Viscous fan clutch blues?

[bengal taiga]

I replaced my Behr viscous fan clutch 6 weeks ago - since it only freewheeled when it warmed up. The replacement Sachs clutch worked fine for a few days and then exhibited the same symptoms (Engages when cold - but freewheels when hot). I returned the clutch and obtained another replacement and was told by the vendor they will not continue to "play games with me" over fan clutches. With the temp gauge in the red, the radiator hoses ballooning, and the engine pinging - demonstrating true overheating issues, I can stop the fan from rotating with only finger pressure. Ergo, this fan clutch is now a useless appendage.

I expect to be adding an electric fan from Summit as a stopgap measure. But with over 25 years of e9 ownership as well as years of owning other cars running viscous fans (Mercedes) I don't believe I should be forced to adapt an electric fan at all. I have researched the issue and noticed several people suggest the viscous clutches can be filled or even refilled with silicone oil to resurrect them. The problem is that these newer clutches do not evidence any leaking oil. Then, there is the issue of grades of oil, oil compatibility, and if there are leak, replacing the leaky seal/s.

Is anyone personally familiar with this process? If so, how would you reattach the bimetallic strip that expands as a valve presumably redistributing the oil to engage the clutch? The Behr uses some sort of putty while the Sachs employs what appears to be a post or rivet connection? What about adding some other substance to permanently lock the clutch (similar to running longer bolts in the original friction fan clutch set up)? Does anyone have or can they point me to a picture or exploded view of the viscous clutch innards? This inquiring mind wants to know!

TIA

 

[sfdon]

As an emergency fix on the road you can simply pop off the strip and the clutch should engage. Access to a drill and a sheet metal screw will allow you to manually adjust the fan clutch tension with the strip still on. HOWEVER- if it was me having this problem I would look to a slipping fan belt, bad pulley and most importantly, I would use a laser temp gauge to make sure that the radiator was not running a large cold spot in front of the fan clutch. Before returning the new fan clutch you should try it in another car and verify it is faulty. BTW- you did check to see if BOTH the upper and lower radiator hoses are hot??? If the lower hose is not hot then you have a faulty thermostat!
Good luck!

 

[bengal taiga]

viscous fan clutch blues

As an emergency fix on the road you can simply pop off the strip and the clutch should engage. Access to a drill and a sheet metal screw will allow you to manually adjust the fan clutch tension with the strip still on. HOWEVER- if it was me having this problem I would look to a slipping fan belt, bad pulley and most importantly, I would use a laser temp gauge to make sure that the radiator was not running a large cold spot in front of the fan clutch. Before returning the new fan clutch you should try it in another car and verify it is faulty. BTW- you did check to see if BOTH the upper and lower radiator hoses are hot??? If the lower hose is not hot then you have a faulty thermostat! Good luck!

The radiator is clean and has been pressure tested. The thermostat clearly opens as one can see the rush of fluid in the reservoir when things get to operating temperature. Both radiator hoses are hot, the hoses are fully expanded and the gauge visits the red. The water pump is no more than 9 months old. The coolant is the gold Mercedes type.

There is no doubt the fan clutches have been bad. If the car is moving, the temp gauge goes down to normal. Again - the fan clutches have worked when the car is first started and there is no doubt due to the rush of air it produces. BUT, after running a few days the clutch does not engage when hot. Although I have not fully disassembled a fluid clutch, I have pried up on the strip and wedged a shim underneath it and found no difference. I have also heated the strip with a propane torch to draw the "pin" out - also without any appreciable difference. I noticed a less expensive Chinese made clutch that bolts onto the fan. It has a coiled bimetallic strip that appears, at first glance, to be a better design that the sachs or behr units. The "triggering" pin turns laterally rather than longitudinally. I may try it.

I am fascinated by your suggestion of being able to adjust the tension of the stock viscous fan clutches. Are you suggesting the pin be replaced with a sheet metal screw? How far do you screw it in? Is this just intuitive or do you have any specific instructions? Is there concern with silicone oil leakage?

In my particular situation, I am unable to simply uncouple the fan clutch and swap it out because it is too close to the radiator. Instead, I have to drain and remove the radiator to access the fan clutch. If this were not the case, I would give used clutches a shot.

In the meantime, I have an electric fan that I am thinking of installing as a manual backup.

Your suggestions are appreciated. If you have any more, please post them!

Thanks

 

[sfdon]

Drill the hole to the left or right of center and use a sheet metal screw that you have ground off the point.

Have you inspected the condensor and radiator for airflow [not full of bugs and dirt]?

Why isn't your auxillary fan clicking on?

There should be no " rush of fluid in the reservoir when things get to operating temperature" in fact the pressure should be constant or decline as the thermostat opens. Keep in mind that if the nipple at the radiator or at the expansion tank is plugged for the small hose you will have an overheating problem.

You really need to use a laser temperature gauge on your radiator also.

 

[bengal taiga]

viscous fan clutch blues

Drill the hole to the left or right of center and use a sheet metal screw that you have ground off the point. Have you inspected the condensor and radiator for airflow [not full of bugs and dirt]? Why isn't your auxillary fan clicking on? There should be no " rush of fluid in the reservoir when things get to operating temperature" in fact the pressure should be constant or decline as the thermostat opens. Keep in mind that if the nipple at the radiator or at the expansion tank is plugged for the small hose you will have an overheating problem. You really need to use a laser temperature gauge on your radiator also.

I wondered whether you were advising that a screw should be used to replace the center pin because it seems (at least on the behr unit) pushing the pin inward increases the clutch resistance/engagement. I have not tested this idea in the car with a hot radiator, but I used a brass machine screw at the bench and the clutch tightens up. I am still wondering how much the action of the clutch is dependent upon any heat absorbed conducted from the waterpump shaft in addition to convection from the radiator. I ponder this because I have noticed with very little air being moved by the fan at idle, the clutch still engages and this would discount the notion that the bimetallic strip is critical. I also notice that the older style viscous clutch (red blades) does not have the bimetallic strip and the two that I have on the shelf, worked fine (I cannot use them without changing the nose of the waterpump).

Since my last post I have swapped the Sachs fan clutch out for a made in China version. I figured maybe they have made a few improvements with the coiled element that seems similar to other asian fan clutches. After driving for the better part of an 3hrs running errands, as soon as the gauge crosses the 1/2 way mark, the fan clutch engages. The last new fan clutch did this when it was first installed but quit within two days. This clutch seems to be working fine- (fingers crossed).

As to your follow-up considerations, the radiator, although not new, seems very clean (no noticeable mineral deposits). The fins are clean enough to eat off of. In fact, the radiator does such a good job that the fan seems to be an afterthought for hot days! There is no auxiliary fan - no air conditioning. That is the one thing I was planing to install although this particular car has never had one in the 20+ years I have owned it.

I haven't pulled the thermostat recently, but have not seen any need to considering every thing seems fine if the fan clutch would only engage when things get warm in stop and go traffic. I think I last checked it this past year when I replaced the water pump (which suffered from bearing noise). The heater fan recently stopped working and this could prove helpful as an additional heat reducer, but I do not think I have ever used it other than to play with it on a rare occasion.

One last thing, it has been my experience with E9's, E-3's and even E24's that when the thermostat opens, more coolant flows through the overflow tank. I understand the coolant is flowing constantly, but there is a discernible difference when the thermostat opens. By the same token, if I lift the cap on my 02, there is no discernible flow until things warm up and the thermostat opens the port. As an aside, using the phosphate/silicate free gold coolant seems to have made a difference in keeping things clean and even (possibly) cooler. But since I change things regularly, I probably could get by with just about anything.

Thank you for your thoughts.

 

[Arde]

I still have the fixed clutchless fan and read a number of times about converting to the viscous clutch. Following this thread and reading some more about how the clutch works I am surprised that the failure mode of this device is to get softer rather than harder. One would think that the engineers would prefers a fail safe mode that doesn't lead to engine overheat.

Maybe it is time to create our own e9coupe upgrade here.

 

[bengal taiga]

I still have the fixed clutchless fan and read a number of times about converting to the viscous clutch. Following this thread and reading some more about how the clutch works I am surprised that the failure mode of this device is to get softer rather than harder. One would think that the engineers would prefers a fail safe mode that doesn't lead to engine overheat.

Maybe it is time to create our own e9coupe upgrade here.

I have shared the same observation and wonder if the clutch were run dry or filled with plain (corrosive) water or just epoxy - if the clutch wouldn't default to a locked condition. I liked the simplicity of the early style friction clutch - except when it was worn. Nevertheless, I presume that the majority of the viscous clutches perform satisfactorily - despite my spate of bad luck. Otherwise, there would be many more cars on the side of the road.

It probably helps, at least with newer cars, that clutch activation by hot air directed from the radiator, may be more concentrated due to fan shrouds and maybe even thicker radiator cores. I wonder what effect, if any, is the heat conducted via the waterpump - especially since the newer cars probably run hotter normal temperatures than our E9's.

In the case of the subject E9, the radiator is "relatively" new old stock brass and even though it is clean and unobstructed, one can easily look through the grill and at the front of the engine through the radiator fins. Since I have never had any real overheating issues with this or any other radiator on E9's or E3's, my theory probably doesn't hold much water.

