Filler Cap worries, Overheating ?

[rfs1957]

There was an interesting exchange a while back here ......

viewtopic.php?f=3&t=9042&p=72924&hilit= ... +up#p72924

regarding Water Consumption and catch-tanks etc, but my recent experiences with filler-caps might warrant a separate thread as they might apply to many of us.

The 5.500km I did (over 2-weeks in the second half of July)

viewtopic.php?f=13&t=10951

involved all sorts of speeds, ascensions, and ambient temperatures around Europe, albeit mainly Northern, and the car temperature remained an impassive 80° to 85° throughout.

The car has a Kenlowe set at 95° which would kick in when the engine was stopped, as the heat surge swelled up, then run for a couple of minutes - probably unnecessarily but it seems a good way of checking and being reminded that the thing is still working.

Before leaving I had fitted a new 4 lb coolant cap, and put the "proven" previous one that was less than a year (1.000km) old into the spares I carry, reckoning it would be a reliable "safe bet".

I would check all fluids every morning, and until the very last day the car did not drink even a tea-spoon of cooling fluid - I can be sure of this as I run the header-tank "overflow" into a plastic bottle in the wheel-arch which means the header-tank can actually be "over-filled" to make the level very visible/tangible, as when it expands and spits off the excess, this excess remains contained within the overflow bottle and is sucked back in with a loud burp after the engine has cooled down at the end of the day.

During the last 300-odd kms south of Clermont Ferrand, the A75 goes up some fairly demanding steady climbs and the car would usually rocket up these effortlessly at 80mph ; it was a hot day, I would guess well into the 30's, and as I'd got used to ignoring the temperature guage my first warning that all might not be well came from feeling the car was slightly down on power.

Probably suffering from the "we're nearly home syndrome", somewhat like the "I am going to land the plane first time" beloved of pilots, I put it down to the beginnings of a whiskered plug (which I've suffered from with N12YCs before, and had assumed was due to my oil-consumption) and just nursed the car more carefully, whilst watching a rising temperature-guage go past 100° and then drop back over the next 5 minutes after each climb.

The last 150kms of the A75 are easier on the car, and so we got home at a steady 70mph and didn't think much more of it - and 5.500km, mostly with the hood down, leaves its mark on a man so I didn't touch the car for a week afterwards.

So it was a surprise to discover that there were now a good 4 litres of coolant missing from the system, with the attendant worries regarding whether what I had actually felt was the engine tightening up ........... as the plugs looked fine.

There was no excess coolant in the "overflow" bottle, but the engine can spill 5 litres out and still leave nothing to see in it as it only has a 1-litre capacity, and the it sucks a good litre back in when it burps upon cooling down. So you could lose 5 litres from active service, and see them reduced to 4 once that litre has gone back in - and I think this is what happened.



My immediate thought was "head-gasket", so I fitted a gauge I use in place of the sender-unit (5/8 UNF in case you need to do the same thing) on the inlet-rail and pressurised the system .............. only to find this was virtually impossible, as the header-tank cap would blow off before any pressurisation, at less than 2lbs, and this with a virtually new 4 lb cap.



The "spare" I had carried around Europe turned out to let go at even less, as far as I could tell at Zero psi !

I have already had, in the past 10 years, issues with brand new filler caps lasting ridiculously short periods before they leaked, but had imagined that by sourcing more recently from a Usual Suspect that these issues were behind us. I mean, a Filler Cap ?! The last one I had on the Mini lasted 20 years ..............

I then tried a later 9 lb cap (3.8's went from 4 to 9 lbs at 85067/879044/861091/888241, according to J30) which started to leak at 7 psi, but was then able to maintain that pressure and am reasonably sure that - at cold at least - the cooling system is water-tight.



I'd be interested to hear what other 3.8 owners run - 4 or 9 lbs - and whether I'm the only one to have had this sort of issue with filler-caps ? On paper the 4lb on is good to 107°C, the 9lb one to 115°C.



The header-tank is a good-quality brass one and the seating looks perfect.



Is my interpretation valid, or am I missing something ?

