can anyone explain how the funny vw cooling system works?

[Master ACiD]

i have never seen a car with a bypass hose which goes from the waterpump right to the upper coolany neck. at first glance this seems counter productive to cooling?

[QuickTD]

The thermostat is double acting, it closes the bypass hose as it opens to the radiator. When the thermostat is fully open the bypass is completely shut and all the coolant flows through the rad.

[chrissev]

IMO the only goofy aspect of it is the oil heater (cooler?).  Flow through it on the golf/jetta is into the bypass hose which is blocked when the engine is hot.

Not really.  The oil cooler has two coolant inputs.  One goes to the bypass hose (which connects at the top to the main coolant input to the engine).  The second connects to the return line from the heater.  The return line from the heater also connects to the base of the water pump, which connects to the bottom of the radiator.  This design allows the thermostat to regulate the flow of coolant to the oil cooler when the engine is cold or warm (ie, change where it comes from).  When the engine is warm, cooled coolant flows out of the heater core return hose, into the oil cooler, and then into the bypass hose, which is connected at the top to the hot coolant return line to the radiator.  When the engine is cold, the flow is reversed.  Warm coolant flows into the oil cooler through the connection to the bypass hose, and then out into the heater core return line and down to the water pump, which then recirculates it back up through the bypass line.  This has the effect of warming the oil when the engine is cold, and cooling it when the engine is warm. Those wacky germans eh?  I love their designs.  Anyway just thought I'd point that out.  

Chris

[chrissev]

The problem with the way you described it is you would be expecting the hot coolant to do something it never does by itself, ie flow downwards.  The cooling system is a thermal syphon assisted by the water pump which sucks from the bottom of the rad.  The coolant that goes out the bottom of the rad and into the pump is therefore cold coolant that has sank to the bottom of the rad after being cooled by the air passing through it.  Hot coolant will never, by itself, flow downwards.  I think what happens is that some of the coolant that flows back from the heater core is sucked up through the oil cooler and into the main hot coolant return line that follows the thermal syphon into the rad and then downwards as it cools towards the water pump which assists the flow.  To have hot coolant flow down the blocked bypass and into the oil cooler would mean that the hot coolant is flowing downwards (which it never does) and against the prevailing flow of hot coolant, which is upwards, into the main return line, into the rad, and then down into the water pump as it cools.  The prevailing hot coolant flow direction, when the thermostat is open, is up, through the engine as it warms up, catching heat as it rises to the top in the cylinder head, out the main return line, and into the top of the rad (and it would smoke away as steam if there was a hole anywhere, because that is where it wants to go, straight up).  There would never be a tendency for the hot coolant to flow downwards as it heats up.  The only reason the hot coolant  flows downwards is because it is cooled rapidly when it enters the rad, and the cooled coolant is pushed down (and sinks all by itself also) by fresh hot coolant continuously entering the rad from the engine.  That is why the design of the oil cooler is so ingenious.  Whoever thought of it saw the big picture and realized that hot coolant would never flow downwards and any cool coolant that reached the oil cooler would necessarily, as it went through the oil cooler and became hot, rise upwards towards the main return line to the radiator.  

Now when the thermostat is closed you have the same effect.  The only difference is the hot coolant coming out of the top of the cylinder head is forced back down through the bypass into the bottom of the engine (If the thermostat ever failed to open, what eventually would happen would be that the top hose to the rad would burst, and the coolant would all flow out that hose, mainly as steam, which would rise straight up, like it always does.  None of it would flow downwards.  The oil cooler would probably be quite cold, while the hot coolant gushed out of the burst top hose).  But as hot coolant, it again doesn't want to go down, it wants to rise.  So it takes the path of least resistance, through the oil cooler, and then is again forced down into the water pump which recirculates it through the engine.  The whole system is designed around the assumption that hot coolant will not go down willingly.  It always wants to rise.

Chris

[Master ACiD]

well heres what im thinking guys. i live in south florida. it gets to be 40 degrees celcius here in the summer, and we dont have a winter.

im thinking of disabling the heater core, because it leaks and i dont need a heater anyways. then im thinking of removing the heater hoses and blocking off the ports just to clean up the engine bay a little bit.
then in order to keep the overflow tank functioning i would just hook it up to the flywheel side of the cylinder head.

do you guys see any problem with this or does this sound like a bad idea? im open to suggestions.

[QuickTD]

im thinking of disabling the heater core, because it leaks and i dont need a heater anyways. then im thinking of removing the heater hoses and blocking off the ports just to clean up the engine bay a little bit.
then in order to keep the overflow tank functioning i would just hook it up to the flywheel side of the cylinder head.

do you guys see any problem with this or does this sound like a bad idea? im open to suggestions.

