Troubleshoot a 617.952 cooling system (84 300TD turbo)
 

The problem: The car runs on the hot side, near 100 C. The operating temps actually vacillate between ~92 C and 100 C, during normal driving (no heavy loads, AC off, ambient temps around 75 F). I know that this is within acceptable range and is considered normal by some, but in comparison to my 85 300TD it is 10-15 C hotter. I bought the car a couple months ago, but haven't driven it more than 500 mi yet (I've been working on it). It was filled with green coolant and had this problem when I bought it.

Previous work:
1) I degreased and citric acid flushed the system. It is currently filled with 45% Zerex G-05:distilled water. There was ALOT of scale build up prior to the acid flush. All flushes were done with the thermostat out, heating the car up to operating temp for 15-45 min while running the defroster, pulling the radiator plug - block plug - temp sender and upper radiator hose and flushing with a garden hose.
2) The thermostat was replaced with a new 80 C Behr unit, installed correctly. I have tested it (in the thermostat housing, out of the car) with boiling water and it extends fully to block off the radiator bypass, but there is no way to tell if it opens "fully".
3) I have tested the temp sender/dash gauge by grounding the temp sensor and immersing it in 90 C water next to a laboratory thermometer. It reads true within a couple degrees.
4) I removed the radiator, filled it with boiling citric acid and let it sit over night. I power washed it at the local car wash from both sides to clean any dirt and debris. I am fairly confident that it should have good coolant and airflow. There are very few bent fins.
5) I have swapped the coolant tank cap on both cars, the system does pressurize when hot and hold pressure overnight.

The questions:
1) How is the fan clutch supposed to work? When the car is cold and I shut off the engine the fan stops immediately. When it is hot the fans spins a few extra revolutions. This is true of both cars. In direct comparisons the fan clutch on my 85 300TD (which runs cooler) has more resistance when spun by hand. Not a huge difference, but noticeable.
2) I have seen posts stating that the water pump impeller blades had worn down. How is this possible?
3) Other than swapping parts between the two cars, what other diagnostics can I do?

EDITED on 07/24/07: For the record it was the thermostat, but please read all the great suggestions.

The water pump blades wear down due to erosion. The metal is harder than water, but water resistance "gets harder" the faster something tries to move through it. It sounds like a combo thing to me, weak water pump and weak fan clutch. Also, make sure the fan shroud is in place. HAve you tested the temp sender? Sometimes they can send a too high or too llow reading to the guage. The gauge an also read wrong. Have you used a non contact thermometer?

What do you mean by "holds Pressure" overnight? Is that from running the engine or from you manually pressurizing the system? It is a bad sign pf pressure stays overnight after running the engine.

yup. what he said.
holding pressure overnight that YOU pumped up with a pump is a good thing.
holding pressure overnight from the engine running is a bad thing that indicates a cracked head or blown gasket...
pressure should build up due to heat of expansion and dissipate when the engine cools.

rrgrassi,

I am certain that the dash gauge is reading true (or darn close to it). See previous post 2nd paragraph #3.

I also did the unscientific (but free) test today of holding the upper vs lower radiator hoses after a 20 minute highway drive. Upper hose was almost too hot to hold, lower hose was only warm. Despite the qualitative nature of this test, it clearly indicates that the radiator more than sufficiently dissipates heat and therefore eliminates concerns about the radiator and fan clutch.

Oh bugger.

So what's the next diagnostic step? compression test?

not sure, but you could start with releasing all pressure in the cooling system, and put a gauge on the cooling system, start the car and watch the gauge. how long does it take to build up pressure? how much pressure does it build? instant pressure spike, blown gasket. pressure jumps over 20 psi in 5 minutes? cracked head... sorry.

False Alarm
 

I think so anyway (fingers crossed).

I must have confused myself about when the system was pressurized. The last two mornings (and after work today), I removed the expansion tank cap without any discernible pressure release. Maybe, MAYBE, a slight gurgle from the upper radiator "overflow" hose.

I hooked up the mityvac as a pressure gauge to the expansion tank (via a #6 one-hole rubber stopper) and tested that it held pressure up to about 8 psi (for a couple of minutes at least). I released the pressure, then started the cold engine and watched the pressure gauge for about 8 or 9 minutes. There was at most 1-1.5 psi.

I see no evidence of coolant in the oil, oil in the coolant (though I have not had either analyzed), coolant loss or white smoke/sweet smelling exhaust. So, as of now I think I can put horrible thoughts of a leaking head (gasket) out of my mind.

