a circular halbach array without some sort of standoff device, but on the K&J site they have a great article

This is super interesting - could be 3d printed if someone were so inclined.

That's the plan! I've already made the design, just need to get access to a 3D printer. The magnets will contact the surface of the filter and also the edges of the neighboring magnets (so pretty much like the FilterMag, but with thicker, shorter magnets). I'll add a channel in the middle of the backing material for a piece of steel to see if that makes assembly any easier, and that might also give a small performance improvement.

Sadly all the libraries around me seem to have broken their printers, so maybe it's time to finally buy one...

 

[UncleDave]

I'll be making a simple retainer for the magnets, and I suppose I could also do some pull force testing once it's done.


Looks like alternating to me, but should be pretty easy for someone who owns one to test.


That's the plan! I've already made the design, just need to get access to a 3D printer. The magnets will contact the surface of the filter and also the edges of the neighboring magnets (so pretty much like the FilterMag, but with thicker, shorter magnets). I'll add a channel in the middle of the backing material for a piece of steel to see if that makes assembly any easier, and that might also give a small performance improvement.

Sadly all the libraries around me seem to have broken their printers, so maybe it's time to finally buy one...

Looked around on K&J and my memory mostly served me correct (getting harder as I get older) both alternating polarity, AND all north pulls harder than a halbach and an MMR even harder than that.

source

https://www.kjmagnetics.com/blog.asp?p=halbach-arrays

 

[UncleDave]

I'm guessing the FilterMag would be considered a Halbach?


View attachment 247991

After spending some time on K&J I believe this to be an MMR.

 

[Hatsuwr]

Looked around on K&J and my memory served me correct (getting harder as I get older) both alternating polarity, AND all north pulls harder than a halbach and an MMR even harder than that.

That's for pull force on a steel plate that is in contact with the magnets. The scenario with an oil filter is different for several reasons - mainly in that we're interested in magnetic field strength on the other side of the steel plate, but the magnets also aren't in full face-to-face contact. The plots I shared in my first post are very specific to the oil filter application, and show the difference in magnetic field strength.

Also, for a somewhat more detailed examination of the K&J scenario, see here the link below that shows better performance with the Halbach array, especially so with increasing distance between the magnets.
https://www.duramag.com/techtalk/halbach-arrays/benefits-and-drawbacks-to-using-halbach-arrays/

 

[UncleDave]

That's for pull force on a steel plate that is in contact with the magnets. The scenario with an oil filter is different for several reasons - mainly in that we're interested in magnetic field strength on the other side of the steel plate, but the magnets also aren't in full face-to-face contact. The plots I shared in my first post are very specific to the oil filter application, and show the difference in magnetic field strength.

Also, for a somewhat more detailed examination of the K&J scenario, see here the link below that shows better performance with the Halbach array, especially so with increasing distance between the magnets.
https://www.duramag.com/techtalk/halbach-arrays/benefits-and-drawbacks-to-using-halbach-arrays/

Interesting, is your intent is not to allow the mags to touch the filter, or each other?

I looked at the attach - thanks - and I cant tell when they say "Gap" they mean air or the thickness of the plate.
Does the above model calculate the effect of having an air gap?

If you are going through all this, why not go all the way and build an MMR?

 

[Hatsuwr]

Interesting, is your intent is not to allow the mags to touch the filter, or each other?

I looked at the attach - thanks - and I cant tell when they say "Gap" they mean air or the thickness of the plate.
Does the above model calculate the effect of having an air gap?

If you are going through all this, why not go all the way and build an MMR?

No, the magnets will touch the filter, although the point of contact is mostly just a line, and there is separation caused my the paint/coating on the filter. The magnets would also be in contact along an edge with the neighboring magnets.

I think the whole idea of the MMR-style and other mounting magnets is a steel cup around the magnet. So it's the extra steel, not the orientation of the magnets, that gives the performance improvement. The same effect can be applied to magnet arrays by the way, but doing so around a circumference would be difficult with good tolerances unless you just wanted to laser cut and stack steel plates. The backing plate I mentioned achieves some of this effect.


Here's the design so far for the housing. It is sized for a 3" OD filter. This is my first 3D print design, so I wouldn't be surprised if there are some improvements to be made.

• It is two identical pieces that would be assembled with a few screws. Two pieces are necessary to allow it to be 3D printed easily.
• Nineteen .5" magnets actually fit almost perfectly around a 3" diameter filter housing, with less than a 1 mm gap between the last two.
• There is .25" of material at the ends and outer circumference.
• The groove around the circumference allows for a backing plate to be added if wanted, and saves a bit of material.
• The two parts are located to one another with a series of rails and grooves. The rails are 1 mm thick and high while the grooves are 1.1 mm thick and deep.
• Nine holes for #6 screws are located around the circumference, but I imagine four will turn out to be plenty.
• This design allows for nineteen magnets, but five is probably going to accomplish almost everything that nineteen would.
• The screw holes are probably going to be an issue with the small details around them, but this size allows me to test a few different things with this initial print.

 

[UncleDave]

No, the magnets will touch the filter, although the point of contact is mostly just a line, and there is separation caused my the paint/coating on the filter. The magnets would also be in contact along an edge with the neighboring magnets.

I think the whole idea of the MMR-style and other mounting magnets is a steel cup around the magnet. So it's the extra steel, not the orientation of the magnets, that gives the performance improvement. The same effect can be applied to magnet arrays by the way, but doing so around a circumference would be difficult with good tolerances unless you just wanted to laser cut and stack steel plates. The backing plate I mentioned achieves some of this effect.


Here's the design so far for the housing. It is sized for a 3" OD filter. This is my first 3D print design, so I wouldn't be surprised if there are some improvements to be made.

• It is two identical pieces that would be assembled with a few screws. Two pieces are necessary to allow it to be 3D printed easily.
• Nineteen .5" magnets actually fit almost perfectly around a 3" diameter filter housing, with less than a 1 mm gap between the last two.
• There is .25" of material at the ends and outer circumference.
• The groove around the circumference allows for a backing plate to be added if wanted, and saves a bit of material.
• The two parts are located to one another with a series of rails and grooves. The rails are 1 mm thick and high while the grooves are 1.1 mm thick and deep.
• Nine holes for #6 screws are located around the circumference, but I imagine four will turn out to be plenty.
• This design allows for nineteen magnets, but five is probably going to accomplish almost everything that nineteen would.
• The screw holes are probably going to be an issue with the small details around them, but this size allows me to test a few different things with this initial print.

View attachment 248115

View attachment 248118

Im intrigued, and impressed with your commitment to the endeavor.

 

[KevinMalone]

Not sure if completely relevant but check out Filtran's magnetic ATF patent: https://patents.google.com/patent/US6139737A/en

#18 is the magnet: