: Water Pump - Failing When Hot



Paul2020
04-10-14, 11:05 AM
Is it possible that a Northstar water pump can work when cold and fail when hot? It seems to be happening on my 1999 Seville. :-)

I pulled and extended the return line at the coolant reservoir and it flowed out into a pail when it was cold. After it gets warm however there is no flow.

How likely would an air lock be if I am continually topping it off (with the engine running) with distilled water? My thought is that there should still be flow since there is a bypass around the thermostat where an airlock might occur.

There is no steam coming out of the exhaust. No sweet smell to indicate a blown head gasket. No water in the oil nor in the oil cap. When I tried a block test it did not turn color before it started boiling over. I don't think it is a head gasket. It churns violently when I put it under load and comes out the radiator cap.

I think it must be the water pump. I wasn't aware that it could flow when cold and shut off when hot though. When it was cold the flow rate did seem a little slow though.

Thanks!

rodnok01
04-10-14, 11:21 AM
The pump turns the same hot or cold, now possibly the tensioner or belt are weak and in turn slowing the pump down. Can't think of that happening here to anyone though and doubtful.

If you are using only water it's going to boil it off...

Paul2020
04-10-14, 12:06 PM
Appreciate that input RodNok. I checked just now and as far as I could tell the pump pulley is spinning good. (Can't see it real well but seemed like it was and seemed to speed up when I gave it throttle).

I don't know what else it could possibly be with the flow completely stopping if it was not the water pump, do you? I've tried to keep the water level up well above the level of the pump.

rodnok01
04-10-14, 12:45 PM
You could have something collapsing or plugged I suppose preventing flow. Hoses are know to collapse under pressure. Since the thermostat is closed at lower temps may be why you are seeing flow in purge line. There have been a few plugged radiators around here that drove owners insane trying to figure out over heating problems. Not saying the rad is bad but a flow problem there may be somewhere.
Are you running 50/50 mix or just water?

Submariner409
04-10-14, 12:45 PM
Thoughts - Your statement says "water". If you are currently running only water and messing with the engine with the coolant reservoir cap off, water boils at 212 degrees, about 12 degrees before fans go to SLOW (at 224, HVAC Off)........... Sooooo, if that's the scenario, then yes, the engine is overheating and boiling over. Even a faulty pressure cap will cause a "low temperature boilover". A proper pressure cap that holds system pressure raises the boiling point of water up to about 245 degrees and for 50/50 coolant, to 265 degrees. IF the water in the engine boils, then the water pump CAN NOT move water - it simply cavitates in the watery foam and all cooling ceases.

You need to have a proper coolant mix installed as well as a good pressure cap in order to watch accurate temperature changes. A 50/50 coolant mix boils at about 218 degrees at atmospheric pressure, so the engine block exhaust gas test should be done with the engine warm, idling in P, and don't try to outguess the test set - bubble the test fluid according to the directions, then stop and shut down or the engine WILL overheat.

Look carefully at Ranger's coolant boil tables - coolant percentages and pressures -- learn how to apply the info to your particular circumstances.

Ranger
04-10-14, 02:39 PM
I don't know what else it could possibly be with the flow completely stopping if it was not the water pump, do you?
Flow stops because the WP pumps liquid, not gas. Your HG is breached, allowing exhaust gases to be pumped into the cooling system. When those gases aka air pocket get to the WP, the coolant flow stops.

Add to that, you said that "It churns violently when I put it under load and comes out the radiator cap" AND it's a '99.

Everything points towards the HG.

Paul2020
04-10-14, 04:08 PM
Flow stops because the WP pumps liquid, not gas. Your HG is breached, allowing exhaust gases to be pumped into the cooling system. When those gases aka air pocket get to the WP, the coolant flow stops.

Add to that, you said that "It churns violently when I put it under load and comes out the radiator cap" AND it's a '99.

Everything points towards the HG.

Add to that, you said that "It churns violently when I put it under load and comes out the radiator cap" AND it's a '99.

Everything points towards the HG.[/QUOTE]

Hmmm... That's what I thought at first too. The thing that made me think otherwise was the fact that there is no steam coming out the tail-pipe. No sweet smell and no other indications I've been able to find. (No water in the crankcase either for that matter).

Have you ever seen a case where the HG was blown, putting hydrocarbons into the coolant but the coolant was not going into the the combustion chamber? Seems like it would always be a two way street. If the exhaust was going to coolant, the coolant should also be coming into the combustion chamber. Have you ever seen otherwise?

