Expansion Tank Pressure when cold.

[The Great Pretender]

I can't see why level would drop between -ve pressure and open to atmosphere.

Didn't the level rise when open to atmosphere switching the LCA off?

[Driver+Passengers]

I can't see why level would drop between -ve pressure and open to atmosphere.

Didn't the level rise when open to atmosphere switching the LCA off?

Yes, got my words twisted - I meant rise. Still that, suggests +ve pressure, not -ve pressure in the header tank. I'll edit.

[The Great Pretender]

I can't see why level would drop between -ve pressure and open to atmosphere.

Didn't the level rise when open to atmosphere switching the LCA off?

Yes, got my words twisted - I meant rise. Still that, suggests +ve pressure, not -ve pressure in the header tank. I'll edit.

No problem, we all make mistakes, especially at this time of night.
Are you saying positive pressure was holding the level down?

[Driver+Passengers]

No problem, we all make mistakes, especially at this time of night.
Are you saying positive pressure was holding the level down?

That's what I have assumed - but I can only get to level dropping with +ve pressure if either hoses were swelling or air pocket was compressing, but don't imagine cold hoses would be significantly swollen at low pressures. I could be wrong about many things.

[The Great Pretender]

No problem, we all make mistakes, especially at this time of night.
Are you saying positive pressure was holding the level down?

That's what I have assumed - but I can only get to level dropping with +ve pressure if either hoses were swelling or air pocket was compressing, but don't imagine cold hoses would be significantly swollen at low pressures. I could be wrong about many things.

I know you have done significant computer tests to solve your problem, and a brilliant job you did. I tend to get down and dirty in the workshop and use instruments and measurements for information. I have a large Boss 2 bar gauge linked to my coolant system that shows the system pressure, depending on use it will hold pressure (dropping over time) for four to six hours but never overnight.
Air pockets would only shrink (compress) as temperature reduced lowering the level in the tank.

[helen&tony]

Hi
Matt...
Water doesn't resist compression, as there's ALMOST always dissolved gases in it...hence it doesn't make good brake fluid... ...just being picky

Ah...strangely enough, I've been watching the "annual drop" in water level as it gets cold over here...a normal cooling system a few weeks ago in the summer, demonstrates quite a drop when the cold starts in here...as much as a 1/4 inch in a day or so, before it settles down for the winter....remove the header tank cap, and there's a fair hiss. Is there a drop when you take the cap off?...well...no, that's why a radiator cap has 2 springs and seatings in it...one large one to hold the positive pressure when the coolant heats up and expands, and the other to cope with the contraction in the water level when it all cools down....the theory behind the small spring and seat is that it is there to cope with contraction which MAY...repeat MAY cause collapse of radiator and heater matrix passages....Having totally dissected the Bongo heater matrix, I expect it would need a lot of vacuum to do that, as they are comparatively large tubes and walls comparing with the old thin brass / copper tubes of years ago. Anyway, any hiss when taking the cap off is just residual minor pressure difference between the closed system and atmosphere. In a normally functioning system there shouldn't be any loss, and what variation in level you do see is just expansion and contraction with the seasons.
We see a lot here, as the difference between summer and winter temperatures is 60 to 70 degrees Celcius....and because of my forgettory (ne memory) , I don't always "click" that the sudden loss is due to contraction, and I buzz round for leak patrol and get all in a fluster....
Cheers
Helen

[Northern Bongolow]

I have a large Boss 2 bar gauge linked to my coolant system that shows the system pressure, depending on use it will hold pressure (dropping over time) for four to six hours but never overnight.
Air pockets would only shrink (compress) as temperature reduced lowering the level in the tank.[/color]

spot on mel, and helen.

during this time of year everyone will experience a very slight drop in cold level noted, this i call the seasonal adjustment. this must be due to lower night temps.
the time noted by mel on cool down is also noted by myself with a similar set up on my bongo, helen will note a larger coolant level drop as she says her season temp adjustments are larger.

as the coolant cools it pulls a neg pressure in the system, this is controlled via the cap inlet valve, it rests at -1 pound (the spring value), so when the cap is removed a hiss is noted, this is air going in to relieve the -1 to neg pressure. on a really cold night after a run you can even hear the cap squeak as it lets the system pressure drop as it cools down. as the system cools the coolant level rises slightly.

[tallbongo]

There's a lot of good info here, but also some stuff in this thread that could lead to confusion, so I'd like to chip in my tuppence worth to hopefully clarify some of it.