For the record, the Chinese clutch is still working fine, but I have a nice big flexalite electric fan, if necessary. :wink:

 

[Honolulu]

B Taiga, just a thought

Critical to your situation, so far unmentioned, and a prime suspect, though you evidence enough mechanical familiarity to have addressed it (sheesh I sound like Murray)

Having boiled several times, when you refill coolant, have you adequately bled the air from the system? Yeah, I realize the f-clutch is showing sure symptoms of failure, but you GOTTA bleed the system with the engine fully warm, parked nose up on an incline, at least once, usually two sometimes three times (short run between).

Your parts shop should not give you ANY static about "games" when you can clearly demonstrate the clutch has failed. On their side, you probably can't/haven't driven over to demonstrate, and it isn't their business to diagnose your cooling system woes.

 

[sfdon]

Bleeding the M30 engine can be tough- Thats why I drill every thermostat I install and never try to diagnose an overheating problem without removing and drilling a fresh thermostat. Keep in mind that a thermostat should be "clocked" inside the housing for maximum flow depending on which thermostat cover you have. Once the thermostat is drilled the system is self bleeding. A few years later the new design thermostat came out that was self bleeding.



http://www.autohausaz.com/search/imageexpand.aspx?imageurl=../secure/PartImages/11531729841.jpg [/img][/url]

 

[Tierfreund]

Could you maybe elaborate on the drilling and how it is to be done and why then the system is self-bleeding?
Also what are the disadvantages?
Sorry to be so demanding, but i find bleeding the E3/9 very difficult but essential, so that info might be of very good use to me and probably many others...

 

[bengal taiga]

Re: B Taiga, just a thought

Your parts shop should not give you ANY static about "games" when you can clearly demonstrate the clutch has failed. On their side, you probably can't/haven't driven over to demonstrate, and it isn't their business to diagnose your cooling system woes.

I would invoke Pele at the Part's counter, except they would wonder how the former Soccer great is when he gets offended. :wink:

The bleeder screw has seen so much use recently that I may have to fabricate a bearing surface to mitigate the wear. With the "Chinese-made" fan clutch engaging far more often then its predecessors, I am also thinking about purchasing some new rain gear for the shower bleeding can produce! Seriously, it is a minor inconvenience that is easy to overlook.

On a related note (SFDon) I have a collection of thermostats - many which I have drilled. I used to routinely drill a pilot hole or two on just about every thermostat that touched my grubby paws. This included even the awkward-to-drill 02 thermostat. I still have mixed feelings about the process since the drilling may provide a "little" insurance against thermostat failure. The trade off is the creation of a lazy thermostat situation. This may cause the engine to take longer to reach operating temperature and while this is not a great cause for concern, it arguably burns more fuel and causes a bit more wear on moving surfaces. Conversely, it may also give the operator a false sense of security when the thermostat is actually failing. Fortunately, this happens infrequently or less frequently than my experience with fan clutches.

Regarding the 02 thermostat, my understanding is that it actually acts in reverse to the E9 setup so that drilling a hole may be good for purging air from the internals but it is counterproductive to helping avoid a boil over. The weird thing is that I have never had to bleed the 02 system and the manual does not even mention this.

I am using the original head on the subject E9 and am mindful of its reputed penchant for overheating. I have never boiled over, but the pinging, swollen hoses and temp gauge gave me every indication I was close to it. I have unfortunate experience with an E3 which never indicated a temperature problem and the fan clutch always seemed to work fine. That car had a great heater and an auxiliary fan for the air conditioning too. It became increasingly harder to start and one day (XMAS Eve) I pulled the plugs and found coolant in one of the cylinders. I drove it 50 miles home - without overheating (per the gauge) and without any impressions of a Stanley Steamer. Running cooler did not prevent the head from coming off Christmas day. It still makes me think seriously about going air cooled.

Thanks to all

 

[sfdon]

The need for bleeding any M30 engine is that as water goes into the top hose and fills the radiator it creates a air pocket at the top section of the lower radiator hose where it connects to the thermostat housing. By drilling a 1/8" inch hole on the top flange of the thermostat you allow the sytem to replace the air pocket with water. Later thermostats have what is called a "jiggle valve" and some even have two. Most vertically mounted thermostats need to have this opening at the top of the thermostat.
Hope this helps and of course you can google "jiggle valve" for a better explanation.

 

[pmansson]

On a close enough matter: are the viscous fan clutches to be trustworthy from a balancing point of view? One one car the fan and clutch are not spinning evenly. It is a recent upgrade/installation.

 

[sfdon]

Did you change to a new fan? There are problems matching fans to clutches and if your combination is not correct you will be replacing your water pump soon.

