Manual Transaxle oil cooler: FILTERING

[Sodo]

[ejimmi]

All im saying is there is no reason to take it too far.

I completely agree that with smaller filtration the more waste you will capture. The question is will there be a specific gain that warrants doing it?

Ok i gotta go out and work on the van before this thread consumes my day, Im replacing the bad bushing in my radiator fan with a skateboard bearing

[thasty07]

[Sodo]

PCforno that Magnom filter element is tiny! I am surprised to see it in real life. But it was built by magnet experts, as opposed to some forum jackass, so it probably works ! BTW you forgot to mention that you have 200HP running thru your trans (at times).

================

I thought my spherical ball filter was going to be THE BOMB, but it wasn't. When magnets cling together much of the magnetic field cancels out. If I were to arrange the spherical clump shaped like a hotdog, there will be a north and south end, but inside the array, where I wantd to filter the oil, there is not much field. So little, in fact, that gasoline could swish the ferrous particles off the magnets. And these are super-strong neobdymuim balls. Now I have 648 spherical magnets in my toybox. Not sure what to use them for. Would sure be nice if a magnetic fields expert would pop into this thread with some tips

For the next iteration I bought some 40x20x5mm rectangular magnets, these things are SUPER strong. The Weddle cannister inside diameter is about 44mm.

Here's a wooden mockup of the plan. Magnets are arranged in a spiral configuration in hopes that it swirls the oil past as much magnetic surface as possible. At 2gpm flowrate (= 7 cu.in/sec) the oil is flowing past the magnets at a velocity of 4.2 inches/sec. That's still moving along!

With brass screws and nylon spacers, I've isolated these magnets from each other.

Dropped the magnetic element in the Weddle cannister and closed it up. We'll see what sticks to it. Hopefully not much metal in there,... but catching EVERY ferrous particle. Whenever the oil stops (oil motionless) I bet the oil in the cannister gets pretty "clean". But what I've seen in the settling project I DOUBT the magnets can pull particles out of moving COLD oil. Possibly a few more particles can be pulled out of hot oil.

This is all an experiment. Anyway, this is plumbed into my van now, just have to wait awhile, stay tuned.

If anyone wants to try this or other schemes,,,,, note: These magnets are brittle like glass. They can attract each other 3 inches away and shatter upon collision! Also I tightened the brass screws a little too tight and broke a magnet. I started with 10 magnets ($21 on eBay from China) and broke 3, so buy extras. The brass screws are #8x2.5" long, and the spacers are 3/16 hole x 3/16" thick.
_________________


'90 Westy EJ25, 2Peloquins, 3knobs, pressure-oiled GT mainshaft, filtered, cooled gearbox
'87 Tintop w 47k 53k, '12 SmallCar EJ25, cooled filtered gearbox
....KTMs, GasGas, SPOT mtb

[Sodo]

More scribbling.....for those who are interested. And Re-stating the assumptions (for those reading on their phones) Please Focus on FILTERING metal contamination ONLY, thx!

ASSUMPTION: Excessive HP/torque generates metal fines in this little transaxle's oil. The premise is that transaxle life can be extended significantly by removing this contamination immediately (rather than driving on it for 1,000s of miles). For those who are using a pump and cooler, this is feasible.

ASSUMPTION: Using a mechanical filter for bigger stuff, PLUS employing magnetic + gravity for the steel particles smaller than the filter rating.

ASSUMPTION: excessive HP/torque is here to stay and we're trying to solve (or mitigate) its peripheral problems .

===============================

Got this idea from the settling proposal thread. I noticed a significant amount of settling in 3 weeks. So if you had a settling cannister on your van, and parked it for a week (or 2 or 3) , you'd have some pretty clean oil. Every "first-time" the pump starts, a quart of super-purified oil pumps thru.

