Weird Cooling problem

[JayArr]

So we have been on this road trip for 10 days and there were no problems with engine temp until yesterday. Driving along and the temp started to creep up and it hits 125C/225F as I pull over. It's been running around 100C/180F for the last ten days.

When I open the hood I can hear the resevoir tank gurgling. There isn't any more coolant going in, just steam. It's full!

So I let it cool down and the tank remains full. Finally at about 95C I remove the rad cap and when I do there is one last burp to the resevoir tank and then the coolant from up there begins flowing back to the radiator.

If I just wait, with the cap off, the reservoir will slowly drain and fill the rad back up. I then put the cap back on and all is well again.

When it happened the first time yesterday I stopped at Napa and bought a new rad cap but it happened again today.

As I understand it the pressure in the rad has to reach a certain point before the rad cap valve opens and coolant can go up to the resevoir, can anyone tell me what that pressure is? Similarly, when the engine cools down, at some point that same same valve should open and let the resevoir tank drain back to the rad to fill it back up for the next trip.

Any ideas?

The only guess I have right now is that when I'm going up a very steep hill and the transmission kicks all the way down into first and revs at 4k-5k I may be pushing coolant out of the system into the resevoir tank and I can't get that back without a cool down period.

I'm wondering what would happen if I removed the little valve in the rad cap.

Any ideas?

 

[Mooseman]

Relief pressure is supposed to be 16psi IIRC. The way you explained how the system.works is correct.

Maybe your fan clutch isn't pulling as much as it should? I know you went with a new ACDelco or Behr but maybe it crapped out again. At that temp, it should be pulling pretty hard. this is the command chart for the EV clutch.

Edit: looking at the chart, at 225f, it's.commanded only to 29%. The way you tow, it should be reprogrammed to pull harder at those higher temps, more like at least 50%.

 

[TJBaker57]

If you even suspect your fan clutcho you may be able to monitor both the actual fan speed and the PCM desired fan speed as well in your Torque Pro app. The actual fan speed was the very first PID I posted in this thread...

More PIDs for Torque App

A short time ago I posted a new pid I discovered for the Fan Speed of a 4.2 in a 2002 Trailblazer. It was properly suggested by MRRSM that I repost that information here so here I am. It is possible that this is my first original post! I have been spending a great deal of time fiddling with...

gmtnation.com

Desired fan speed came later.

 

[JayArr]

I think this may be even simpler than expected.

Tonight after I had left it for a couple of hours, I poped the rad cap off and I couldn't see any coolant but there was still lots in the resevoir. I expected it to drain in but this time it didn't.

I removed the line at the rad and blew into it and after some initial resistance I was able to blow through and create bubbles in the resevoir.

I put the line back on the rad and it filled up.

So now I wonder if I have some bottom dwelling sludge in the resevoir that gets blown up into the resevoir by the pressure when the rad is expelling coolant but then gets sucked into the resevoir exit and/or the hose line plugging it and keeping the coolant from flowing back.

Tomorrow, with just the tools on hand in a provincial camping site I'll attempt to remove the resevoir and clean it out.

Wish me luck

PS - Thanks TJ I'll add those two fan PIDs to my torque Pro.

 

[Redbeard]

.

Tonight after I had left it for a couple of hours, I poped the rad cap off and I couldn't see any coolant but there was still lots in the resevoir. I expected it to drain in but this time it didn't.

Make sure all the hoses are tight and that there are no pin holes in the hoses. As the engine cools it is removing the excess antifreeze from the resevoir and "sucking it" back to the engine. If there is any holes or loose connections it will suck air and not coolant. Hence a full resevoir but much less coolant in the radiator.

 

[JayArr]

Well, I notice that there is no clamp on the line where it connects to the resevoir, I suppose it could be sucking air there. I don't have a spare hose clamp but I can put a couple of tie wraps around the joint to try to tighten it up until I can get to NAPA.

I pulled the reservoir and cleaned it out, I didn't find any conspicuous sludge that could block it. I ran a 1/4" rod all the way through the line to make sure there was no blockage there.

I also found this website:

Radiator Caps Explained

It seems to indicate that for an open reservoir system like our Envoys the return poppet should have no spring and be open all the time. This would allow coolant to flow back to the rad while driving even with pressure below the 15psi that the upper valve opens at.