I was intending having the engine it bits this winter anyway, so if I've done any damage it will be fun to find out exactly what I've done.

Perhaps one of the problems in having such an apparently under-stressed and powerful engine is that you can have issues that are not immediately obvious, or don't appear significant, when you're cruising on just a whisker of throttle ? I'll listen more carefully next time .............

PS - Anyone for the Kettle and Fruit-Thermometer Method of gauge-checking ?

[ETTony]

Bizar, yet the header tank must have remained nearly full as you had water temp readings throughout.

I am interested in your catch bottle and the coolant being sucked back into the system as it cools, I had been pondering on adding the S2 reservoir system, but will now follow your idea

[christopher storey]

I think it extremely unlikely that you have done any harm. In the first place the cooling system holds from memory 32 pints ( truly enormous by today's standards ) so your loss is no more than 25% of full capacity. Secondly, if an engine starts to overheat to a damaging extent from coolant shortage, under load the first thing that happens is that you get vociferous pinking. Thirdly, I don't know what altitude your climb involved, but at 30+ ambient temperature, if your altitude was say 1800 feet above sea level, your density altitude was a good deal higher than that, at least 4000 feet and if it was 35 degs it would be about 4500 feet. Virtually all normally aspirated engines start to lose power from 3000 feet density altitude upwards, and this loss is increased if there is no mixture weakening device, which I have never seen on a car. I cannot remember what the partial pressure of oxygen is at density altitude of 4500 feet but my guess is that it would represent a power loss of the order of 15 to 20 per cent

[Duckham]

I think the change to the higher pressure cap required the non-convoluted top hose, Rory, so it will be worth changing that if you stick with the working cap. Ridiculous they can't get the caps correct though.

Bizar, yet the header tank must have remained nearly full as you had water temp readings throughout.

At least on my slightly later 3.8 the sensor is in the path of the water flowing into the tank, so I think that even if the tank itself is empty it will still read correctly. Right up until the water stops flowing of course !

Joe

[Heuer]

My OTS has been subjected to all sorts of trials (including 40c+ in Biarritz) over 50,000 miles and I have never had a heating/coolant problem or seen the need to adopt the S2 system. It does have a Cambridge Motor Sport alloy rad/header, 7lbs cap and Kenlowe fan but is has been like that since 1999. Ditto my 3.8 FHC which has sat in a hot summer traffic jam with a replica Marston radiator, 4lb cap and standard fan/motor and did not overheat or lost coolant.

I wonder Rory whether your addition of the overflow bottle contributed to your problems without adding anything to the running of the car? Personally I would go back to stock or migrate across to the full S2 system rather than a half way house which can obfuscate the real issue. The standard system allows you to overfill and then self regulates by burping out the excess. Once the car is happy no more coolant should be ejected and a quick 'finger check' into the header every morning should confirm nothing is amiss.

I reckon you have a leak rather than coolant being ejected which means hoses or radiator. If the head gasket has gone the coolant in the header would be discoloured.

If the XK engine overheats the first causality is the head which, being alloy, warps. That could of course cause the gasket to fail but the coolant condition will tell you that. So I suggest you dump the baby bottle, top up the coolant and go for a drive keeping an eagle eye on your gauges. Get back to the garage, leave the engine running and check carefully for any hose leaks (hose at bottom of rad is favourite because it is a step down hose) or holed rad. Switch off and car should burp any excess coolant - it should not loose the excess as long as the engine is running. Check the level in the header. If you can't spot anything chuck a couple of bottles of Barr's Leaks in and see what happens.

[Duckham]

David - I agree with the principle, but surely the key here is that 2 of the caps are opening at low or zero pressure. That will cause a problem, although there may still be a separate leak of course.

I don't see why the burp tank should be a issue, although personally I would remove it until the header tank had reached it's natural level and then put it in place. That way the system is working as intended but it is easy to see any abnormal expansion occurring.