More info: It appears as though the car will only run hot while being driven (locally or on the highway). If I start it and let it idle it will run at around 85 C. If I park it while it is around 100 C and leave it idling, it will eventually return to ~ 85 C.

Next up I will pull the water pump and inspect it. Gaskets and time are cheap.

Now on to the part of every repair job, where I break something:

While I had the mityvac on the system, I decided to try to pull a slight vacuum. I was kind of thinking that I might be able to pull any trapped air out of the system (OK, so it was not exactly clear thinking). I could hear the system sucking air from somewhere in the front and bubbles in the radiator. I tried to pump pressure back in to the system and now there is a small leak somewhere in the radiator. The fan sprayed the coolant around. Doh!

IIUC the water pump fails when you start to weep water out of the built in hole, can't fathom how you can wear down impeller. I've snapped off a hold down bolt on water pump, be careful! You found a hole in rad, correct?

That occurs when the shaft seal fails, which is totally different than the impeller failing.

Fan clutch, fan clutch, fan clutch! swap clutches to verify. Years ago we would test a clutch by blowing compressed air against normal rotation thereby stopping the clutch with engine running. Once engaged, the air could no longer hold it. Check temperature on dash when you can no longer hold the fan with air.

I also have a very hard time imagining how the water pump impeller could be worn down. But with the amount of pitting and scale I found on the thermostat housing (around the hoses), I can't rule it out. But the t-stat housing is aluminum and the water pump is iron, it still seems unlikely. The only way I'll for sure is to pull it.

Besides, at this point, insufficient coolant flow seems to be the most likely explanation to me.

I think I actually created a hole in the radiator, or unplugged one to be exact. I'm certain that this hole in the radiator has nothing to do with the car running hot.

Its also on the list, right after the water pump.

To check the waterpump I have to remove the fan clutch anyway. If the waterpump shows no serious signs of deterioration, then I will swap out the fan clutch from my 85 300TD.

Thanks for the diagnostic tip though. I'll make a comparison between the two cars.

Here's the thing I don't like about the implication that it is the fan clutch:

When I pull over while the car is running hot, the lower radiator hose is "cool" to the touch (c.f. the upper radiator hose). I don't have a hard number, but I would estimate that the hose is less than 55 C based on feel. This is the temp at which I pour bottles of molten agarose in the lab on a daily basis, I do know the difference between 55 and 75 C by touch.

If the fan clutch were faulty, then the radiator would not return "cold" coolant. Right? In addition, I would think that the car would run hotter at idle, when the fan is needed most.

A poor circulation issue on the other hand, would produce hotter coolant within the engine and colder coolant returning from the radiator, because the coolant spends more time in each respective part of the system.

As to the source of poor circulation... it could be any number of things: clogged radiator, thermostat not opening fully, weak water pump or clogged engine passages. The first two I think are OK, the last two... I don't know.

Cool!
 

I would suggest trying another thermostat first, it's easy and inexpensive compared to some other things you might do. My '85 ran consistently between 80C and 90C, a little higher in hot weather with a/c, a little lower in cool weather w/o a/c. When my mechanic replaced the bypass hose last month, he put in a new thermostat and now the temp is right at or just under 100C. Both the old and new thermostats were marked "80C."

The additional difference I notice is that the temp no longer changes with the environment or a/c. It's always in pretty much the same place on the gauge, just at or below 100C. Nothing else was changed, incidentally -- water pump, radiator, hoses, etc., are all in good condition but have been on the car for a couple of years.

The mechanic claims that the engine actually runs more efficiently at a higher temperature and I have read that (theoretically) a heat engine is more efficient if the temperature difference between the "hot sink" and the "cold sink" is as high as possible. Whether this is something that you would be able to see as increased performance (horsepower or fuel economy) in a "real" engine I do not know but I'll put 2000 miles on the car in the next week and will report back.

Jeremy

What about trying this for troubleshooting purposes:

While idling hot, hose off the front, aroung the condenser and radiator fins and see how the temp gauge behaves.

If temp. goes down, it's air circulation across radiator (fan clutch)

If no change, it's fluid circulation (could be water pump, thermostat, or clogged radiator tubes)

....just a thought

Agreed, pulling it is the only definite way to determine whether or not there is a problem with the impellers. I would go with Jeremy...check t'stat, again. I imagine you'll replace belts, if necessary, were you to decide to check water pump. Will still need to deal with radiator, keep us posted.