A rad cap with a spring on the return poppet would need an actual vacuum to suck back coolant not just a pressure below 15psi.

This makes total sense to me so correct me if I'm wrong. Don't we want the coolant to flow back as it cools? If I heat up going up a mountain hill and I push coolant into the reservois don't I want it to flow back when I'm going down the other side?

So I looked at the NAPA rad cap and sure enough the return poppet is spring loaded.

Does annyone have an OEM rad cap? Is the return poppet spring loaded or normally open?

Could this be a bad cross reference at NAPA?

 

[Mooseman]

I think it has to have vacuum regardless. The poppet, or return valve, has a very light spring but pressure from the radiator side will keep it closed until it cools sufficiently to create a vacuum. I don't know how it would affect the system if it were to be boiling and creating steam.

This cooling system is marginal at best, hence why they never went to electric fans and stayed with an oversized mechanical fan to compensate. If the fan is not.putting out as much as it should under heavy loads and temps, it might create the problem you're experiencing.

 

[JayArr]

Well, I programmed in TJs PID codes for desired fan speed and fan speed so I'll be able to monitor those as I drive on Monday. I think I'll stop every hour or every time the temp goes above 100 and we'll just have to keep our fingers crossed.

We're going through the Rockies at Jasper but from Alberta to BC is the easy way because we're already at a high altitude.

 

[JayArr]

Help me out here, I'm having a hard time understanding the idea of the vacuum sucking the coolant back.

Let's say we have water in the rad (because I can't find tables for glycol)

at 90C the pressure is 10.19psi and at 100c it's 14.71.

So if the rad water cools from 100c down to 90C does that represent a "vacuum" of 4.5psi? that should suck the coolant back in through the return poppet?

The coolant all ends up in the reservoir so my instinct says that either there isn't enough vacuum or there is a blockage.

Could the 18 year old rubber hose from the rad to the resevoir be collapsing under vacuum?

 

[budwich]

I would be surprised if the hose is collapsing but you should probably see that during the "cooling" event (engine off, cool down).

As with most "vacuum driven" systems, air plays a big role. The system is relying on the expansion volume of water and then the resulting contraction there after to act as a "pump". However, IF there is significant air ingression, then it is unlikely that there will be a large vacuum action thereafter. How does the air get in? The overflow return line is one source (readily checked). The other "source" is the "other end of things"... the rad and the engine. The "boil event" might be such that the system has pressured past the "normal expansion" of "heated fluid" causing the overflow event. BUT, since the event may have been caused with some "air injection" then the system is now "compromised" and thus the resulting "vacuum event" may not be as great as expected because the loss of contraction of liquid volume is not as great because of the ingression of air. Of course, an "air event" is somewhat normal when doing repairs like rad replacement and heater core replacement, etc. The "burp" process is usually followed. However, other "air ingressions" aren't usually so nice. Maybe a simple "boil over" is NOT so simple but could be.

I am going to watch mine as I tow heavy and have seen the temps up near 220 at times. I don't believe that I have ever seen a tank fill but I guess could happen during an engine stop event shortly thereafter.

Hopefully, yours is nothing more than a "larger than expected boil over" (poor cap maybe), that can be resolved by appropriate "fill /purge process".

 

[JayArr]

Well two 25 cent hose clamps on the line from the rad to the tank seem to have helped a great deal, I'm still running a little hot up the hills but it cools down and sucks the coolant back from the tank on the down slopes.

I think Mooseman hit the nail on the head when he said the cooling system is marginal at best. I can drive for an hour with no problem but then I reach a big hill and it seems like the cooling system can't get rid of the extra heat generated by a big hill. The result is I'm running along at 92C and I go up a big hill and the temp rises to 94 or 95C but then it stays there if there is no big downhill to cool it off. Sometimes I drive for miles at 95 or 97C and then I hit a downhill and it cools back down to 92C again. I must be at the limit on flat ground with no extra capacity for hills.

Yesterday I was running at 94C climbed a very large incline for about 5 miles and then turned into the capsite. I was at about 102C when I turned off the highway and by the time I drove a mile into the campground the heat soak from the engine had driven the temp to 132C (270F!) with the dashboard chiming like crazy and the resevoir churning and gurgling.

