Coolant system pressurising

[Toilet Duck]

Morning all,

I have an issue with the Honda K20 (civic type r) engine in my Ariel Atom. Driving on the road everything is fine. When driven hard on track, the coolant system is pressurising. The coolant level in the expansion tank rises and eventually leaks out of the cap. Air was present in the radiator at the front of the car (I opened the bleed bolt in the top of it). I tried a new cap but the symptoms persisted, I then drained and refilled the cooling system. I got a garage to do a sniffer test on the expansion tank for combustion gases as I feared head gasket woes. The car was left running for around 20-30mins and the fan was kicking in, so up to "normal" temp. No trace of combustion gases were present. They suspected I had had an air lock that had been resolved when I drained and refilled. Went to another track day and after the first session, I checked the expansion tank; the levels had risen to the top again and coolant was escaping out of the cap, air was again present in the radiator. Letting the air out the rad causes the coolant level in the expansion tank to drop (with the expansion tank cap off).

I took the car to a local trusted garage and got them to bleed the system and do another sniffer test. Again, no trace of combustion gases present. I got the car back and drove at 70mph in 3rd gear for a few miles, with some hard acceleration when appropriate to try and simulate track conditions as best I could given the constraints of driving on the public road. The coolant level started to leak out of the expansion tank cap. I let a load of air out the radiator and the coolant level in the expansion tank dropped back to normal. I immediately repeated the test and got exactly the same results; another load of air in the radiator and coolant leaking out of the expansion tank cap. Bubbles were visible periodically rising in the expansion tank.

At no point did the coolant temp rise above "normal" readings, and driving "normally" for road use the problem does not manifest, it only happens under hard use.

I'm not sure what to do next. The vehicle has a mechanical and electrical water pump. The electric pump is working, and I assume the mechanical one is as well. The fan is operating normally. My gut is saying it's a head gasket issue causing the coolant system to pressurise, but the lack of gas detection on the sniffer test goes against this. However, the sniffer test has not been carried out after "hard" use. Is it possible a dodgy head gasket can only leak combustion gases under hard use and not "normal" use? I'm wondering whether to change the thermostat, but I don't see how that would cause air to forum in the top of the radiator?

Any help or advice welcome, got some charity days coming up soon so need to get this resolved!

Cheers

[SteveSlowboy]

As a first stab I would try replacing the radiator cap - perhaps it's not holding the pressure required?

[Toilet Duck]

SteveSlowboy said:

As a first stab I would try replacing the radiator cap - perhaps it's not holding the pressure required?

There isn't a radiator cap, just a bleed bolt on top of radiator. However, I have replaced the expansion tank cap (as stated above) but the coolant is still being forced out past it.

Cheers

[stevieturbo]

Could be many reasons.

System filled too much from the outset.

Air locks within the system and no way to self bleed

Inadequate or malfunctioning stat/rad. OEM type temperature gauges are woefully inaccurate and slow, so not always a good indicator.
How are you reading temp ?

First simple test is to run the system with slightly less water.


And when you way water is blowing past the cap, surely there is another reservoir for this water to collect in, and then be drawn back into the system when it cools ?
Which is fairly normal behaviour

[anonymous-user]

How big is the header tank?

If you've an engine at the back, and a rad at the front, chances are you've got something like 2 or 3 times as much water in the system as for a std engine/rad at the same end of the car (with the same amount of extra expansion!


As a test, let the coolant escape, but DON'T re-top up the header tank. Let it cool, then drive it again and see if the coolant level returns to the same place as before

[mighty kitten]

One thing you could do is use the bleed screw to test the cap if you can rig up an adaptor but if it's the normal jap type small cap with the two way valve to draw coolant back in I'd just get a new one , they're only a few quid from Honda

[AER]

That has all the hallmarks of air remaining in the system. Coolant will expand some with temperature, but not all that much. Put some air in the system, especially in a region that traps it and you'll have happen exactly what you are describing.

[mighty kitten]

With the rad vertical it's hard to get any air that ends up at the front out . My rad is vertical but I have a small bleed line that runs from the top of the rad all the way to the header tank at the back so it can self bleed . Elise , mr2 and mgf sit the rads at a shallow angle or horizontal to make it easier for the water pump and let any air find its way back to the header

[Steve_D]

It could be that your engine has a hot spot somewhere which is locally boiling the coolant but only when giving it beans.
Second issue is that the air/gas caused by the boiling is getting trapped in the top of the rad so, as stated elsewhere, connect your rad bleed point back to the header tank.

If you still feel you have a head gasket issue get a garage to do a 'leak down' test which connects an airline to each cylinder in turn. In your case you will be looking for bubbles in the header tank.

Steve

[mighty kitten]

Are you losing coolant ?

[Toilet Duck]

Cheers for all the replies, much appreciated.

To answer some of the questions:

The expansion tank holds, at a rough guess, around 1 litre. Unlike most normal systems, there isn't a return feed to it. There is one connection, a hose attached to the bottom which must feed into the coolant system. The system doesn't appear to have any "self bleed" mechanism that I can tell. There is a single outlet/inlet connection at the top of the expansion tank but this is capped/blocked off. I checked with the factory and this is correct.

The radiator is vertical at the front, and I am not loosing coolant from anywhere other than via the expansion tank cap (which I have replaced) when the engine is driven hard.

I am measuring the coolant temp via the dash display, and also via the OBD port using a Bluetooth dongle and a tablet (there is approx 4 degrees difference between the two, one sensor for the dash and one for the ECU).

I will investigate further and try some of the suggestions and report back.

Thanks again

[anonymous-user]

Is running two water pumps having any effect? when I run a Craig Davies all the instructions said not to run the mechanical pump.

[k20erham]

Hi please do make sure the coolant has the correct mix of antifreeze in it too as that along with running at as higher pressure as you can raises the boiling point, so if it is an engine hot spot that's the problem correct coolant mix and the right operating pressure will help, as a poster has said some expansion tanks actually vent into another catch tank then pull it back when it cools, there is a valve in the cap, I have a K20 in my Caterham and I had to put a 15psi cap on to contain the pressure

[CarsOrBikes]

Replace the thermostat, if it's sticking partially closed it will cause that, I had the same problem on the Honda v6. If you do and still see a fault, remove and check the water pump (mech) in case the impeller is detached or under rotating even broken. Only my2p

[Jimmyarm]

Has it always done this or has it suddenly started doing it ?

Another couple of things to check - if the electric fan is two stage, are both stages working correctly (it's possible only the slow speed is working and the resistor has failed so the second speed doesn't kick in) and have any heat shields been removed near coolant pipes and the exhaust.

What temperature are the readouts giving when it is boiling over ?