Joe

[Heuer]

Rory has not said whether he tested the car with fresh coolant, no bottle and leave it to idle until it reaches normal temp. If the cap was faulty water should start pouring out very quickly so no need to use a pressure test other than out of curiosity. Repeat until you find a cap that works. Now I agree it is of interest that the calibration of some caps is off but not sure how you can tackle that with a vendor. Also he may find that real world operating conditions i.e. hot coolant may affect the release pressure of the cap.

Never seen it myself so it has never come up on my radar. With the convoluted hose Rory needs to use a cap of no more than 4psi

[MarekH]

Is one reason why you have lost more coolant than you expected to that it always vented when you were at altitude? In this case, the spring holding may have been rated at 4psi, reflecting a boiling point at sea level of X'c, but once venting open to atmosphere, the water will have been relatively superheated as the corresponding boiling point will have been X-delta 'C.

Looked at another way, a 4psi cap opening against pressure at sea level will be like a lower psi cap opening against rarified air, i.e. it opens more easily.

I expect the effect may be quite small, but it might add up.

The other possibility is that you were venting water, not steam. If you are pointing steeply uphill (or downhill, depending on whichever way round it has to be to be abnormal) when overheating, some of the air/steam at the top of the header tank may have moved around towards the other end of the engine so when the cap vented, water rather than steam, was getting out.

Last time at Bicester, AndrewH (on this list) turned up and there was steam coming out of the Cheney clamps holding thre big hoses at the bottom of the engine. The catch tank may not have been the only place that the steam escaped to.

kind regards
Marek

[Moeregaard]

I'm with David on this one. This doesn't sound like a blown head gasket, otherwise you would see the aforementioned coolant discolouration, and possibly contaminated oil in the sump. Since the car isn't "marking its territory" on the pavement, I would suspect a pinhole in the radiator core or a leaking core plug--either of which will evaporate before reaching the ground. Bars Leaks is a excellent way to deal with either of these until a more permanent solution is available. Have you checked all of your heater hoses, which tend to be "out of mind" during the summer months? I'm just trying to toss all possibilities out there.

Regarding radiator caps, I always ran a 7-lb. cap on my early 4.2 cars with no problems.

[Duckham]

Thinking a little more about this, if the rubber seal and tank seat are fine then for a cap to show essentially zero pressure means it must have almost no spring resistance. Does it feel that way?
If it feels OK could your tank have too large a gap from top to bottom seats preventing the bottom seal from experiencing enough spring pressure, hence what appear to be multiple cap failures.

I can understand why others suspect a leak somewhere, but I would have expected even a pinhole to be visible when you pressurised to 7psi if the system was cold. Hot makes it more difficult to spot.

Joe

[PeterCrespin]

I was just about to write the same thing Joe. Zero psi would feel impossibly loose and floppy when tested with one's thumbs pressing the cap seal spring. The reason for lacknof system pressurisation probably therefore lies elsewhere.

There are two common neck designs* which require what I think of as 'long reach' and 'short reach' caps to sit correctly with their lower rubber washer against the lower sealing flange. It is easy enough to do a rough check by ensuring the cap needs to be pressed down a little to engage the cap tangs in the upper flange cut-outs before twisting tight to effect good upper seal clamping with the second unsprung rubber washer against the upper neck flange. A short cap in a deep neck may never even contact the inner seal and a long cap in a short neck may bunch up and seem to tighten early. In between lie a range of situations caused by wear or the deterioration of fixed or moving parts.

I like brass components in cooling systems but it has to be admitted that after years of use the the upper flange can wear or distort so that it gets pulled up around the rim where the ramp tightening occurs. This can be repaired by careful dressing flat with a small hammer and should be felt as firmer final tightening. Alternatively, with old parts, the ramp edge itself gets worn away (like a Chef's favourite knife after years of re-sharpening eh Rory?) This results in absent or only marginal sealing until the last bit of turning when the tangs abut the less-worn final section of ramp.

I expect Rory's tank is not clapped out, so there is perhaps a cap/neck mismatch of some sort?