At 2gpm, the velocity of oil pumping upwards is 15mm/sec, I bet that would carry the fines up & out but perhaps NOT the ones that sank to the bottom. With a cap at the bottom you could take a stool sample from time to time to assess your fluid condition. I think I'm going to have to build this unit!
_________________


'90 Westy EJ25, 2Peloquins, 3knobs, pressure-oiled GT mainshaft, filtered, cooled gearbox
'87 Tintop w 47k 53k, '12 SmallCar EJ25, cooled filtered gearbox
....KTMs, GasGas, SPOT mtb

[Merian]

your assumption neglects chemistry; contaminants can alter the additives chemically
_________________
....

[Sodo]

Wellllll... if anyone is still interested in one crazy old sap's closeup and verbose observations about the specific subject of "filtering transaxle oil".

This trans is my "hoarded trans". It was "freshened up" by a premier builder. He said he opened it up and it was in such great condition, and 100% OEM, he just "freshened it up" and put it back together. So it should NOT be expected to generate a lot of 'break-in swarf'. Maybe a little. I don't know whats new, if anything. Perhaps the mainshaft bearing, pinion bearing & syncros?

This trans runs cool. I had set my Cooler/pump/filter system microprocessor to operate at 150*F, but the trans ran so dang cool it almost never ran the pump. There was comfort from cool running but distress too, because - I_MUST_FILTER ! So I reduced the setpoint to 145*F and it runs more often, enough to satisfy me. I may have to reduce setpoint to 140* to 'filter' periodically as we go into autumn.

Here's an update at 4159 miles. Backpressure across the filter was increasing,,,, as if it was plugging up. This seems impossible, so I unscrewed the filter cup to have a look. I also have an inline magnetic filter BEFORE this filter, installed 10 miles ago so haven't checked it.

The 'dirty' oil enters the filter on the outside, exiting (clean) from the inside of the filter. So the stuff you see was outside in the cup that was rejected by the filter. Also draining down off the outside surface of the filter element.

Closeup. The stuff responds a little to a magnet, mostly just changes glint in the sunlight. A LOT of it looks like brass, which don't react to the magnet.

Heres a pic of dumping the clean oil OUT of the inside of the filter. Nice and clean, green. Since my 2gallon per minute pump system exchanges the entire volume every 30 seconds, I have to __assume__ the fluid inside the trans is "that clean".

I poured gasoline into the center of the filter to wash "stuff" out of the filter-fabric (like a gauze). This is a pic of the wash gas - it's BLACK. What the heck is this black stuff? At the magnet you can see a little line where the shiny ferrous metal fines have lined up. In a video of this; moving the magnet, the amount of steel fines becomes much more visible, bigger particles with steel dust billowing - I would post the video (4 seconds) but YouTube hates me. The stuff is floating throughout the wash gas, and settling slowly. Comforting that the filter is catching and holding it.

I don't know what to say about other transaxles, this is just one trans . 173 HP EJ25, cruising 75 MPH, 4.86, OEM 4th, 27,8" BFG tires. If other folks get interested and start observing at this level, posting pics etc; we may learn more, sooner. Agreed it takes quite the obsession to look at this level, but I'm enjoying it. Anyway theSamba needs entertainment for the hardcore otherwise they'll go elsewhere.

Ferrous metal fines can be referred to as "hardened steel grinding paste". That's how you think of it when pre-cleaning the parts in an ultrasonic cleaner for assembly, right? Its the enemy. You CANNOT SEE it, you just go thru all the witchcrafty motions to ensure only CLEAN parts go into the new trans. And then.... in the bottom of the solvent.... after washing "CLEAN" parts, you see black sediment. It exists, and its the enemy. Working with a fella who triple-filters his shop air, some of this obsession has clearly found a home with me. If a normal trans lasts 150-300k miles swimming in this stuff, I have a hunch that removing it as soon as it's generated, could extend transaxle life a LOT.

