Engine coolant pump replacement?

[Vincent998]

Has anyone ever changed the "water" pump on a GT motor?

I seem to have a very small, slow leak. It was discovered when I got a bunch of spatter on the motor side of the internal window, and it has only happened once (last week). I cleaned the window off and then drove the car about 20 miles at a variety of speeds and it has not happened again. A thorough inspection from the bottom (belly pan off) and the top could not find any evidence of leakage past a hose clamp, but inspection using a borescope showed that the pump housing was wet below the drive shaft and there was a coolant drop clearly visible. The spatter was caused by a drop of coolant falling on the drive belt and being thrown upwards onto the engine side of the window.

There have been a number of such events described on the Forum - there was a thread by junior on 06-21-2008 at 4:14 pm describing this exact kind of event, but I can't find any discussion of what eventually happened, nor any discussion of water pump leakage.

So, what to do? Is there a risk of a catastrophic failure (e.g., it would lunch the motor) or is this a known not-too serious problem that can be fixed but is not an emergency?

I await your collective wisdom.

 

[fjpikul]

You're sure it's not the supercharger snout?

 

[Vincent998]

You're sure it's not the supercharger snout?

Yes. I declined to taste the drop from the pump body (I know that tasting it is old school, but it IS poison ...), but the spatters on the window dried and were white, and the drop we found on the bottom of the pump body was coolant, not oil. It was yellow/gold, not colorless like s/c oil. There were dry white streaks on the inner lips of the lower pulley. I have photos that I can post if anyone is interested. Plus, I wiped under the s/c snout and although it was oily, there wasn't any accumulation etc.

 

[2112]

Subscribing

 

[nota4re]

I'm sure Junior will chime in but I know the problem continues to persist. We have a couple of the OEM hoses and clamps here awaiting install during the next service but it sounds like we now have another suspect.

 

[ultrasportracing]

Do a cooling system pressure test, should show up any leaks.

 

[NorthwoodGT]

as another suggestion, make sure you have replaced all the factory OEM hose clamps. that design has proven not to be the best and does create leak issues after awhile. jmo

 

[junior]

Is this the kind of splatter you've noticed? (see pictures below). This was noticed on my car with less than 1000 miles on the clock. Team Kendall/Ryan (Nota4re) replaced few suspect hose clamps, issue went away at the time. However, it has re-surfaced lately. This is more noticeable after a cruise on a hot day (80+ Deg F). Ryan re-tightened the non OEM clamps again a couple of months ago, and now much less splatter experienced. So not sure if it's a hose clamp issue, a tiny pinhole in a hose somewhere, or a water pump issue (good observation on the water pump possibility), but I'm barely at 7k miles so doubtful in my case at least, unless I have a dud water pump from the factory. Coolant level is checked religiously, no loss noticeable in the coolant tank. Also, notice the Calcium like deposit on the pulley right under the short coolant hose below. Other GT owners with OEM hose clamps are not reporting this issue which reduces a clamp type issue. Hopefully we can get to the bottom of this during the next service.

 

[Vincent998]

Hi junior! Many thanks for your post.

Yes, the splatters on my window are similar to yours - but much less dense. I notice that yours, like mine, are spread in a horizontal band across the window, and most dense in the center where the s/c belt is located. I've attached a photo of mine so you can see. It has only happened once to me, at about 27,000 miles (as I said, last week). I've also attached a photo of the pulley from the bottom so you can see the streaks of what I believe is dried coolant on the inner lips of the pulley.

Yesterday evening I pressure tested the system, as suggested above by ultrasportracing. I used a Stant pressure tester, pumped up to the max (as per instructions) for a 16 psi cap. The instructions say that if there's no drop in 2 minutes, any leak will be very small. I left it on for 4 hours to see the pressure drop. I can definitely say that there was a little drop in the first 20 minutes, and after that it fell very slowly. I checked every 20 minutes for the first two hours, then every hour until four hours. I attach a photo of the pressure gauge face. I marked on the glass with a pen; there are five marks. The one farthest to the right (Mark #1) is the start - it's hard to see, but was the left edge of the white area, aligned with the yellow arrow. Mark #2 is 20 mins, #3 is 1:20 hr, #4 is 3 hrs, and #5 (farthest left, actually right on the needle end) is 4 hrs. I'm not sure if this decline really means a leak in the system or just a very slow leak in the tester and its cap. The first drop at 20 mins could also be the results of cooling of the air pumped into the tank; air heats up when compressed, so it could also be that warm air pumped in simply cooled to ambient temperature.