Pete

*I am not 100% certain of how/when or even 'if' there are still two distinct depths of neck design. It may have been in the early days of switching from various manufacturer-specific screw-down or clamped caps before de facto standardisation, but I have definitely encountered cap that did not reach the lower sealing flange. That scenario would give zero system pressure.

[ralphr1780]

I have experienced the same what Pete is describing well: the new cap from one of the usuals had a shorter neck than needed, I was losing coolant on every ride without any visible leak and no headgasket issue. Went to a spares shop and bought a longer neck cap rated 7 psi, problem solved for good.

[rfs1957]

Great to have contributions from all the usual suspects, at last no sign of sowing on stony ground !

Currently preoccupied with traversing France on the very same roads and hills as with the Jag, but this time pulled not pushed, and by an A series 1275 producing 95hp, enjoying the double takes as the Minivan sails past at 85 mph, uphill.

Nice to drive something with the lambda rock solid in the green, no pinking, and a water gauge nailed to 90 ....... makes a change !

Will ponder the various issues and advices,and explore leaks in 10 days time, believe the caps are correct in that spring pressure is present upon fitting, and no sign of any steam-blown traces anywhere around the engine, but will do some forensics and report back.

[Ashcloud]

Hi, I just purchased an S3 OTS and picked it up. I started a comfortable drive along flat level country roads.
At the first gas stop I saw coolant running out from under the car. Some investigating showed it only came from the overflow hose and stopped after 30 seconds some burbling could be heard, no steam. I refilled and continued my journey watching the temperature very close. It stayed perfectly in the center of the gage after a little warm up time. A minute after each successive stop to check the coolant it blew water out again. After the third stop I had enough and called the "rescue" truck To pick up the car and he drove us home.

After investigating:
1) the system did hold 1bar/15psi without leaking
2) the 14psi cap barely touched the seat
3) the fans never came on
4) the right fan motor is dead
5) the sensor and tempgauge appear to indicate correctly (never overheated)
6) later on I found 2 instances of coolant under the cold engine but it still tests good at 15psi ???

Finally got a 9psi cap but with a longer spring and sits firm now.

When I ran the engine on idle for a while no overflow yet, no burbling (yet).
Checked the temp an different places:
1) prior to the thermostat left side 85C, right side 105C
2) after the thermostats on either side about 65C
3) radiator 55C

Just purchased new thermostats and both will go in next week.

Btw.: here I cannot get Bluecol "U", what would be a good equivalent coolant for the car?

Greetings Bernhard

[Sarthe72]

Just caught up with this thread. In 2012 the club toured Europe, and on the way back we did quite a few passes in the Alps before stopping for the night. My temperature gauge showed no overheating at all, so we did more of the same the next day and I checked the coolant level that evening after dinner. I put a whole five litre bottle in and still there was space for more! In 2012, the car being somewhat doggy after 25 hard years, I had an economy rebuild done, including an engine rebuild by VHE. This what they found:-

3. More cracks by Tony Brown, on Flickr
One cylinder head fit only for scrap, and:-

tony brown 11 by Tony Brown, on Flickr

One cracked block. Now the funny part was that apart from timing chain thrash the car was still running well except for using a pint of oil per 1,000 miles, something she had never done in 80,000 hard miles, and well, I was always topping the water up so hadn't really noticed any difference until that time in the Alps. No mayonnaise in the oil either. I have no idea as to whether your engine has a problem, but I offer this history as a possible warning. And I was using a 7lb cap by the way

[Ashcloud]

Thanks for those informative pictures. I just wanted to exchange those thermostats today and as careful as I am I still broke 3 out of 6 bolts while removing the thermostat housing. I even used an inductive heater to loosen those bolts which worked for one of them but the rest was rusted more or less away.
Now I am sitting here and I am thinking about a complete engine revision before more goes wrong. I believe that in the long run things will get worse. If bolts are rusty there is probably more to come when removing those heads.
I think if I get the required special tools I will take the engine out and do the job this winter. At least at the end I know my engine and have the peace of mind that I need in order to enjoy a trouble free engine.
Greetings Bernhard

[rfs1957]

Blimey, where did three years go ?

