OK, I tested my visco clutch - when cold it would spin say 20x after the engine stopped, when over 100 degrees it would spin the same. Didn't take a lot to figure out the clutch did not work.

There still was some resistance against rotation but the resistance went away once the fan was turned a few times.

I took the clutch off yesterday and had a closer look. Everything seemed to be fine with it, when heated with hot air the bi-metallic strip would bend upwards and release a tiny rod, when cooler the strip would bend back down and press the rod inside.

I was lucky enough to have another visco clutch, non working that I put aside a few years back. I opened that clutch and collected all wax that was there. Drilled a hole in the first clutch, heated the wax (though the matter reminds me more of honey rather than wax) and poured it in. Made a thread, put a screw in the drilled hole.

Had it tested, it still rotates rather freely when cold but when over 100 degrees it rotated only say 8x instead of 20x. The problem now is that I have no bad clutch to open to collect the wax from.

Yet ... the wax when rubbed between my fingers reminded me strongly of one STP engine oil additive. The same colour, same consistency (being sticky and dense). I think I will buy a bottle today and pour some of it in the clutch. I can't make it much worse than it already is, but I can definitely make it better. Maybe. If not I will end up with having to buy a new clutch, which I would have to anyway had I not repaired it. If I succeed I will know a remedy for a shot visco clutch for a few $$$.

Btw - the rod operated 'valve' is no miracle - the rod, when in, presses on a metal strip which end covers a hole in a rotating plate. The cover is neither very tight nor very exactly shaped. Given the density of the wax it probably suffices though. When the bi-metallic strip bends up, the rod doesn't press on the metal strip and the hole opens a bit and the wax slowly gets beneath the second plate in the clutch causing more friction. I saw that the clutch could never really lock up, only increase its friction.

When the engine cooles down the hole closes (but as I say it isn't tight at all) and the wax that is there gets spinned off by centrifugal force and returns to a reservoir in the middle of the clutch.

Anyhow, will inject some of the oil additive in to the clutch and test whether it was a good idea.

Woohoo!!

I'm looking at doing this (along with the electric fan) myself...would love to hear your results. Any pictures, by chance?

Thanks!

Poured in a syrenge full of the honey stuff (oil additive, reduces engine wear, consumption, makes your wife prettier and other bull) and guess what? Although slightly tougher, it still rotates rather freely when cold but when over 100 degrees when I stopped the engine it only rotated some 1.5 times! Woohoo! Don't know how long for but it seems I have done it! A definite improvement from a free rotation before. Even if I have to change the fluid every half a year or so this is a good choice. The bottle was $4, the clutch works. At least for now.

Given this was an oil additive I suppose it should withstand friction wear and high temps. Ideal for use in a clutch.

Sorry, no digital cam, no pics.

Jassper,
Well done.
Out of curiosity I've often wondered how a viscous clutch fails.
Weak bimetallic strip over time?
Worn out entrance pin-valve?
Worn out exit valve?
Loss of silicone gel?
What are your feelings after having resurrected one with fluid/gel replacement? Is there a failure mode you would call primary? Or do you feel it's a compilation of several factors over time? I know your sample size is "one", but that's more than anyone else.

__________________
The Golden Rule

1984 300SD (bought new, sold it in 1988, bought it back 13 yrs. later)

To be honest I have had no experience with visco clutches until now. I was lucky enough I had one spare I could open up and study. The clutch basically consists of the body (2 disks, one pressed in the other and either glued or siliconed), the visco fluid - very sticky, white to yellowish coloured. I have no idea how much of the fluid actually is supposed to go in there. There are 2 metal plates (disks) inside that rub against each other. The viscosity of the visco fluid between them does that the clutch gets 'rotated' by the engine but not with the same speed- there is some viscosity slip. It's like you had 2 waffles with honey in-between. When you rotate one the other catches up with some delay, owing to the nature of honey.

Then, apart from the central bearing, the only part I remember seeing is the rod and the valve. The valve is very simple - there is a hole through the front plate with a piece of metal strip laying slightly above over the hole. When left as it is the hole is open as the strip rests some tiny distance over the hole. When the rod presses against the strip the strip is forced to lay over the hole blocking it.