First some facts on gases and liquids. Both are fluids, but gases are generally treated as compressible and liquids not and pretty much everything I can think of resists compression (or they'd collapse into nothing under atmospheric pressure). However, air is much, much more compressible than water. If I recall correctly, I think the figure is about 20,000 times more compressible. It is therefore a fair assumption that in a system containing air and water, any compression of water is negligible compared to that of air.

However, as indicated by the Great Pretender, the density of the liquid is different to it's compressibility. 13 litres of 50% water/50 % propylene glycol mix put into the system at a temp of 30 deg C will only be 12.75 litres at -10 deg C. The compressibility of the air makes up for these differences, resulting in a change of pressure inside a closed cooling system.
The bongo cooling system contains about 13 litres of liquid, and at an estimate, about 2 litres of air in the expansion tank.

The radiator cap opens when the pressure inside the coolant system is 1.1 bar greater than the pressure outside, i.e. atmospheric pressure. So if you fill your system on a very high pressure day (say atmos pressure of 1.04 bar), put the cap on and, without having driven, open the cap on a day when the temp is about the same, but the atmos pressure is very low (say 0.97 bar), then your system will vent pressure when you open it because the system has 0.07 bar more pressure than the atmosphere. If you do it the other way around it'll take in pressure when you open the cap. Similarly, if you open the cap when the temperature is higher than when you filled it there will be positive pressure, and negative pressure if the temperature is cooler.

Of course we all drive our Bongos resulting in temp changes in the coolant. Say we filled the system with 13 litres on a day when the temp was 20 deg C. Assume atmospheric pressure to be a constant 1 bar and that the cooling system has a constant capacity of 15 litres (i.e. no expansion of pipes and 2 litres of air in the expansion tank after filling). Pressure x Volume / temp (in Kelvin) is constant for an ideal gas, and air is close enough to an ideal gas. Then as the engine temp heated up the following would happen in a sealed system.


(These figure correlate well with the 10-13psi that Matt quotes for a healthy system, corresponding to a running temp range of about 88 deg C +/- 5deg, so I guess that the 2L of air estimate is quite accurate.)

When the pressure reaches 2.1bar (i.e. 1.1 bar above atmospheric pressure) the vent cap opens and excess air is slowly expelled as the engine heats up maintaining a pressure differential of 1.1 bar. When the coolant temp starts dropping this vent cap closes and the system is again sealed. As the coolant cools to ambient temperature the pressure in the system drops.

The temperature and amount of coolant you use when filling dictates this curve. The more coolant you put in, the more extreme the effect, as the volume changes by more with temperature. I calculate that the vent cap will open at 15 degrees lower temp if you overfill by only 0.5 a litre (i.e. 13.5 L coolant, 1.5 L air). Of course the difference in atmospheric pressure between the time the vent closes and when you next open the expansion tank cap could also result in the effects described a couple of paragraphs back.

It's often said on here to fill right up to the max line. I don't believe that this is a good general rule. The max and min lines are to allow some user error and to allow expansion and contraction of the coolant with ambient temperature. If you do have a leak, overfilling won't buy you significantly more time, as you will be pressurising that weak spot more than had you filled to the min line, potentially accelerating the failure.

Any contraction of pipes etc during cooling will result in a positive pressure, possibly keeping the pipes partially swollen.

So there are many reasons for both an intake or expulsion of air when removing the expansion tank cap. There is a limit to the upper pressure (1.1 bar more than atmosphere), and from the info on this thread it appears there is also a negative pressure limit (I'll have to check my expansion tank cap because I thought there was just the one spring on mine). The upper limit should only be reached when the engine is working very hard, and a lower limit of -1psi only reached during significant changes in environmental conditions from the last time the vent opened.

However, in any situation, the coolant level should not change by solely by opening the expansion tank cap on a cold system. If it does it means that something else is changing volume, not the coolant.

One reason could be very dodgy rubber pipes - they could have expanded when pressurised and, when the cap is removed, contracted. Any change in their volume will directly correlate with the expansion tank level.

Another reason could be air trapped somewhere in the cooling loop. In this case the change in level will be proportional to the ratio of trapped air to the volume of air in the tank. The system needs urgently bled if this is the case.

Exhaust gases could cause the pressurisation while running, but should be vented. What they could easily do though is add a compressible gas trapped inside the coolant loop as above.