I tellya, if this "lifetime gear oil" thing wasn't foisted upon us by VW and bought into,,, hook-line-and-sinker by Bentley I think there would be pristine trannies out there, (after repairing broken 3/4 hub) with 4, 5, 600,000 miles. Funny this "lifetime" thing. A car that's wrecked at 50k miles had a "lifetime". Shortened, but still a lifetime, so they're not deceptive, it's just NOT how the antique vehicle enthusiast prefers to define it. It's not advice, rather its "self-fulfilling prophecy".
_________________


'90 Westy EJ25, 2Peloquins, 3knobs, pressure-oiled GT mainshaft, filtered, cooled gearbox
'87 Tintop w 47k 53k, '12 SmallCar EJ25, cooled filtered gearbox
....KTMs, GasGas, SPOT mtb

[Pcforno]

Hey Sodo- my 6 micron filter is in the line before my magnetic filter and it caught most of that black metal. The magnetic filter which I have to believe is quite good based on the design had some on it, but was extremely small. I totally agree with you. That stuff needs to get the heck out of your transmission if you expect it to last. Another interesting point- I watch my real time torque numbers as well as transmission temps and the only time torque really gets higher than low 100s is going uphill at speed. You can drastically reduce this by simply shifting into third and taking the hill at 60 rather than 70...

[Syncro Jael]

Sodo, I am leaning towards a settling tank with a magnetic rod up the center that would be attached to your stool sample cap. I would also add a "fill" cap at the top to refill the settling tank when you check it.

I have also planned to install the "Magnom" magnetic filter but still have not pulled the trigger. It is nice to see Pcforno using one and I was hoping to see if it caught some swarf.

As you can see in this video, here is some ferrous metal fines that the magnets around the Weddle filter caught. It was like a paste on the inside of the filter housing, but when the oil is cleaned out with gasoline, this is what was left.


Link

_________________
1987 Syncro Westfalia Hightop - NAHT
Subaru EJ25 Forged Frankenmotor, Triple Knob.
Jael = (Mountain Goat)

[Pcforno]

Some light reading on why filtration at the micron and submicron level is important and supposedly achieved by magnom filters

http://www.magnom.com/wp-content/uploads/2015/03/Magnom-White-Paper-IFPE.pdf

They have a lot more technical articles on their pages - it obviously makes sense that the swarf I found on my magnet was small and minimal as it was placed after the filter which I believe is the proper place- filtering large to small..

[Sodo]

SJael, I want to make a 1 quart settling chamber but need to find the right "donor scrap". Current plan is to put my clump of 648 ball bearing magnets in the bottom to create a non-turbulent quagmire so stuff that settles out doesn't get washed back into the system. But...where to PUT the dang thing?

[thasty07]

[Sodo]

I just read - again - SyncroGhia's excellent 2011 post (discussion with memeber SnowSyncro & others) regarding high loads passing thru a transaxle.

============

Further musings on metal fatigue as related to THIS topic of tiny metal fines; the spectre of Metal Fatigue can apply like this. (I think)

Think of gear face contact. It is a line contact between two oppositely curved parts. When all dimensions are perfect, shaft positions perfect, bearings tight, it is a rolling contact (on paper). It is perfect on paper, beautiful actually, zero friction, just like a ball-bearing.

At low power transfer there is no metal-to-metal and the only heat comes from smashing an oil film away. In real life there are clearances, looseness, thus no genuine line-contact. More looseness, and line contact becomes a "smear" which increases heat generation. Keep in mind higher heat causes a drastic reduction in oil "lifetime".

The engineer designed a gear face 20mm wide, for 90HP. But you have made your choice between driving enjoyment and component longevity, and are putting 180HP thru it with a smile. A gear face 30mm wide (line contact 30mm long) and 50% more teeth is required for that force. Basically a much bigger gear is REQUIRED. We don't have bigger gears. All we have to mitigate this is:

A few percent better oil
Cooling
Cleanliness

Transmitting low torque, the zone is narrow, and perhaps cushioned by the oil film, generally less than a micron thick. At high torque the zone becomes wider, the oil film is smashed. At higher torques the oil has vaporized to a puff of smoke and it's metal to metal for that moment. And your oil has incrementally lost some of it's "lifetime". Perhaps soot particles darken the color.

