Coolant ideas - swirl pot or header tank

[tvrolet]

OK, time to start 'wiring up' the cooling system. Engine is a small block Chevrolet, but the principles are the same as anything else.

There is no heater, and although the flow leaves the block through the thermostat housing, there's no thermostat, just a restrictor plate.

Interested in the pros and cons of the 2 solutions I can see, and I've diagrammed them below.

[pic]http://www.indexcomputers.co.uk/cool.gif[/pic]
First option is just to have a swirl pot, with the water level above the thermostat. All the flow goes in the top of the tank, then out the bottom.

Second option (which I was told is better to keep the system purged of air) is to run a small bore tube out the thermostat to a header tank, then another small bore (but a bit bigger!) tube back to the pump.

Got to say I favour the first option as it's so simple, but open to suggestions, comments etc.

WB

[GreenV8S]

Are the heights accurate i.e. bottom of swirl tank no lower than top of radiator? as long as the pipes from the top of the rad and from the thermostat housing both go up hill all the way to the filler cap, I don't see any paricular problem either way. The swirl tank design would probably fill and vent better, the separate header tank would give slightly better degassing, but not a huge deal either way.

Any particular reason not to use a stat? I strongly recommend you do use a stat. If I were you I would ditch the restrictor plate and put an inlet stat in parallel with the radiator, this will allow you to run maximum water flow without worrying about extended warmup times or overcooling under light load conditions. Rover stat PEM101020 (84 deg nominal) would probably be suitable.

Your diagram doesn't show any provision for conventional bypass or heater feeds. Removing these isn't inherently harmfull (as long as you arrange adequate flow when cold, as you have) but be aware that some pumps carefully have these inlets positioned at the top of the pump and use them to vent air from the body of the pump when filling. If you use one of these pumps and just block off the inlet ports you may find the pump is very reluctant to prime and stops working the moment any air gets into the system.

[tvrolet]

GreenV8S said:

Any particular reason not to use a stat?

Only that the engine was built and dyno's without one...and the engine builder (who in fairness specialises in race engines) doesn't use them normally.

GreenV8S said:

Your diagram doesn't show any provision for conventional bypass or heater feeds. Removing these...

They're not there to be removed. Car never had heater (Tuscan) and isn't going to get a 'water-fed' one either. (I have other plans ) Pump is a nascar style item, with only a single inlet, and outlet to the block. It has the provision to open/tap other channels, but I hadn't planned to do so.

WB

[eliot]

my sbc has the expansion tank attached to the top of the radiator (ala range rover classic because thats what it is). Never had any problem overheating or with air locks etc. I run a thermostat and a heater. Heater is plumbed into said hole in top of pump and returns to inlet manifold near thermostat. System has plenty of thermal load placed on it, including 400+bhp, water cooling a pair of turbos and having two large oil coolers placed in front of it.

[shpub]

The 520 does the header tank with a tee in parallel to the heater. I also run without a stat due to advice from my engine builder. I do block off the rad though to keep the temp up on cooler days. Without the stat is fine on the track though.