Now, when the temp is low, the bi-metal strip outside presses against the rod that presses against the metal strip that closes the hole. When the clutch heats up the bi-metal strip bends up releasing the pressure off the rod and the metal strip inside raises to its original position over the hole thus effectively opening the hole. I suspect then the visco fluid gets rotated thru the hole and causes additional friction not only between the 2 plates but also between the plates and the front of the body of the clutch.

Now to your questions about possible modes of failure :

This (only valve inside, there was no other one) is a very simple design and I can't think of any way how the valve could fail. The clutch doesn't get opened and the only part that gets stressed seems to be the metal strip that closes the hole inside. It travels up and down, being pressed by the bi-metal strip. The travel is minimal and I don't think there is anything that can go wrong with it unless some drunko in the factory welded the strip on improperly.

There are 3 parts on the clutch that are either dynamically stressed or stressed by heat. It is either the pin thru which the visco fluid leaks with time (this was my case) or the central bearing which can either fall apart or leak or both. On my visco clutch I saw no other openings thru which the fluid could escape. There were those 2 only - the pin and the bearing.

In other words - IMHO there is nothing that can go wrong with the internal parts of the clutch because if its design. The only reason a clutch would lock up, I could imagine, would be when somehow the 2 internal metal plates (disks) somehow get stuck with each other because of the heat and friction after the fluid escaped.

Also the bi-metal strip outside could somehow fatigue with age and not bend (that way the valve wouldn't open). Though this is tested very easily - heat it up with hot air and see if it bends. When it bends it bends very visibly.

One way or the other, if the bi-metal bends and the rod travels up by itself it means you have only lost visco fluid.

Drill a hole (the outter part at the bi--metal, where the reservoir is), pour some high viscosity fluid in, seal the hole and the clutch must work. I filled my clutch full with the oil additive. I had no idea how much should go there, was only a guess. An original clutch stops after half a rotation, mine does after a rotation and a half. This is because the oil additive isn't as dense as the original filling but it still does its purpose well and when the clutch locks I can finally hear some roar by the fan.

So, to be concise, I think, after having the clutch apart, that the reason for 99% of visco clutch failures is a worn out entrance pin thru which the fluid escapes.

Jassper,

Too bad you do not have a digital camera. I still wonder what makes the fluid go back in the reservoir when the unit cools down. I was assuming centrifugal force sent it out but I cannot figure out why it obediently returns when the unit cools off. Especially if the entry port is blocked off by the bimetallic spring having cooled down first. Any idea? Since you have had yours apart maybe it is obvious to you. From here it sounds kind of peculiar. Thanks, Jim

__________________
Own:
1986 Euro 190E 2.3-16 (291,000 miles),
1998 E300D TurboDiesel, 231,000 miles -purchased with 45,000,
1988 300E 5-speed 252,000 miles,
1983 240D 4-speed, purchased w/136,000, now with 222,000 miles.
2009 ML320CDI Bluetec, 89,000 miles

Owned:
1971 220D (250,000 miles plus, sold to father-in-law),
1975 240D (245,000 miles - died of body rot),
1991 350SD (176,560 miles, weakest Benz I have owned),
1999 C230 Sport (45,400 miles),
1982 240D (321,000 miles, put to sleep)

"I still wonder what makes the fluid go back in the reservoir when the unit cools down."

No idea. Unfortunately the parts have been trashed and are no longer available. If you had the clutch open you would still wonder what makes it tick because there is almost nothing inside and I had to keep guessing about the function of the 'valve' because there simply is no obvious function to it. Only guessing how the circulation happens.

See the attached chart of the clutch. I am not a good painter, the drawing is yukky but gives you an idea how the clutch looks inside. You will see that there is nothing apart from the 2 disks, even the valve is very simplistic. It's so simplistic it is very hard to tell the flow.

