Coolant circulation.

Seems strange. Laws of physics and fluid dynamics dictate which way it'll go to an extent.

Conor formulated on Monday :

I thought I heard/read somewhere that it was usual to pump in the opposite direction to to the natural thermal flow.

Perhaps the explanation might be one of feeding the coolest water straight from the radiator, into the head first - the hottest part of the engine most in need of cooling?

Not on any car I've seen, the top of the radiators always hottest.

Given that hot water rises that would mean a hose from the bottom of the rad to the head - and I've never seen that.

Dave Plowman (News) wrote on 15/04/2009 :

You are correct, sorry was not thinking straight :-)

The forces circulating water thermally are tiny, and easily overwhelmed by a very modest pump.I'm pretty sure all the water cooled cars I have pump hot water in at the bottom of the radiatior and take it out, colder, at the top.

Ian

It's something I'd not really thought about - but on mine the top of the rad heats first as the thermostat opens.

"Dave Plowman (News)" gurgled happily, sounding much like they were saying:

On the Saab, the top of the rad is fed from the head via the 'stat housing, with an over-temp cutout switch for the aircon in the top hose. The bottom of the rad goes straight to the water pump, and thence into the block. There's a tee off the bottom hose feeding water from the turbo back into the circuit - taken from the head.

Yes, but the water is fed from the bottom of the radiator into the engine via the pump. It exits the engine through the thermostat into the top of the radiator - hence the top of the rad being hottest.

It's the bit that you also need to keep at the right temperature, which is quite hot if you're going to get it too run properly.

That's the same direction as the normal siphon action.

Which is what would happen with no pump - hot water rises to the top of any container.

Logically if you ignore thermo flow, the ideal would be the coolest water being pumped to the hottest part of the engine - ie bottom of rad into the cylinder head?

and "cold" water in rad sinks to bottom. I say "cold" because it's still over 60°C so quite hot enough to damage skin.

Overcooled head, undercooled block. Heat transfer to water depends on flow rate and temp difference. Higher temp diff (head - cool water) indicates a lower flow rate is needed for head. Now feed that warmed up coolant to the block and temp diff (block temp - hot water from head) is lower requiring a higher flow rate than before but head requires reduction.

It will tend to trap air as bubbles won't want to go down from head to cylinder block with water flow. Putting water in at bottom means any air pockets in head are scoured out to the thermostat housing and eventually vent to expansion space. This should occur during first engine warm up on a short test drive.

My head gasket closes all the holes in the head/block except the 6 round no4 cylinder - the cylinder furthest from the pump/stat. This forces the coolant to go past all cylinders and back though the whole head. There are small 4mm bleed holes to vent air from 3 or 4 places along the block. Earlier 8V SOHC version of the engine allowed coolant flow though all the holes, this meant the coolant tended to miss out on the trip round 3 and 4.

With good thermal contact between head and block as with a copper gasket there have been engines made with just the head liquid cooled. The heat from pistons is transferred to the bore walls and then conducted to the head though the gasket

Not necessarily, there's a lot of conflicting requrements & the heatflow in the head's easier than the cylinders.

Now I'm getting cross. How can I not be sure about something so simple. I'm going to have to go out tomorrow and run a car up to temperature to see what happens. Curse you, Plowman.

Ian

I am pleased you are not my GP

I thought it was simple too. ;-) Give me a nice straightforward auto to repair any day.

quick look in the workwhop manual for my van (iveco turbo daily) shows the flow diagram.....hot comes from thermostat to top of rad, where it is cooled and hence sinks downwards, sucked up by pump from bottom of the rad,

Indeed - that's the same as my SD1. However, at one point it seems the flow was reversed in the RV-8 fitted to Range Rovers and I wondered why. My car has a cross flow rad and I assume the RR the same.