Don't do it, at least not the way you described. If you wan't to defeat the heater core just connect the hoses together or place a short rubber elbow from the coolant neck on the flywheel end to the metal pipe that runs to the water pump. You can cut one of the heater hoses to suit. You  don't want to dead end the coolant fitting on the head, the number 4 cylinder will get very little flow if you do.

[QuickTD]

They have a valve in the line from the flywheel end of the head to the heater core. When that valve is shut all coolant flow on that circuit is stopped.

Wasn't aware of that, I've never eyeballed a north/south diesel in person. I guess it can be plated off then. It was my understanding that the coolant entered low on the front and exited high on the rear of the engine. I would have thought the flow would get a little stagnent around #4 (like it does in a 2.2 dodge...) with the heater hose blocked. But if the factory does it, it should be OK.

[chrissev]

Now I'll say you're completely wrong.  The force of thermopsiphon has almost no effect on the VW cooling system.  Actually the force is so weak compared to the waterpump that you could effectively say it has *absolutely no effect* and you would be right except perhaps in the case of the radiator where it *might* reduce turbulance *slightly*.  The vanagon radiator actually draws cool from the *top* and inputs the hot at the *bottom*.  The force of thermosiphon is so weak that it has *no aprreciable effect* on the flow through the oil cooler.  In fact if thermosiphon were the force of movement through the oil cooler, the cooler would be utterly and completely useless due to the very small surface area and the temperature differential required to move coolant against the pressure differential of the waterpump.  I used to imagine that the flow followed thermosiphon as you say until I was educated about it.   Nope, you're totally and completely wrong about the coolant flow path.

Andrew

I don't think I'm wrong.  The vanagon engine sits on it's side doesn't it?  I've never seen one so I'm just guessing.  But you know that hot coolant rises and cool coolant sinks right?  And the cooling system is partially pressurized so some of the coolant that is right at the top, if it was allowed to escape, would do so as steam?  Think about it.  Steam rises, it doesn't sink. Hot water sits on top of cold water.  There would never be a situation where hot coolant, that is beyond it's boiling point but not yet steam because it is under pressure, would flow downwards and displace colder coolant.  It's just physics.  

The water pump is definately necessary to move the cold coolant back into the engine, but the hot stuff is going to rise all by itself.  Don't need the pump for that.

[chrissev]

Not sure on that MasterACid.

A couple more thoughts on the thermosiphon effect.  

If it were a dominant factor in the flow path, then there would not be any flow to the waterpump prior to the thermostat opening, and so the thermostat would never open as it would never get any hot coolant to it, etc...

not really.  As you probably know (if you've ever installed a thermostat) the element is on the inside.  So it takes its temperature from the coolant that is coming back down the bypass line, not from the coolant that is in the radiator.  

With regard to the force of thermosiphon as opposed to the force of the waterpump, imagine a 220v hot water heater element in a still water tank.  How many ripples form on the surface even when the element in near the surface?  Now imagine a waterpump, how many ripples are formed when it is near the surface of the water?

The hot water at the bottom of the water tank would rise to the top as the 220 volt heater heated it.  It's just physics.  

As I mentioned before, in order to push the water up to the high pressure hose the coolant would need to be well past the boiling point in the oil cooler circuit and thus would be utterly useless as a cooler.

Andrew

not really.  The coolant is cool when it comes into the oil cooler.  It then takes heat from the oil, making it warm, then it rises.

[BlackTieTD]

well heres what im thinking guys. i live in south florida. it gets to be 40 degrees celcius here in the summer, and we dont have a winter.

im thinking of disabling the heater core, because it leaks and i dont need a heater anyways. then im thinking of removing the heater hoses and blocking off the ports just to clean up the engine bay a little bit.
then in order to keep the overflow tank functioning i would just hook it up to the flywheel side of the cylinder head.

do you guys see any problem with this or does this sound like a bad idea? im open to suggestions.

sounds like a plan. only thing i will add is that some early mkIs has non-overflow radiators... meaning they had a filler right on the radiator and no need for an expansion tank. if you want to clean up under the hood you could try eliminating the overflow altogether and see if you have any cooling problems. i have not done this personally (yet), but i have spoken with people who have, and despite their worry that the car would run hotter, in practice, it runs at approx the same temps.

[fspGTD]

The carburated VWs are the ones that came with overflow tankless radiators.

I run a carburated VW Radiator on my 1.6lTD for weight savings and to clean up the engine bay... no cooling system problems.  But - without the clear plastic overflow tank, you can't as easily check the coolant level.  What I usually do to do a quick fluid level check is give the upper radiator hose a squeeze.  Can usually tell that way if there is coolant inside it and if the pressure is right.

[Red Rabbit]

I've got a 1984 Jetta with 1.7 cis throttle body, it came stock without the overflow bottle style.