Lets say it gets worse..... the oil has delivered a few metal bits to this interface (it MOST certainly has!). But how big are the particles, and how much metal fines are polluting the oil? How many particles are ---on this line--- at this moment? In relative terms think of the oil film thickness as your head-height, ~6feet. A 3 microns particle is like the size of an apartment building.

So you are mashing these apartment buildings thru your gearface contact. A new apartment building, deforms the gearface at one tiny specific location. And the apartment building gets smashed too, flattened somewhat. Later on, its smashed further. Eventually it becomes like a bunch of flakes, and feels like a "metal paste" because all particles are rolled out, going thru gearfaces, and bearings, tens, hundreds of thousands of times.

Heres a pic of what flattened apartment buildings looks like. You cannot "feel" them it's like a paste.

I suspect that once a particle becomes flattened like this it's mostly benign in comparison because it's surface is large, flat, low psi, encapsulated by oil, with EP additives bonded to it. It's like an apartment parking lot. That's why you can run unicorn blood for thousands of miles.

But focus now on ---- that one spot --- that was previously deformed by that apartment building. The steel there was stressed beyond its fatigue limit. It can tolerate "more than a few" of these hard hits, if another apartment building comes along. But also keep in mind your transmission will release more than a few NEW, sharp apartment buildings into the oil in an on-going basis. Many are coming along, and there are millions of gear-faces coming together.

Statistically, this point will get hit again, and again. Hit-count increases, and eventually the fatigue limit of that spot is reached, and a new apartment building breaks loose from that surface and tumbles off into your oil, along with a cadre of smaller particles. See THIS website, it has a VERY GOOD diagram and very good, simple reading on the subject. They are selling a product, but.....still good stuff.

(pic is repeated into theSamba gallery in case nextgenfluids goes dark, pic has credit to nextgenfluids.com)

Now think of a ball bearing. Such as your mainshaft ball bearing, lonely up there, high& dry. Small ball, much larger race. Like a bowling ball sitting on your hardwood floor. Increase the force with a large engine that pushes apart on 4th gear teeth. An equal (too high) force is beared out by the mainshaft ball surfaces; far beyond what the engineers designed, like a bowling ball pressing into padded carpet. In that deformation,,,, heat in the metal is generated, and it starts conducting outward into the metal of the races. Heat in the oil, from being squished (or vaporized) reduces the oil film to zero. Fatigue cycles in the metal surface start "counting up" to their numerical limit. Now toss an apartment building or two into that interface.

At higher heat, it's worse, and oil is weaker.

Its been explained to me theres another situation in the VW van transaxle that hassles the mainshaft ball bearing. In 4th gear, the Mainshaft bearing's inner race is in direct, non-rotating contact with 4th gear (a thermally isolated idler). On long, fast highway driving, 4th gear gets HOT. Gear thrust presses it against the mainshaft ball inner race. Because it's an idler, separated from the shaft by the needle bearing the only way it cam transfer its heat out is to the mainshaft ball's inner bearing race. The heat reduces the bearings oil film thickness further. Trouble is brewing. This bearing is going to wear faster, let fines loose into the oil.

Metal is going into your gear oil in an on-going basis. If you REMOVE it - on an on-going basis - your OVERDRIVEN transaxle will SIMPLY run longer. I suspect MUCH longer if you arrive at a solution that captures these particles ASAP and remove them from the gear oil.

Well anyway I think most folks who are interested enough to read this far into some fool's musings on gearboxes may think this is friggin' obvious, but I was just sitting here with a keyboard in front of me and the desire to write something. So I hope you aren't too put out.... being FORCED to read this diatribe.... Also hope I'm not making mistakes or perpetuating myth, as I'm building understanding as I go, there can certainly be errors.
_________________


'90 Westy EJ25, 2Peloquins, 3knobs, pressure-oiled GT mainshaft, filtered, cooled gearbox
'87 Tintop w 47k 53k, '12 SmallCar EJ25, cooled filtered gearbox
....KTMs, GasGas, SPOT mtb

[Pcforno]

Thasty07 - I highly recommend Pegasus racing for the canton filter as well as mounting clamps and any AN fittings or hose. They have a lot of cool stuff

[Syncro Jael]

After reading much of this material, I pulled the trigger and have a Magnom Mini coming to install into the system. Thinking about how to make a magnetic filter while continuing to drive my Syncro just doesn't make sense. It will be installed after the Weddle Filter, inside the quick disconnects, before the cooling radiator, to catch the small stuff. I still like the idea of a settling bowl, but getting something in the line now will be saving the transaxle longevity.