I see no bubbles in the coolant reservoir. When I attempted the block test for exhaust gases it did not turn yellow. It is acting just about like a DeVille I worked on lately that had a broken water pump belt. Just boils up real fast when you rev up the rpm's.

----------

Well, I pulled the housing off the water pump and took a look. The thermostat looked like it was in bad shape and as I studied it I could not see any route as I thought for the coolant bypass the thermostat and go to the heater then back up to the coolant reservoir where I am doing the flow checking. So if it is a bad thermostat stuck closed then I would not see any flow from that return line as far as I could tell.

I put in a new thermostat and put it back together. Still doing the same thing though. No flow after it warms up.

When I looked at the pump impellors they were pretty rusty but I thought they would probably still pump. Maybe that rust would keep them from pumping? I would have changed it but could not easily get it out without that special tool.

I am aware of the different boiling temperatures between coolant and water. I started out with coolant boiling over and switched to water since it is a little cheaper :-) It should not really be getting much over 190 degrees with the little bit of load I'm putting on it and the outdoor temp today in the 60's or 70's. The gauge does not show it really overheating. Dead center. I put an IR temp gun on it and it showed 170's to 190's around the upper and lower hose casings.

I guess I'll go ahead and change the pump and see if it makes any difference.

Submariner409
04-10-14, 04:53 PM
Your cooling system is way different in flow pattern from other systems. The thermostat is in the return line from the radiator. The thermostat housing itself is the bypass circuit. Other than the small purge liquid dribble there is NO "flow" through the surge tank. It is an expansion/contraction tank that sits on the cooling system circuit - connected by a 3/4" hose at the bottom - which T's into the heater circuit. The purge line merely bleeds the system high point which allows trapped air/gases to collect in the airspace OVER the coolant in the tank - the "clean" coolant then finds its way back to the main system.

Look at the orientation of the thermostat wax pellet cylinder - it's in the water pump cover, NOT pointing to the radiator.

Coolant flows to the pump, through the block and heads AND the heater core circuit, then back to the water pump. When THAT circuit stabilizes and coolant temp reaches 188 the thermostat (acted upon by block/heater circuit temp) begins to open and coolant slowly begins to flow to/from the radiator. In these engines there is almost NEVER full coolant flow through the radiator - the thermostat is fully open at 206 degrees; most Northstars run at 195 - 200, so the thermostat is only 50% - 80% open. In a good, cold winter it takes a LONG time to get any radiator flow - your heater demands will cool the whole enchilada.

Very seldom does a Northstar with failing head gaskets put coolant into the lube oil.

More cooling pictures in my albums - click my username, open the profile. left column, Albums. 6 pages of stuff for your car.

Ranger
04-10-14, 08:40 PM
Hmmm... That's what I thought at first too. The thing that made me think otherwise was the fact that there is no steam coming out the tail-pipe. No sweet smell and no other indications I've been able to find. (No water in the crankcase either for that matter).
You won't see steam in the early stages. When it gets to that point there will be little question as to what the problem is.


Have you ever seen a case where the HG was blown, putting hydrocarbons into the coolant but the coolant was not going into the the combustion chamber? Seems like it would always be a two way street. If the exhaust was going to coolant, the coolant should also be coming into the combustion chamber. Have you ever seen otherwise?
Logically one would think that has to be true, BUT consider an early stage breach so small that the only way exhaust gases are getting into the cooling system is by being pumped in under tremendous pressures created in the combustion chamber, but the breach is too small for coolant to flow the other way when shut down (until it gets worse).


I see no bubbles in the coolant reservoir.
If the purge line is working as designed there should be none.


When I attempted the block test for exhaust gases it did not turn yellow.
No, but there was a very slight color change and that MIGHT be indicative of a very small amount of exhaust gases in the tank.


It is acting just about like a DeVille I worked on lately that had a broken water pump belt. Just boils up real fast when you rev up the rpm's.
Didn't you verify that the pump is spinning?


Well, I pulled the housing off the water pump and took a look. The thermostat looked like it was in bad shape and as I studied it I could not see any route as I thought for the coolant bypass the thermostat and go to the heater then back up to the coolant reservoir where I am doing the flow checking. So if it is a bad thermostat stuck closed then I would not see any flow from that return line as far as I could tell.

I put in a new thermostat and put it back together. Still doing the same thing though. No flow after it warms up.
That rules out the thermostat.


When I looked at the pump impellors they were pretty rusty but I thought they would probably still pump. Maybe that rust would keep them from pumping?
Wishful thinking. Not a chance.