My engine had picked up on one of the bores, but impossible to say when or why.



I'm now about to run my new engine on an engine test-bed, out of the car, but am still a bit puzzled by the same radiator-cap questions as were raised here three years ago.

I blew compressed air into the cooling system today to check for any major leaks (it helps to have not forgotten the four 5/16 UNF blanking plugs in the water jacket, under the exhaust manifolds) and whilst the radiator cap blows off at approximately the correct pressure, it doesn't re-seal completely once the pressure drops below the blow-off, and continues hissing until the pressure is almost down to zero.

Now, I know air isn't water, and even less so when everything's cold, but I don't think it ought to do this.

It got me to wondering again about my tank neck, and sizes.





From what I can see (and guess) the Jag and the Mini have conventional "short" necks, which is to say there is a 3/4" / 19.5mm difference between the upper and lower seating surfaces.

I suspect this is as per the SAE J164 specs hidden behind an $83 pay-wall here

https://www.sae.org/standards/content/j164_199702/

and confirmation appears to be provided here

https://www.northernradiator.com/knowledge/radiatorcaps

where there is a familiar range of 2.2" caps with a 3/4" nominal depth, which I think are ours.



The Made In England caps (the 9 and the 15 seen here) I have are both of similar dimensions, which is to say that they are barely clamped at all in the first position, just after the bump, and the pre-load therefore appears to be afforded solely by the ramp on the track, which I measure at 2.5mm.



The Fah-ler shiny cap (7 lbs) that I was going to use is actually about 3mm longer, so would have about a 5.5mm effective pre-load once clamped.



The true loading of course is a function of both the pre-load AND the spring rate.

So what's up for debate is a) how different is air to water ? b) are these neck and cap depths what everyone elses has on their cars ? and c) does my tank neck sealing surface look ok ?

[PeterCrespin]

I think your tank neck looks OK but you could always check the upper edge is flat and dress it down a bit if it has high spots. I'll leave you to explain the irrelevance or otherwise of 'pre-load' and/or 'ride height' on the further movement of a linear rate spring.

NB, the rubber gaskets may(?) seal more effectively when tested hot, due to increased plasticity conforming better to surface imperfections? On second thoughts, that probably has zero real effect for a cap in good condition,,,

Less Mt. Ventoux, more Promenade des Anglais for you matey!

[rfs1957]

The Forum is a very good measure of senility, a sort of "temperature gauge" for the neurons.

Unfortunately I appear to be off-scale (although, with a nod to Peter Crespin - if he's still around - I can do the Ventoux)



Anyway, digression aside, it appears I'd already taken and published the pictures of caps, sealing ramps, header tanks etc in 2017 and 2020, so the same ones I've just taken are irrelevant.

Anyway, here's the question.

Just HOW full is FULL ?



I had the filled the coolant level to 12mm below the true sealing flange (the lower one) before a 250km jaunt recently, with stable temperature readings, even in half an hour of stationary/crawling traffic in over 30°C, yet when I got back the tank required 300ml of coolant to get back to the same 12mm level.

Further usage, probably another 100km, in cooler conditions but with similar gauge readings, but this time with a baby bottle on the overflow hose, has revealed a couple of table-spoons of coolant ejected.



Now, I'm not 100% confident about the 9lb cap I have, and am about to fit a replacement, but wondered whether I'm also over-filling.

Jag referred to topping up "to the bottom of the filler neck" whereas others seem to regard that feeling the coolant above the baffle plate is enough.



What are others' experiences please ? Does it make any difference ?

I might yet ask you again, as in a few weeks' time I'll have forgotten the answer .............

PS - and will anyone be selling Bar's Leak at Goodwood

[david muir]

I have experimented with using a pressure catch tank i.e. plain cap on radiator and pressure cap on catch tank. I had used this to good effect on a tuned Mini ( probably suggested by Vizard). Benefit , on both vehicles, was that one could just glance at translucent catch tank and see what level was (knowing that header tank would be full to neck). Worked great on Mini but I removed it from E reckoning that I was being 'over the top' re coolant loss.