It does sound like waterpump failure to me im guessing you have checked that the drive belt is still fitted ok tho?

Other than a visual inspection of the belt and pulling it a bit to check if it's tensioned, not really. It has about 25.000 miles on it. How often should it be replaced? It doesn't slip as far as I can tell.

The heater won't blow hot air if the thermostat is stuck shut even if the coolant is boiling because it's not circulating.

 

Thermostat - heater matrix circuit isn't moved by the water pump ? It needs to open the thermostat first of all ?

^ I was thinking about this last night.

I couldn't figure out why the heater matrix would go cold at first, but then I had an idea...

 

If the coolant starts to boil, due to the stuck-closed thermostat, there will be gas bubbles pulled round with the flowing coolant to the water pump.  The water pump is designed to pump water, not gas, so once there is enough gas surrounding the impeller, it probably will completely fail to push any more coolant round the circuit (like the opposite of priming a pump).  This will mean that the coolant does stop circulating, even though the water pump isn't damaged, and the matrix will cool rapidly because the coolant in it will not be flowing, but the cabin fan will continue to extract heat from the static volume that is there. The engine overheating will of course continue even faster if the coolant isn't flowing.

 

Maybe?

Edited 8 March, 201610 yr by Wino

^ I was thinking about this last night.

I couldn't figure out why the heater matrix would go cold at first, but then I had an idea...

 

If the coolant starts to boil, due to the stuck-closed thermostat, there will be gas bubbles pulled round with the flowing coolant to the water pump.  The water pump is designed to pump water, not gas, so once there is enough gas surrounding the impeller, it probably will completely fail to push any more coolant round the circuit (like the opposite of priming a pump).  This will mean that the coolant does stop circulating, even though the water pump isn't damaged, and the matrix will cool rapidly because the coolant in it will not be flowing, but the cabin fan will continue to extract heat from the static volume that is there. The engine overheating will of course continue even faster if the coolant isn't flowing.

 

Maybe?

 

The water pump is usually at the lowest point in the cooling system for this very reason, the only part which is lower is the radiator bottom hose. The thermostat is simply designed to shut off coolant flow from the cylinder head to reduce warm up time, usually the heater matrix is fed from the cylinder head as well.

I thought the 'short circuit' that doesn't include the radiator during thermostat closed operation usually does include the cabin heater though?

Maybe that's what you're saying.

Edited 8 March, 201610 yr by Wino

not usually because the matrix is often taken off the cylinder head.

Thermostat - heater matrix circuit isn't moved by the water pump ? It needs to open the thermostat first of all ?

 

Yep.

 

This is the diagram. As I can tell from the flow arrows, pump (4) pushes pushes the fluid independent from the thermostat (10), through the engine block to the thermostat, which if it's closed, to the heater matrix, then it gets back to the pump. It has nothing to do with the expansion bottle at all when the thermostat switch is closed. I am not sure how that ventilation pipe (6) is working. The overflow (with a closed thermostat) is getting back to the expansion bottle ?

Edited 8 March, 201610 yr by Alexandru

That diagram shows that the coolant circuit is isolated from the heater matrix by the thermostat which prevents flow out of the block.

The inlet to the heater matrix is coming from the thermostat.

 

EDIT: sepulchrave beat me to it.

Edited 8 March, 201610 yr by TMB

So it allows no flow in both directions, heater matrix AND radiator as well ?

Just start it from cold and feel the radiator hose and matrix feed hose where they come out of the coolant regulator housing. I bet the matrix hose starts to warm up well before the rad hose.

That was my initial thought as well, if there is no water flow through the heater matrix it would take ages to get the engine at 90 degree, open the thermostat and then have heat in the car during winter.

So it allows no flow in both directions, heater matrix AND radiator as well ?

 

Yes, the car warms up much faster.

It makes perfect sense then. I was thinking it was cycling somehow the coolant though the small heater matrix when thermostat was closed, at least to get heat in the car before the engine would be running at the standard temp, controlled by the thermostat.

Edited 8 March, 201610 yr by Alexandru

Do the experiment, then we'll all know.

Just start it from cold and feel the radiator hose and matrix feed hose where they come out of the coolant regulator housing. I bet the matrix hose starts to warm up well before the rad hose.

 

If that does work then I'm favouring your bubble theory :p

Na water pump is at the bottom of the block so its not going to be vapour locked easily.

The heater gets flow regardless of thermostat so that going cold indicates either air lock from boiling the coolant or the coolant isnt getting pushed around the system.. or its blocked with radweld or similar lol.

The hot side of the heater comes from the engine side of the stat so you get heat from the engine quickly without waiting for the stat to open.

Edited 9 March, 201610 yr by felicia16v

I know if from the winter time that I can feel warm-ish air coming through the vents even if the temp needle is not all the way up. I've done a 'test' today, ran the car for about 5 minutes, the needle was half way up, the pipe going to the heater matrix was hotter than the one going to the radiator. It's not precise though,  after I restarted the car the needle went all the way up. The main radiator was cold though.

If the water pump is successfully shifting water up through block and head, and thermostat is closed; apart from a small-diameter hose that goes to the expansion tank, there doesn't seem to be anywhere else except the heater matrix that it can flow to/through...

 

It looks to me like you might be able to poke a finger through the hole that the coolant temperature sensor lives in, into the start of the matrix feed pipe, to confirm that nothing blocks that, ever.

Edited 9 March, 201610 yr by Wino

So the pump should've been running hot coolant through the heater matrix when my engine overheated.

So the pump should've been running hot coolant through the heater matrix when my engine overheated.

 

No, look at the diagram you've posted, it has arrows denoting flow, the pump pumps coolant into the block, the outlet from the block is controlled by the thermostat, once the coolant in the block is hot enough the thermostat opens allowing flow through the rest of the system. Until the thermostat opens the pump simply cavitates which is why the coolant is often a bit frothy as it flows back into the expansion tank just after the thermostat starts opening.

 

I'm trying really hard to explain this incredibly simple concept as straightforwardly as possible, you have the diagram, look at it.

First photo in this post shows, to my eyes, that the thermostat cannot close off the feed to the matrix (the dark hole showing within the o-ring of the right-hand part).   :no:

First photo in this post shows, to my eyes, that the thermostat cannot close off the feed to the matrix (the dark hole showing within the o-ring of the right-hand part).   :no:

 

Look Wino, the thermostat is mounted tight against the engine block in that housing, it shuts off the coolant flow from the pump which is pumped through the block to the thermostat, once the 'stat opens it allows flow through those stubs on the housing to the matrix, the rad and the expansion tank as per the flow diagram above.

 

Edit: Actually forget about flow for a moment and think about high pressure and lower pressure, flow can only occur from a high pressure region to a lower pressure region, the pump is in the block and the water jacket is the high pressure region, the 'stat is at the far end of the block and everything else is in a low pressure region. Once the 'stat opens, flow can then take place but not before. If your 'stat is stuck shut then flow cannot take place but heat can spread on the low pressure side by convection because the pump is cavitating. Ok?

Edited 10 March, 201610 yr by sepulchrave