I'm one day from home and my Stant kit that tests the cooling system for leaks. I've also not been able to get an OEM radcap anywhere in my travels but the GM dealership closest to my house has one in stock - go figure! To top it all off I've burnt out the doughnut gasket between the exhaust manifold and the downpipe to the cat so I've got this deep throaty sound as I run down the road. Oh and yesterday I blew out a trailer tire running down the highway and had to change it on the side of the road.

Good think I like adventures!

 

[Mooseman]

Maybe try adding some Water Wetter to the system. I know it helps my sleds with heat transfer.

 

[TJBaker57]

was at about 102C when I turned off the highway and by the time I drove a mile into the campground the heat soak from the engine had driven the temp to 132C (270F

Have you tried to read your fan speed and desired fan speed with Torque Pro?

 

[JayArr]

Yes, it's running around 1300 when "desired" is only like 300-400.

I don't think it's a matter of the fan not cooling.

I let it cool down and it sucked the resevoir tank back into the engine except for a cup or two. I just took the cap off and the last cup or so drained into the rad and the resevoir was back to "full when cool" level.

There is no actual coolant loss.

Curiously... since I added your two PIDs for the fan I've lost my transmission fluid temperature PID, it went to "-" and then when I tried to delete it and reload it disappeared from the list entirely. I don't know if that's a coincidence. Not the main concern right now, the trans is performing well.

Would an exhaust leak where the manifold connects to the downpipe to the CAT cause the engine to overheat or run hot?

 

[TJBaker57]

Would an exhaust leak where the manifold connects to the downpipe to the CAT cause the engine to overheat or run hot?

That CAN cause a bad O2 sensor read for certain. And that could in turn cause a lean burn condition, which could cause a hotter engine in general.
OR, I may have backwards and it may cause a rich condition?? Cannot remember, maybe someone else knows. But a leak there can definitely cause a bad O2 sensor value.

I have had issues finding a PID in the list in particular when the displayed name (short name) is substantially different than the long name used in menus.

I have not experienced a PID completely disappearing from the list.

The transmission fluid temperature PID is one of the [GM] PIDs, listed as "[GM]Transmission Fluid Temp (GM Method 1 or 2 or 3) Method 2 is for CANBUS vehicles with separate ECM and TCM and methods 1 or 3 should work for the rest.

 

[TJBaker57]

Yes, it's running around 1300 when "desired" is only like 300-400.

OK that sounds about what I see at highway speeds. That is about what I see it run when the PCM is NOT commanding the fan clutch on at all.

At idle speed of 600 or so I see about 720 fan rpm when there is no engagement requested from the PCM.

What this does NOT tell us is does the fan speed increase at low speeds when the desired speed is higher than the actual speed.

In other words we cannot tell if the fan clutch actually engages when commanded to engage by the PCM due to engine heat or high AC high side pressures. Like when you come off the highway and travel at low speed for a mile or so and the engine coolant gets hot.

 

[JayArr]

There is a way to test this with the Tech II, I did it before when I first iinstalled the fan. I think the results may still be posted in a thread here somewhere. It's a little late to be running the engine at high RPM in a campground but I'll do this in the morning to confirm that the command is implemented.

I'm a short drive from Kamloops so I may unhook the trailer and drive into town and see if I can get a muffler shop to replace the doughnut gasket on short notice. Maybe that will help.

The O2 sensor is varying, I've been watching it as I drive but I don't know if it's in the right range and perhaps the leak is making it think it's just right when it's not.

 

[JayArr]

Just hooked up the Tech II and did the "viscous clutch test".

I can command up to 100% and when I did that and revved to 3000RPM the fan speed matched the desired speed (about 3500RPM).

This pretty mutch matches the numbers I got when it was new.

My conclusion is that the fan and fan clutch are OK.

 

[JayArr]

I tried the PIDs for commanded and measured AFR but they don't work, I guess this PCM doesn't send that data.

 

[Mooseman]

So @aaserv just posted this :

So I go see said guy and he says no problem. I mention I would like to get the fan speeds increased and the transmission shift points raised as well. No problem. Well as shocking as this may sound so far its working perfectly!! Yea when driving in town with 106Deg temps I can hear the fan spinning like a helicopter but dammit Ive yet to see the gauge go over 209.

I think this would be the way for you to go as well. Increasing the fan speed according to the temperature would be good. There's another way you could achieve this by putting 12v directly to the white wire on the fan harness however @TJBaker57 has mentioned that he burned one out doing this.