I guess I'll go ahead and change the pump and see if it makes any difference.
You are wasting your time and money. If the pump is spinning, it's doing it's job. It's just that simple.

Caddy-EaRL
04-10-14, 10:53 PM
I just replaced a water pump that was spinning but when removed saw that the impeller was loose on the shaft.

Submariner409
04-11-14, 08:36 AM
Now that you mention kit, a while back we had one or two in here that had shed the cone-shaped "shroud" over the blades. No pumping there, either.

Ranger
04-11-14, 10:11 AM
Anything is possible, but that is about as rare as hens teeth and wouldn't explain loosing coolant flow only when hot.

Paul2020
04-29-14, 05:59 PM
Sorry for the delay in getting back here.

I got a different block tester kit and found out that Ranger and Submarine were right. There are hydrocarbons in the coolant.

I did this after temporarily bypassing the heater core (re-routing the heater output line to the return line). It had a serious leak that flowed to the back seat. (I had about an inch of water/coolant standing in the back seat floorboard.)

The information you guys furnished me has been extremely valuable. Submarine, you got me thinking about the flow of the coolant through the pump and the heater/thermostat circuit. Based on the diagrams I still have a question in my mind about the exact flow of this circuit. Do you know of any books or manuals or other threads that would give detailed info on that.

When I get a chance I'd like to share a picture with a specific question I had about how it is possible to flow through the heater bypass circuit when the thermostat closes the passage except for those small holes in the thermostat plate and when the flow looks like it is moving away from that and towards the pump impellers and through the engine.

Thanks!

Submariner409
04-29-14, 06:58 PM
You have the flow bassackwards. Please go back and read Post 8 again. The thermostat is in the radiator return line - it NEVER passes coolant until the engine/heater circuit is up to at least 188 degrees - one reason a Northstar "warms up" so quickly.

The water pump ALWAYS moves coolant - let's take an ounce of "water" and move it through the system, beginning at the water pump impeller. The water leaves the pump impeller and moves through the block, up through the heads and back to the water crossover. SOME of it (now relatively "hot") gets diverted out the rear of the crossover into the heater circuit, where it flows through the heater core and back to the line that T's into the bottom of the surge tank. That line is the one that goes into the water pump cover suction side and mixes the heater circuit water with the block/head water, and so back to the water pump. (See my earlier diagram) As the block/head/heater circuit temperature approaches 188 degrees the thermostat begins to open, allowing a dribble of cooler radiator liquid to mix with the hot block circuit liquid. As the entire system comes up to temp the 'stat opens further and there is a dual flow of coolant through both the heater and radiator circuits: The thermostat will equalize the flows to maintain ABOUT 190 - 205 degrees in the entire system.

Because the thermostat is not fully open until 206 degrees and the car usually runs at about 190 - 205 degrees, the safety factor allows a LOT of radiator flow when necessary - summer traffic and A/C operation. Actually, a Northstar thermostat runs about 75% open most of the time. In COLD weather the heater circuit will cool the engine with NO radiator flow.

Find Ranger's Northstar system flow diagram and trace it out. Most elegant.

Ranger
04-29-14, 09:27 PM
Submarine, you got me thinking about the flow of the coolant through the pump and the heater/thermostat circuit. Based on the diagrams I still have a question in my mind about the exact flow of this circuit. Do you know of any books or manuals or other threads that would give detailed info on that.

This help any?

Paul2020
04-30-14, 09:12 AM
...back to the water crossover. SOME of it (now relatively "hot") gets diverted out the rear of the crossover into the heater circuit, where it flows through the heater core and back ....

Thanks Sub! As I trace it through now, that makes sense. You guys really have this down!

Submariner409
04-30-14, 09:55 AM
Just remember: That surge tank is under pressure and is an expansion/contraction "spring" only - other than the purge line discharge there is essentially NO "flow" through the surge tank - it simply rides on the system, connected by the T in the line from the heater circuit. As the coolant warms and expands, the cap-sealed airspace in the surge tank is what provides the "spring" to maintain system pressure at 16 or 18 psi (depending on year/model).

Here are the boil points for different coolant mixtures at different pressures. Cadillac allows a maximum coolant concentration of 70/30 coolant to distilled water, with 50/50 being the most often recommended. (Chart courtesy of Ranger)

Remember - EVERY automotive engine on earth now runs some form of "long life" DEX-COOL or a variant, so don't get antsy about coolants - BUT change it every 3 years.