So basically follow Simon Jones' advice and if that doesn't show anything you may have a problem similar to Matt (Driver and Passengers).



(Regarding some of the other stuff mentioned here:
1) Water boils and forms gases at a relatively low temperature. Gases are very compressible - not good under hard braking just when temps rise. I like my brakes to work below zero degrees C, when water becomes solid. Other liquids, as well as being more suitable for these reasons, are also more easily sealed. These are more important reasons why water is not used as a brake fluid.

2) If water were suddenly to become far more compressible, then ships would experience problems. However, we could still design them such that they would float. Helium balloons float in air. Air is compressible. )

[helen&tony]

Hi.
My remark regarding brake fluid was a bit of a jest...but as I said, water doesn't resist compression...and it can exist in 3 of the 4 states of matter
Cheers
Helen

[tallbongo]

Hi Helen,

The important point is that water is much, much, much, much harder to compress than air. I don't believe that you are arguing that water will compress in the Bongo's cooling system, but some may interpret it as that.

Can we agree that if the coolant level changes simply by opening the expansion tank cap, then that coolant system is not healthy?

You and others are absolutely correct about the annual drop and hissing from negative pressure. If the OP is accurately reporting the situation, then that isn't the whole story here. Removing the cap should never make this "winter level" rise and switch off the alarm.

David.

[The Great Pretender]

Hi David, some interesting information there. I’m sure Helen will have a response to your last post as she can type quicker than I can think………….

With regard to the graph from your first post it appears to show that the pressure cap will open at 100 C

[helen&tony]

Hi
David...
Not arguing the point at all...I'm often poking around at the cooling system, improving and dis-improving it ...never seen it drop on opening the cap...that would mean that water is reversibly extensible, (or compressible ), and, thus, elastic, and no darned good in a pressurised system at all....BUT, at high levels of compression, I believe that it may???? be possible to compress in a cylinder while fluid , but that may be at a transition phase between liquid and solid....and bearing in mind that pressure in a cylinder is as a result of , say, external influence, pressure would increase the temperature , so how best to calculate the transition point under increasing pressure would be a bit beyond me in this season , weather-wise!...
That last bit about the season, is my little bit of personal humour, as at this time of year, not only do we see the coolant drop in the Bongo, but it's "stroke season", where there are quite a few people popping off due to sudden pressure changes, as strokes. I've had a couple me'sel, and this year I've had more problems....still trying to decrease blood pressure, and pretty well unable to do much.....I mention this, as it's rather interesting (to me) the co-relation of the human body's control systems in comparison to that of the average car....development in control systems in machines have nearly caught up with the inbuilt redundancy of our bodily systems....FASCINATING....all I need is a set of 4 wheels!
Mel...
You have no idea how long it takes me to write anything, and thinking is pretty much out of the window this week.
To write anything on a keyboard , I constantly correct as I go, then I find I have written a whole lot of words as one, and halfway through, I go back, check, and correct....PLUS, I check twice at the end of my effort, and hope that the editing is complete....As for speech...that's affected by my strokes in a minor way...which I find rather amusing, as Tony's often asking what I'm trying to say....and I have a stock reply...."Why are you asking me what I'm trying to say.....how the hell do I know?"...
Cheers
Helen

[tallbongo]

Helen,

I hope your health improves soon. Glad to see we're in agreement too!

Hi David, some interesting information there. I’m sure Helen will have a response to your last post as she can type quicker than I can think………….

With regard to the graph from your first post it appears to show that the pressure cap will open at 100 C

Yes, the graph shows that, but it's just to illustrate the principle. There are far too many assumptions to take it as being correct.

1) I estimated the air volume in the header tank poorly. There's significantly less, which would mean the vent should be at a lower temp.
2) I used an extremely simplified model, where the coolant wasn't really cooling! The coolant will be at different temps at different parts of the coolant loop. In general though it will be cooler than the engine. This increases the engine temp at which venting will occur.

The errors in 1) and 2) have pretty much cancelled each other out at around 90 deg, but there is no guarantee that is the case at other temps.

There are many more factors not allowed for in the model. Mix and expansion properties of coolant, fill level, thermostat, speed of travel, ambient temp and pressure, wind, radiator condition and probably loads of other factors will affect the pressure in the expansion tank.

The only way to really know what is correct would be through extensive data logging, but there are a lot of factors to account for.