Maybe getting the shift points raised might help too.

 

[TJBaker57]

I tried the PIDs for commanded and measured AFR but they don't work, I guess this PCM doesn't send that data.

Tried these where? In Torque Pro? I only see Commanded AFR at a mode&PID in Torque of 22119E. and my own 2005 P10 PCM supports that. I have seen some PIDs requiring the addition of "01" after the mode&PID that comes with Torque. So where 22119E comes back with no data I would try 22119E01.

 

[TJBaker57]

@TJBaker57 has mentioned that he burned one out doing this

I did burn out more than one doing this. But the service literature states that is a valid test! Go figure.

 

[JayArr]

Drove the car into Kamloops yesterday without the trailer. It still got warm on the hills but never more than 97C and it cooled quickly.

I stopped at a garage and the mechanic came out with his stethescope and probed around the manifold as I reved the engine.

The doughnut gasket at the juction between the manifold and the downpipe to the cat is fine but there is a leak somewhere under the heat sheild over the manifold. I figure I either have a cracked manifold or the mechanic suggested it might just be broken studs to the head and a burned gasket. Either way - I think this is the cause of the overheating.

It's not something to be fixed here, parts would take days to get here and then it's a day in the shop. Even tearing it down further does nothing but make the vehicle inoperative and stranding me here. I don't want to spend $1k on a manifold replacement 3+ hours from home, I think I may have a spare manifold in the garage from the second engine. It's driveable as-is as long as I don't pull the trailer.

Hwy 1 is closed through the Fraser Canyon due to a wildfire and there is no way I'm going to try to tow the trailer home on the Coquihalla with the Envoy,

My wife came up with a great solution... leave her and the dogs here at the campground. I need to be back to work on Monday but she doesn't start back until after Labour Day. I'll drive home on the Coquihalla without the trailer and either fix the car or rent a truck and come back next weekend and get her. She's perfectly happy to sit here and read and drink wine and relax for the extra week.

That's the plan, I'll try to get home tomorrow and get the manifold off and order parts Monday and put it back together by next weekend. I've got this faint hope that with the exhaust fixed back up it will tow without overheating like it did for the first week of the trip.

JayArr

 

[JayArr]

I don't think raising the fan speeds will help with the heat soak, the problem there is that the engine RPMs go down once you leave the highway and even if the PCM commands 100% the RPM of the fan will only be slightly higher than the idle of the engine. So where it needs a fan speed of 3500RPM it's getting 750. The coolant just boils at this point.

Theoretically I should put it in neutral at this point and rev the engine to 3500 so the fan can cool the engine but I'm always too afraid to do that for fear of blowing the head gasket.

I do have this theory that when the temp spikes up around 270F what the sensor is really measuring is steam because the colant is all up in the reservoir tank and the engine cavities are no longer full of liquid. I can't confirm this but it's a suspician since the sensor is high on the engine.

 

[TJBaker57]

stopped at Napa and bought a new rad cap

A couple years ago I had an issue where I would drive somewhere that took 25 or 30 minutes, then after shutting off the engine and about to walk away from the truck I would hear the coolant boiling into the overflow reservoir. The engine wasn't so hot that it should be hitting the roughly 267°F required to boil a 50-50 water/antifreeze coolant mix. Eventually tested the radiator cap and found it was not holding pressure. It leaked at the seal that is supposed to open under high pressure. Without that 15 or so PSI pressure the mix boils at like 226°F and yes, I could see that happening when a hot engine shuts down.

I bought a new cap at Napa. Still had the same issue. Inspected and tested the new cap. It leaked at the large O-ring seal that seals the cap to the radiator neck. I happened to be going to the scrapyard so I brought along my cap tester and found one there. Problem solved.

If you lose pressure with a hot engine you are going to boil. Got a cap tester?

 

[JayArr]

I've got a full Stant diagnostic kit at home to pressure test not only the cap but the coolant system of the car so I'll do that later today when I get home. I

I've now purchased multiple rad caps:

Napa1: curently on the car - made in India
Napa2: in case Napa1 failed - made in India
Gates: Identical to Napa and made in India.
-- I believe the Napa units are made by Gates

None of them look very good and I wonder at the quality.