Once I get it installed, run some miles, remove and inspect, I will return to this thread and post results.

I don't ever want to think of this stuff circulating through my box again!

_________________
1987 Syncro Westfalia Hightop - NAHT
Subaru EJ25 Forged Frankenmotor, Triple Knob.
Jael = (Mountain Goat)

[gears]

I think I'm going with a Magnom, too. Thanks for the info ..
_________________
aka Pablo, Geary
9.36 @ 146 in '86 Hot & Sticky
'90 Syncro Westy SVX
'87 Syncro GL 2.5
https://guardtransaxle.com

[Syncro Jael]

This came today and I will have it installed after work tonight. I will also be checking the Weddle Filter and report.

_________________
1987 Syncro Westfalia Hightop - NAHT
Subaru EJ25 Forged Frankenmotor, Triple Knob.
Jael = (Mountain Goat)

[Syncro Jael]

The best part about pulling the filter system is to "see" how much swarf the magnets and filter pick up.

It was another good maintenance day for the transaxle!

Here is all the swarf that the Weddle filter with rare earth magnets gathered up.

This is 3 months and 7,000 miles worth.


Link


I rubbed it across some white paper to show that this "paste" is very small particles of metal. I would believe this is normal wear for a gear box.

And here the Magnom mini / Weddle filter system with quick disconnects attached and ready to keep the transaxle clean. I installed the Magnom mini after the Weddle filter. With the quick disconnects you only lose a couple drops of gear oil and the filters can be inspected, cleaned, filled with gear oil, and installed back on the vehicle in a few minutes.

This transaxle was changed last October 18, 2015 at 70,850 miles. It now has 86, 561 miles or 14,711 miles on this rebuild. My last transaxle failed due to ring and pinion failure at 32,192 miles on the rebuild.

So now I am really inspecting the gear oil, changing it every year, and using a cooling / filtering system that includes the Magnom magnetic filter and Weddle filter, to capture the metal swarf in the gear box. Getting the metal out of the gearbox as it is cooled and filtered will prolong the life. I will be very interested to see how much the Magnom Mini collects.

My next gear oil change I will be switching to Swepco 202 full synthetic, now I am running Swepco 201. With all of these modifications I am sure I will get many happy miles out of this rebuild.

For those of you who don't know, my van has an EJ22. A rebuilt engine was installed Feb. 4, 2015 and now has 35, 674 miles on it. It still purrs like the day it was installed!

EDIT: I NEVER WANT MY GEAR OIL TO LOOK LIKE UNICORN BLOOD EVER AGAIN!

Link

_________________
1987 Syncro Westfalia Hightop - NAHT
Subaru EJ25 Forged Frankenmotor, Triple Knob.
Jael = (Mountain Goat)

[rmcd]

Just want to say big thanks for those boldly going before! We are all going to learn a lot from the scientific method you all are using. Thanks.

At some point I'll have a near virgin rebuilt tranny hooked up to a tdi and a right foot that has a thing for on ramps. Maybe I'll have the funding to be another test bed.
Thanks for your efforts.
_________________
VW LT40 build. Like a Vanagon but 30% larger in every direction and 40% slower even in metric.
https://www.thesamba.com/vw/forum/viewtopic.php?t=749359&highlight=

[Pcforno]

Very nice syncrojael. I have nearly the same setup with quick disconnects and the magnom. May I humbly suggest that the Weddle filter pore size is far too large and I would suggest a filter under 10 microns. I used a canton racing filter rated at 8 microns. The weddle filter is a whopping 200microns. IMHO it basically only serves as a prefilter