 

[pmansson]

But the fan came from W&N attached to the viscous fan clutch....
I have heard of badly balanced such units, and will now check this carefully. Know what you mean about the water pump.....
Don´t want that to happen !!!

 

[bengal taiga]

Update

But the fan came from W&N attached to the viscous fan clutch.... I have heard of badly balanced such units, and will now check this carefully. Know what you mean about the water pump.....
Don´t want that to happen !!!

The steel double row pulley on my water pump has two tack-welded balance weights, so there may be something to what you have heard. I guess, depending upon the quality of their original construction, they could rust or be mishandled and fall off resulting in balance issues. Like clutches, I have had waterpumps with lengthy service lives and some that died at birth or soon after. I have not attributed any failures to un-balance rather than poor construction (bad seals) since most of the pumps used the same rotating parts (fan,clutch and pulley).

I would expect most unbalance situations could be observed and felt from the first run - unless the fan were to sacrifice some of its blades. I tend to think running a belt with too much tension can be even more destructive to pump bearings and seals.

As an update, the Asian fan clutch is working just fine. Every day - you can actually hear and feel it engaging (similar to turning on an electric fan), something I barely noticed with the Behr and Sachs clutches except when they were first installed and cold. I still have one on my desk. It looks perfect and you can feel plenty of resistance when it is turned. For me, the proof has been in the pudding and if they stop engaging when the need and temperature (a)rises, they are nice paperweights. :roll:

 

[sfdon]

a picture is sometimes better than words:

http://www.kfz-tech.de/ViskoluefterZ.htm

 

[bengal taiga]

Zer gut

a picture is sometimes better than words: http://www.kfz-tech.de/ViskoluefterZ.htm

Thanks for the link.

Figuring out how the standard (upgraded) viscous fan clutch is supposed to operate is easy, but pre-diagnosing a failed clutch is a puzzlement. Since the clutch fluid is clear, a leak may easily go undetected. Even on a new clutch, damp or discolored packaging may be discounted as some sort of metal preservative, although it may really be leaking clutch fluid.

Removing the bimetallic strip on a Behr clutch reveals the brass pin discussed above. It fits within a soft silicone seal. It is easy to see where this seal could leak if the oil was permitted to pool in this area during storage. Because the pin and the seal around it, are centered, when the clutch is spun during normal use, leakage from this area is minimized by the centrifugal force on the oil.

The more likely culprit for oil leaks would be the outer perimeter of the clutch housing. Such leaks might also go easily undetected, not just because of the fluid's clear color, but also, because of the velocity at which such oil would be slung. One of my disassembled clutches has a pressed together housing with what appears to be a very thin elastic looking gasket material that reminds me of electrical tape. This construction explains why a poorly manufactured and/or mishandled clutch could develop a micro leak and have a premature failure. A semi-analogous situation might be the seals used around a watch crystal or a watch’s press-fit back cover although watch seals are ordinarily not designed with silicone oil retention in mind. Detecting a leaky clutch, before installation on the waterpump, could be difficult if not impossible.

Some recommend adding fluid (silicone oil) with a syringe by removing the brass pin. The problem with this is knowing how much fluid to add and then affixing the bimetallic strip that retains the pin and actually regulates its movement. As simple as this design seems to be, the Chinese-made clutch has a coiled metal strip on the front of the housing that resembles the coiled metal strip found on mechanical thermometers and thermostats. Having observed them on Asian cars, the design is evidently proven. I have not taken one apart (my only example is still working fine) but the valving of the fluid is presumptively the only difference between this style and the European favored design. I suspect the seals are quite similar and equally prone to leak, if the clutch is mishandled.

In retrospect, I cannot state the Behr and Sachs clutches I tried all failed due to fluid loss or leaks, but I think it a good bet. Each clutch looked new and pristine, but I vaguely recall cleaning each one because of what I thought was an oily film covering them. If this was leaked silicone oil, I cannot say for sure. Even assuming this was silicone oil residue, I am at a loss to explain why the clutches would seem to work fine for a short while and then stop - without any apparent warning. This may have been due to low oil, oddly displaced oil from the factory or some other unnoticed phenomena. Parenthetically, I wonder if heat transferred from the pump to the attaching point at the rear of the pump isn't another factor in clutch operation.) If it was lost or low fluid, whether this was due to shabby handling, poor design or even a short fill at the factory is anyone’s guess. Maybe there are things such as bad batches and jopefully, I discovered all of them making this thread an abberation!:roll:

 

[sfdon]

I think you are onto something regards the leakage- I have been told by knowledgeable folks to never buy a clutch that has been stored flat on the shelf. I did pick up a e32 water pump and fan clutch [much bigger and better design] a few moths ago and will see if it fits.