' I was assuming centrifugal force sent it out but I cannot figure out why it obediently returns when the unit cools off. Especially if the entry port is blocked off by the bimetallic spring having cooled down first. Any idea? '

I think there must be a constant flow when rotating, once the valve opens the flows goes thru the hole, gets spinned off by centrifugal force, goes thru the hole again. I don't think the liquid goes there once and for good when hot, it slowly flows to and fro. When the engine cools off the rest spins off, returns but only between the 2 plates and not thru the hole because the hole is blocked.

It could actually work this way though thinking of it, when the clutch rotates the entire liquid should keep to outer circle of the clutch body. This would mean all liquid would accumulate on the outter circle and stay there until the spin stops. But the inside of the clutch body is grooved in a special way, I can think this has to do with redistributing the fluid so that it flows in a special pattern and does not accumulate.

'Since you have had yours apart maybe it is obvious to you.'

Not at all

Attached Thumbnails

 

The visco clutch will be "locked up" on startup, until the fluid is pumped back into the reservior by centrifugal action. This is the other quicky test -- should draw air on startup, then slow down.

Only three things every go wrong -- bearing locks up so the fan runs at engine speed, bimetallic spring breaks or pin sticks, or the fluid leaks out.

I'll remember this fix if I every get around to the 280 -- visco clutch is bad, and a new one is $350!

Peter

__________________
1972 220D ?? miles
1988 300E 200,012
1987 300D Turbo killed 9/25/07, 275,000 miles
1985 Volvo 740 GLE Turobodiesel 218,000
1972 280 SE 4.5 165, 000 - It runs!

Interesting fan clutch reading

http://www.nichols.nu/tip482.htm

__________________
1995 S-350
370K + SOLD
1952 220B Cabriolet
39K kilometers + SOLD
1998 E300D
310K +
2012 E350 BlueTec
120K

Well, now, I won't even have to drill a hole!

Don't use too much fluid -- if you put enough in there to overfill the reservior, it won't freewheel when cold. Of course, if it was leaking out, it will eventually work correctly, then fail again.

I'll be making a trip to Toyota soon. The 10,000 centistokes fluid will give the fastest fan speed, the 3000 the lowest.

Peter

__________________
1972 220D ?? miles
1988 300E 200,012
1987 300D Turbo killed 9/25/07, 275,000 miles
1985 Volvo 740 GLE Turobodiesel 218,000
1972 280 SE 4.5 165, 000 - It runs!

Peter,

Where is the reservoir? I keep thinking the reservoir is in the center and get confused on how the fluid returns to the center.

I also can't figure out how the air in the clutch, when heated, drives the fluid out of the reservoir, especially during the filling operation, when the seal is open (with the pin removed) and the air has the option to just leave through this hole. My experience with Mother Nature is that if you offer a means to comply with the basic laws of physics that does no work for you, that is the path that Mother Nature will select.

Also, if you just fill it up and seal the hole the "valve" goes through, the viscous operation will function more or less the way it was intended, but you won't get the benefit of the fan freewheeling under certain conditions, provided the fluid does not decide it would like to change phase at the elevated under hood temperatures. As long as the torque transmission of the clutch is similar to the design value, this should not be dangerous or significantly affect your fuel mileage. So why not just thread the pin hole after filling and Lock-Tite a small screw in the hole? I would guess actually filling the unit with fluid will make the operation of the bimetallic valve actuator moot anyway.

So, how does this fluid get back to the reservoir anyway? Thanks, Jim

__________________
Own:
1986 Euro 190E 2.3-16 (291,000 miles),
1998 E300D TurboDiesel, 231,000 miles -purchased with 45,000,
1988 300E 5-speed 252,000 miles,
1983 240D 4-speed, purchased w/136,000, now with 222,000 miles.
2009 ML320CDI Bluetec, 89,000 miles

Owned:
1971 220D (250,000 miles plus, sold to father-in-law),
1975 240D (245,000 miles - died of body rot),
1991 350SD (176,560 miles, weakest Benz I have owned),
1999 C230 Sport (45,400 miles),
1982 240D (321,000 miles, put to sleep)

"So, how does this fluid get back to the reservoir anyway? Thanks, Jim"

May have an answer from the 126 repair CD, job 20-310. Seems to be a spring and "plunger?" at the periphery of the clutch to allow the silicone back to the reservoir. Now to find a spare to experiment on.

