Coolant loss

[lambert.john]

Ever since I have had my Excel, I have had to top up the coolant regularly. It seems that when the coolant expands with the heat, some is transferred, as it should, from the first chamber of the header tank to the second chamber. However, it does not return, as it should, to the first chamber when it cools. I have cleaned out all the crud from both chambers, checked that the connecting pipe is clear, and replaced both caps, but no improvement. I can only think that there must be a pinhole in the tank somewhere (it LOOKS in perfectly good condition) that allows air to be sucked into the first chamber rather than coolant from the second chamber. Anyone else had this problem?

[amarshall]

Which cap is on which chamber?

[lambert.john]

Interesting suggestion, Angus! However, the caps (on my model, anyway) are not interchangeable. The pressure cap is on the second (overflow) chamber.

[Lotus-e-Clan]

So would this sum up the situation?

• 1) the caps are on correctly and tightly
  2) the expansion tank appears in good condition /no obvious leaks /coolant stains
  3) coolant transfer to the expansion chamber occurs correctly WHEN HOT
  4) coolant recovery from the expansion chamber FAILS as the engine cools
  5) you need to top up the coolant - therefore you must have an ABSOLUTE loss of coolant on every heat cycle
  6) there is no sign of coolant loss on the floor or dried coolant stains around hose joints or waterpump bearings

With the engine COLD I would whip out ALL of the spark plugs and look out for water ejection (however minor) from the plug holes as you (or mate) crank over the engine (place thumb over hole to detect smaller water drops).

You could have an unevenly clamped /failing head gasket. A failing /unevenly clamped head gasket will seal properly when hot but when the engine cools the pressure in the water jacket is enough to force water past the failing fire ring of the head gasket and into a cylinder or adjacent cylinders as the aluminium head and Nikasil liners contract and therefore clamping pressure subsides.

Alternatively have a feel around the 'nose' of the water pump. If the water pump seals have failed it will feel wet around the 'weep hole' on the nose of the water pump. You won't notice slight coolant loss when hot because it readily evaporates - but your water pump will have a slight runny nose when the engine is cold -but not enough show as a pool on the ground.

[lambert.john]

No, I don't think it's that serious. The loss is not "total' as it remains in the second (overflow) chamber until, of course, that overflows as well! When I can be bothered I siphon what's in the second chamber back to the first, manually.

[AndyC]

Could it be there's a leak between the two chambers, in the lower half of each? That would prevent water getting back via the linking tube?
Is the tube OK, as in not blocked? If partially blocked it may be enough that water can be pushed past it when the engine's hot but too much for the vacuum to pull water back when the engine's cold.

[Lotus-e-Clan]

No, I don't think it's that serious. The loss is not "total' as it remains in the second (overflow) chamber until, of course, that overflows as well! When I can be bothered I siphon what's in the second chamber back to the first, manually.

Ah! There is no loss of coolant ....

The thread title misled me a tad!

Maybe as Andy says, you have a rust flake/debris in the expansion tube that is acting as a one-way flapper valve? And /or the blanking cap is leaking air but not any significant water?

Pressure test the tank. Use a bike pump / car tyre pump connected to a suitable hose. The hiss from the loss of air will be obvious. ..and the pressure will collapse.

[lambert.john]

Thanks for the input, guys. I will probably bite the bullet and get a new tank; S J Sportscars have them available off the shelf.