That's what I can say at this time. Tomorrow I may pull the middle pan so I can get a better look at the hoses and see if there are any drips from the pressure testing. Tracerline makes a fluorescent coolant dye that is Ford approved. It would be well worth using it to find any leaks.

 

[Vincent998]

This morning there was a 6" puddle of coolant on the garage floor when I raised the car on the lift, so I took the middle pan off and there was a fair amount more trapped on the upper side of it. Using the borescope looked at the coolant pump again and although I could see dried coolant where the drop was seen yesterday, today there was no wetness. To make the story short, there seems to be no more leakage from the pump, but both coolant hoses on the coil cooler below the oil filter housing were leaking past the clamps when under pressure; when the clamps were tightened the leakage stopped. I'm leaving the pressure on all day to see if the two hoses stay dry and to see if any other leaks develop. I also periodically crank the motor a little bit so the coolant pump stops in a different position under pressure to see if it will leak.

 

[junior]

Well done on the trouble shooting. Your light splatter pattern is what I see now on mine under hot weather driving and after re-tightening of the non OEM clamps, I am still leaning toward a bad hose/clamp connection cause. Anxious to learn about the 2 suspect hoses and possible other leak areas, please update when you know. Thanks!

 

[junior]

You're sure it's not the supercharger snout?

Very doubtful, on mine at least, the specs residue is non oily and wipe off easily with no streaks. BTW I checked the SC oil level at the time and it was and still is A OK.

 

[Vincent998]

Just returned from a 400 mile drive over three days. Total mileage is now more than 27,000. No splatters or any sign of coolant leaks. At this point it seems as though it was a very small leak - maybe only a single drop or two - and perhaps the heat etc of the long drives softened things a little bit so it no longer leaks. Before I left on the trip, I could not make it leak at the water pump seal even though I pressurized the system and left it for 20 hours. Both coolant hoses to and from the oil cooler did leak at the cooler, however, and tightening the clamps a bit stopped it.

I have a professional and very experienced mechanic friend who is looking for self tensioning clamps of the same diameter but significantly greater width, which can be used to replace the stock narrow clamps. If they're found, I'll post the specs and purchase info.

 

[Vincent998]

Here is a possible source for better clamps than those used by Ford: http://www.tacomascrew.com/Products/Hose-Clamps#!Hose-Clamps

If you go to the web site and look at hose clamps, you will find both the constant torque clamps like those on our GTs and "lined" clamps. The lined clamps are worm gear type, but with an internal shield that keeps the hose from being squeezed through the slots in the clamp when it's tightened. The constant torque clamps are 5/8" wide, so much wider than the stock clamps on our cars. The lined clamps are also wider than the constant torque clamps we have, and therefore both would provide more sealing area when they are tightened.

 

[Sinovac]

The constant torque hose clamps with the Belleville washers work well.

 

[PeteK]

That's an excellent design. I've never seen them before. Thanks for the info.

 

[Sinovac]

That's an excellent design. I've never seen them before. Thanks for the info.

I won't use anything else. I've never had one fail.

 

[Vincent998]

After driving my car more than 600 miles since the spatter, I got some more. Is there a weep hole in the water pump housing? If there is, where is it?

 

[junior]

Had my coolant change service done by Cool Tech LLC's meister Ryan over last weekend, suspect clamp was replaced with a factory clamp, system vacuum tested successfully and coolant changed (engine and supercharger), and there was no evidence of leaks from the water pump. I am 99.99% confident (so is Ryan) the culprit clamp was this one shown in the pictures below. Hot (very hot) tomorrow here in So Cal, near perfect test conditions, will test drive and hopefully no coolant splatter on the fire wall.


Hose clamp installed


Hose clamp not installed for better illustration of its location on the hose:


Vacuum test, gauges indicating no pressure losses after clamps installation:

 

[KJRGT]

Great to hear! Let us know of the results of your testing.