[935racer]

As far as I know all mk1 sciroccos came with the built in over flow tanks in the rad. I use them on all the vw's I own, I hate the gross looking white overflow tank. I retail them new for $160CAD if anyone wants one.

[chrissev]

My comment regarding the thermostat never opening was in regard to the fact that if thermosiphon were the motivating force as you say, then the Hot coolant would never flow *down* the bypass line to get to the thermostat to open it.  Thus no flow.  That is obviously not the case.

It is the case.  Hot coolant never flows down the bypass line.  Warm coolant is forced down by the pressure created by hot coolant coming out of the engine and also assisted by the water pump.  The hottest coolant stays at the top.  The thermostat opens fairly early in the coolant's heat load profile.  

I understand very well the physics of thermosiphon.  It just simply does not create the pressure differential to ever come remotely close to overcoming the force of the waterpump.  I understand heated water rises in a water tank, but it doesn't jet above the surface to the the foce of the thermosiphon.  The hot water pipe in my house then goes down the wall into the crawlspace and then to the faucets in the house.  Again, I do get hot water at my faucets despite the fact that the hot water flows downhill in the pipe from the waterheater.  The reason is because the force of thermosiphon is very very weak.  It is dependant on the expansion of the molecules by heat making the substance lighter.

the hot water in your house pipe is not pressurized and is not beyond it's boiling point.  The coolant in the car's cooling system is.  This creates a strong thermosyphon that is assisted by the water pump.  The main pumping action occurs at the base of the rad and into the bottom of the engine.  There is a rather small amount of suction in the heater return line (look at a disassembled pump and you will see the hole is small and off centre).  The pressure differential is stronger in the car's cooling system than in a house hot water system because the car's cooling system is under pressure due to the coolant being beyond it's boiling point and under pressure so it doesn't boil.  

I will be installing an active solar panel water heating system on my house shortly.  Even passive systems will benefit in efficiency from a circulating pump.  Active systems where the panel is mounted above the tank require a pump to move the water against the thermosiphon effect.  The pump does not need to be anywhere near as forcefull as the vw diesel waterpump to overcome the thermosiphon effect.

That is fine and true as long as the water isn't boiling.  If it is, then the thermosyphon effect is huge.  Look at nuclear electricity generation stations.  The turbines are steam driven.  There is no outside force that pushes the turbines, just super heated compressed steam.  It spins those turbines and makes mega watts of electricity.  All it is is boiling water under pressure.  

Again, to go against the flow of the waterpump the coolant in the oil cooler line would need to be well above the boiling point to ever produce remotely enough force overcome the force of the waterpump pulling it in the other direction.  *It's just physics*.

not really.  Hot water always rises, so as the cool coolant gets warmed up in the oil cooler, it is going up.  It is sucked up by the strong thermosyphon occuring at the top radiator hose.  

I don't have time to get the hard figures on the force of the waterpump and the pressure differentials in the system.  A pressure gauge at the inlet of the pump and the head outlet would show it.  I also don't have time to look up the maximum pressure differential caused by thermosiphon.  I merely present the info in hopes that your misinformation does not delude others.  I'm done with this.  If you want to still carry the torch and have the time, how about bringing back some test results regarding the pressure differential at the waterpump and what pressure differential can be achieved by thermosiphon.  Short of some hard facts presented I am done with this conversation.  I initially explained the coolant system much the way you do until, as I mentioned previously, I edjucated myself about it.

Well that is fine.  But you must admit, even if you don't agree with my description of the mechanism, that the oil cooler does somehow manage to cool the oil and does not heat it.  I know this because I have run vw diesels with and without oil coolers and measured the oil temp and the cooler reduces it.  So it is in some magical way cooling the oil.  It is also important as we all know that if you run a turbo on a 1.6 diesel that you have an oil cooler somewhere in the turbo oil line otherwise the oil will be too thin and wreck the floating bearing in the turbo.  

 I am *completely sure* (which is actually rare) that the results of any testing or further research would show that you are wrong.  Again, *it's just physics*.  I don't like beating a dead horse and will not repeat myself more.

I don't think so.  If I am wrong then the oil cooler does not cool the oil, it heats it.  That means we should all remove our oil coolers as they don't do any good.  This is of course wrong.  And I am keeping mine in place because I know it cools my oil.

[chrissev]

sounds like a plan. only thing i will add is that some early mkIs has non-overflow radiators... meaning they had a filler right on the radiator and no need for an expansion tank. if you want to clean up under the hood you could try eliminating the overflow altogether and see if you have any cooling problems. i have not done this personally (yet), but i have spoken with people who have, and despite their worry that the car would run hotter, in practice, it runs at approx the same temps.

I did that on a 79 rabbit diesel after cracking two expansion tanks.  Got sick of replacing the things and got a rad out of a 77 rabbit with an internal expansion tank.  It worked fine.  No difference in cooling.