__________________
The Golden Rule

1984 300SD (bought new, sold it in 1988, bought it back 13 yrs. later)

What happens is that the fluid is continously scavenged from the clutch region, where the two plates are close by the centrifugal pump cast into the hub. Its passive (no moving parts). When the bimetallic spring pops out and opens the hole in the reservoir (which is the cavity the hole is in), the fluid leaks out and spins down to the plate region.

The viscosity of the fluid then drags the fan. Maximum fans speed is determined by the viscosity of the fluid to pretty much the optimum speed for moving air. Pretty neat, keeps the fan quiet while premitting nearly maximal air movement potential.

With the pin in and the hole more or less sealed, most of the fluid stays in the center, so the fan more less freewheels.

If the fan is completely free with the engine off and cold, there is no silicone fluid in the clutch anymore, and you won't get any cooling at all.

The air in the clutch doesn't have anything to do with the operation -- and it is really a two speed clutch -- about 800 rpm and about 3000 rpm.

Peter

__________________
1972 220D ?? miles
1988 300E 200,012
1987 300D Turbo killed 9/25/07, 275,000 miles
1985 Volvo 740 GLE Turobodiesel 218,000
1972 280 SE 4.5 165, 000 - It runs!

Peter,

In the photos on the website provided by old south, the reservoir is in the center of the hub pretty much where I imagined it. I can follow how the clutch works once the "valve" opens. It is how it gets the fluid back into the reservoir (through the hole backwards?) when the engine is shut off and cools or cools down any other way.

Once the fluid gets out of the center, through the hole and is pumped by centrifugal force (the plates speed the fluid up, giving it angular momentum as they spin) to the largest diameter of the plates. In my view the fluid will just drain to the bottom of the housing when the engine is shut off. I cannot figure out why it would then climb back up a few inches and, presuming the "valve" has closed or started to close, seek out the little hole, now filling with pin by the force of the bimetallic spring, and squeeze itself back into the reservoir. Bosons, maybe? (edit: no, even bosons would not climb up to a higher level and stay there...)

I do not mean to be irritating but I cannot figure this thing out and would have expected the unit to work better if it just had a certain fill, kind of like a torque converter, and nevermind about these bimetallic or memory shape alloy "controls" that never last. My old dishwasher had one (more like 4 or 5 in its 20 year life) of those for dispensing the detergent. In the hot, humid environment inside the door, with electricity running through them to heat them up, they would corrode pretty quickly and stop working. I gave up and just poured the detergent in there for the last 10 years. Thanks, Jim

__________________
Own:
1986 Euro 190E 2.3-16 (291,000 miles),
1998 E300D TurboDiesel, 231,000 miles -purchased with 45,000,
1988 300E 5-speed 252,000 miles,
1983 240D 4-speed, purchased w/136,000, now with 222,000 miles.
2009 ML320CDI Bluetec, 89,000 miles

Owned:
1971 220D (250,000 miles plus, sold to father-in-law),
1975 240D (245,000 miles - died of body rot),
1991 350SD (176,560 miles, weakest Benz I have owned),
1999 C230 Sport (45,400 miles),
1982 240D (321,000 miles, put to sleep)

Jim:

It DOES sit on the bottom of the housing when not running. It will also leak out of the clutch if you store one on it's face rather than standing up, as my mechanic finally pointed out to Benz a few years ago. They no come with the box clearly marked "this way up or part will be damaged". The fan will spin at engine speed on startup, until the oil pumps out of the working region. This means, of course, if you don't get ran "roar" on startup, the fluid is gone.

I'll dig up the W108 manual tonight and see if I can give a better description of the pump, I'm not too clear on it myself.

Peter

__________________
1972 220D ?? miles
1988 300E 200,012
1987 300D Turbo killed 9/25/07, 275,000 miles
1985 Volvo 740 GLE Turobodiesel 218,000
1972 280 SE 4.5 165, 000 - It runs!