13 replies to this topic

[TheHunterHofman]

Hi all,

 

As the title suggests; I still have exhaust gasses coming through the coolant system, as my CO2 test yields negative results.

 

After replacing the head gasket (with a Payen 450 one), I have had the cylinder head skimmed; this didn't fix the issue (still negative results). Then I had the cylinder head pressure tested; they found a leak surrounding one of the plugs (see below). They then refitted new plugs and skimmed it again. This also yielded negative results with the CO2 test.

 

Does this possibly mean my engine block is cracked somewhere?

 

Here's a video I made about it: https://youtu.be/iiDQu8jehb8

 

Relevant timestamps are:

 

02:58 (skimmed cylinder head)

08:10 (cylinder head with fixed plugs too)

09:06 (footage of the leak in the cylinder head)

10:28 (co2 test and look at the water in the coolant reservoir)

 

Additional remarks: I have torqued the cylinder head correctly, the oil is not milky whatsoever, there is a very light film of (what looks to be oil) on the water in the coolant reservoir and finally, although I have flushed the water in the coolant system directly beforehand, it quickly turns brown when letting the engine run and get up to temperature.

 

If anyone has any ideas or suggestions, kindly let me know; any input is appreciated.

 

Regards,

 

TheHunterHofman

 

[Mr Piggy]

Have you pressure tested the cooling system? Leave it under pressure for some hours, take the plugs out and see which cylinder is getting coolant into it, or where else it's going. A bore scope would be useful in this case. I had the brass plug problem on a 1300 mini years ago. The plug had actually sunk into the head a few thou. It used to start on 3 cylinders until it cleared the water off the spark plug.

[TheHunterHofman]

I have not, no. Thanks, will be trying this out. For whatever it’s worth by the way: I have done a compression test on all cylinders and all had very similar readings (all around 185psi, when cold).

Someone else suggested that I should also get myself a new head gasket as the current (new) one has already been torqued down one or twice (after replacing the cylinder head multiple times), what are your thoughts on that? Can’t imagine having to buy a new head gasket, the one in the video has run for maybe 40 minutes in total.

[Spider]

Head gaskets in my books are single use only, unless it's a Cometic. Some in a pinch can be re-used, maybe once more, but that's about it.

 

If you take the head off again, can I suggest rubbing the deck of the block over with an Oil Stone, just say 3 - 4 wipes to start with. This will show you if it's truly flat or not (forget the steel rule / feeler gauge method - they really need to be flater than that).
 

When torquing up the cylinder head nuts, I'm not sure of your technique here. The factory suggests pulling them down in 2 stages, the first to 25 ft/lb (34 Nm) and the 2nd to 50 ft/lb (68 Nm). I've always done them in 4 stages, the first at 15, then 25, then 35 and finally 45 ft/lb. Be sure to oil the threads and under the flanges of the nuts. Don't add coolant to start off with, just plain water, do a couple of heat cycles (the gaskets have a heat activated sealant), the let it cool, drain the water, flush it, while it;s still cool, re-torque the nuts (back them off one at a time, 1/4 of a turn, oil them, then pull to torque), then re-fill with coolant.

[TheHunterHofman]

Head gaskets in my books are single use only, unless it's a Cometic. Some in a pinch can be re-used, maybe once more, but that's about it.

 

If you take the head off again, can I suggest rubbing the deck of the block over with an Oil Stone, just say 3 - 4 wipes to start with. This will show you if it's truly flat or not (forget the steel rule / feeler gauge method - they really need to be flater than that).
 

When torquing up the cylinder head nuts, I'm not sure of your technique here. The factory suggests pulling them down in 2 stages, the first to 25 ft/lb (34 Nm) and the 2nd to 50 ft/lb (68 Nm). I've always done them in 4 stages, the first at 15, then 25, then 35 and finally 45 ft/lb. Be sure to oil the threads and under the flanges of the nuts. Don't add coolant to start off with, just plain water, do a couple of heat cycles (the gaskets have a heat activated sealant), the let it cool, drain the water, flush it, while it;s still cool, re-torque the nuts (back them off one at a time, 1/4 of a turn, oil them, then pull to torque), then re-fill with coolant.