The one that was on the car when we started out was a knock off and it broke apart when I tried to remove it the first day we overheated. The top cap part twisted off leaving the body with the o-rings in the rad neck, the springs all popped out and I had to remove the body with a pair of pliers. It was soft and deformed as I twisted it out of the neck.

Interestingly the description of the OEM cap is " metal covered by plastic" which may be an indicator of a better quality, there is one in stock at the dealership near my home and I'll get it on Monday. The knock offs may be fine for everyday use and not up to the task when towing.

 

[TJBaker57]

My specific issue/complaint with my Napa cap was the large O-ring had a raised seam on it. How good can something seal with such a deformity on the sealing surface?

 

[Ilikemy3s]

Since I have not read (may have missed it) any mention of this .. I figured to ask, if just to eliminate the possibility .. but how old if the coolant in the system ?

 

[JayArr]

Hi All

First of all the coolant is a couple of years old but that's a good question so I just tested it and it's boiling point is north of 268F. Basically top of scale.

I finally got my hands on a genuine GM rad cap today and what a difference! Holy Crap!

I'm testing them with the upper o-ring removed so that I can see the pressure that the cap opens at. With the Stant testing adapter if you leave both rings on there is no escape route when the valve pops open.

GM Cap (AC Delco) Part#15042975

Pressure valve opens at 17.5PSI - this is where it will send coolant to the resevoir tank.

After 1 minute it still holds 17PSI
After 2 minutes it still holds 17PSI
After 5 minutes it still holds 17PSI

So the bottom o-ring seals properly and no coolant will flow past it to go to the tank, the only time coolant should flow to the tank is when the 17.5PSI valve opens.

NAPA Cap ( I think this is made by Gates)

Pressure valve opens at 17PSI - this is where it will send coolant to the resevoir tank. This is correct BUT!!!

After 30 seconds it only holds 11PSI
After 1 minute it only holds 9 PSI
After 2 minutes it only holds 7 PSI

This is terrible and likely the cause of my problem. It means that at a pressure somewhere around 7-10PSI coolant can leak past the lower o-ring and go up into the resevoir without having to push open the pressure valve built into the cap.

It means I'm not ever getting 15 pounds of pressure and my coolant is boiling at a much lower temperature.

It's why I pull over and my rad is empty and my reservoir is full.

 

[JayArr]

I also had issues with the exhaust and thought that it may be causing some or all of the overheat problem but I'm now fairly convinced it was the bad aftermarket rad cap.

Still... I pulled the exhaust manifold and it was leaking past the gasket on cylinders 5 and 6. The bolts were all easy to remove having only been in a few years and installed with copious never-seize. These are also those special length bolts that ensure maximum thread without bottoming out.

The doughnut was also burned up and leaking - AGAIN! This will be the fourth doughnut in three years so I'm trying to figure out why. I have ordered an OEM doughnut to see if the quality is higher than the cheaper Walker doughnuts I've been using. I also noticed one of the studs on the bottom of the manifold where it attaches to the pipe is misthreaded and crooked. Maybe I haven't been getting a good full, tight connection and it's been allowed to move and wear.

I also noticed that I have a Doorman aftermarket manifold for an 08 engine on an 03 cylinder head. I used the correct 08 gasket but the head has the small oval ports dumping into the larger ports of the later model manifold. I have a hazy recollection that I thought the later model manifold would breather better or something and I don't think there was a new manifold available for the earlier model at the time.

I found a genuine early model GM manifold on a shelf, I don't know if it's the original or where I got it but that's going back on with an 03 gasket. The quality is miles better, particularly at the bolt lands. the Doorman is all rough and not even parallel but the GM one is nice and flat and clean. I think the bolts will have more purchase and stay tighter longer.

I'll post up the results in a day or so but I think I've got everything figured out.

Good thing too, I left my wife and one dog in our trailer at a campground near Cache Creek and I've got to go get them and bring them home on the weekend.

 

[JayArr]

Went and fetched my wife this weekend and the new cap made a world of difference.

No overheats, the temp went up and down as anticipated, the reservoir never filled up and each time I let it cool down and checked the rad was full of coolant.

The lesson to be learned is don't believe that a new part is a good part.

If you're going to buy your rad cap from Napa or Gates or an aftermarket source then test it and make sure it will hold 15 pounds for 10-20 minutes. If not then return it and get the OEM.