 

Thanks Spider. I'll order myself a new Payen head gasket (either a BK450 or an AF470).

As mentioned in my original post, I have torqued the head down per factory spec (the first round to 25 ft/lb (34 Nm) and the 2nd round to 50 ft/lb (68 Nm)). I will try your method next time, as well as the couple of heat cycles.

Any further suggestions or ideas?

Thanks again, much appreciated.

[GraemeC]

I do the first heat cycle dry - just run until you can feel the head start to warm up - 30 seconds to a minute.  The allow to completely cool before adding plain water and heat cycle again.

If you go wet to begin with there is a chance the water will get under the gasket and the sealant won't stick

[TheHunterHofman]

I do the first heat cycle dry - just run until you can feel the head start to warm up - 30 seconds to a minute.  The allow to completely cool before adding plain water and heat cycle again.

If you go wet to begin with there is a chance the water will get under the gasket and the sealant won't stick

 

Noted, thank you.

[TheHunterHofman]

Can anyone enlighten me with how many heat cycles I should do before re-torquing the cylinderhead nuts? I just let it run with the new gasket for 2/3 minutes (without water) and it is now cooled down again.

[GraemeC]

I would re-torque now.  Then fill with plain water and run again - you can either do this as a heat cycle or run with the plain water for a little while and then drain and fill with coolant - you could retorque again at that point, but there would be little to gain in doing so.

[TheHunterHofman]

So, I've installed the new headgasket (Payen BK450), retorqued as per Spider's instructions, heat cycled dry, retorqued the head nuts after letting it cool down completely, heat cycled again with water, let it cool down completely again and still the CO2 test gave out bad results (i.e. it turned yellow after about 10 minutes of running).

 

My thought would be that there must be a crack in the engine block somewhere, or am I missing something? I didn't note any water being evaporated after 15 - 20 minutes of running; could it possibly be that the CO2 test is unreliable? The car didn't overheat (looking at the temperature level of the water, it stayed bang on in the middle).

Apologies for coming back on this again, but really running out of ideas (and motivation) here now.

Thanks in advance for any input.

[Spider]

If there's no visible big bubbles coming up through the coolant in the radiator, there's no oil in the coolant, no coolant in the oil and it doesn't overheat, then if it were me, I'd drive it but keep an eye on it for a while.

It certainly sounds like it's doing all the right things.

[bpirie1000]

Maybe try your test solution system on a car you know is operating correctly and make sure the test kit is giving a tip top result......

[whistler]

I have had a pinhole appear in the inlet port of a cylinder head due to someone grinding too close to the water jacket and eventually it broke through. Could also happen the same way in the exhaust port. Old heads do get rusty internally. Maybe have the ports pressure tested.

[weef]

Reading through this post you appear to have covered all the bases one would go through in trying to correct/detect a combustion to water leak.

With no showing of overheating etc I suspect there is no problem with your engine, rather in your understanding of your test results.

To get a datum connect your test test gear directly to the exhaust tailpipe so some of the gasses pass through the test fluid, note the colour of the fluid. Now take a sample of the radiator water and mix this with some fresh test fluid, sometimes this will generate a change of the test fluid colour.

Compare the colour of two datum fluids with that of the test fluid you have from your engine running test and note which end of the colour spectrum your sample is.

This colour comparision may give you some indication if you have a contamination problem. 

With your tests lasting 15 minutes or so can allow the steam from the radiator to condenste out and contaminate the test fluid rather than measuring exhaust gasses.

If you have a friendly MOT station nearby you could ask them to "sniff" at the radiator cap with their emitions tester to see if anything untoward can be detected.

This is my take on things and I hope it helps.