Cooling system confusion

[NigelB]

The M50 325i engine runs with a coolant pressure of 2 bar which I guess is 2 x atmospheric pressure or 30psi. On that basis I can buy radiator caps from a number of sources rated in bar.

If I want to buy a radiator cap rated in psi I can buy anything from 0 psi up to 20 psi, but certainly haven't found anything at 30 psi.

But what is a 0 psi radiator cap. Does it mean 0 psi above the normal 15psi atmospheric pressure (at sea level) ie a 1 bar cap and on that basis can I get a 15psi cap to give me the 2 bar that the M50 operates at (in production vehicles). If that's not what it means, why would you ever need a pressure cap rated below 15psi.

Am I missing something. Should I be sat in the corner with a dunces cap. (No answer required!!) Or is it just confusing.

Nigel

[mashtun]

Could be the blind leading the blind here but I won't let that stop me It's an interesting subject and one I'll have to address myself soon enough.

Originally Posted by NigelB

If I want to buy a radiator cap rated in psi I can buy anything from 0 psi up to 20 psi, but certainly haven't found anything at 30 psi.

But what is a 0 psi radiator cap. Does it mean 0 psi above the normal 15psi atmospheric pressure (at sea level) ie a 1 bar cap and on that basis can I get a 15psi cap to give me the 2 bar that the M50 operates at (in production vehicles). If that's not what it means, why would you ever need a pressure cap rated below 15psi.

The idea of pressurizing the system is to raise the boiling point of the coolant, so I guess a 0 psi cap doesn't raise the boiling point above what it would be in an open vessel? Presumably the characteristics of a system that uses a 0 psi cap are such that the coolant doesn't reach its atmospheric pressure boiling point under normal running conditions? Or that air rather than coolant is lost (and replaced on cooling) if it does? Sounds inefficient but possibly normal practice in the past?

Originally Posted by NigelB

Am I missing something. Should I be sat in the corner with a dunces cap. (No answer required!!) Or is it just confusing.

It's confusing alright - I see Bentley quotes these figures:

• Radiator test pressure ......... 1.5 bar (21.75 psi)
• Radiator cap test pressure .......... 2 bar (29 psi)

Does that mean the radiator is the weak point in the system? Sounds unlikely that you'd let a radiator fail rather than a cap vent... I think what it does suggest is that the pressure in the system under normal running conditions is less than 1.5 bar.

The other thing to bear in mind is that the water:antifreeze ratio also affects the boiling point of the coolant so I suppose that changing the ratio could be played against the pressure rating of the system (not sure to what extent though).

More questions than answers from me, I'm afraid - sorry!

Mike's M50 engine's in and running - I wonder what he's used? Is morris at that stage with his M50 yet?

Mark.

[morris]

I've been taking all ratings for caps as relative to atmospheric pressure. I've therefore also been assuming that a 0psi cap is simply just a cap and in no way inhibits the flow past the neck to the overflow.

I've pretty much come to the decision of fitting a 1.4bar cap on my header, a 0 to <1.4 on my rad and then a catch tank next to the header which will suck back any over spill of coolant as the system cools down.

I'm working on the prinicple that BMW built the engine to be driven by Canadians in winter and Australians in summer and therefore they need the coolant system to work at both extremes hence the higher pressure for a higher boiling point in hotter climates along with 50/50 water/antifreeze mix and that weird throttle body warming system for cold climates.

As I only need my engine to run between 10 and 30 degrees I think the lowering of the pressure cap isn't going to cause an issue. All this is of course based on a possibly misplaced "he who dares Rodney, he who dares" attitude rather than any real experience in this kind of thing

[Mike]

My simplistic approach was: Copy someone else!

Chris Cunliife has built his own M50 Cabrio, and he fitted a VW Golf header tank, and his works fine. I chose the same, and hope for the same!?!

Truth is, I do not know what pressure the cap is on the VW header - I must say I have often wondered what pressure it is, as it is plastic, not the conventional steel cap seen on most radiators.

I do know it is pressurised, as my rad boiled when I first started my engine up, as the cooling system was full of air, and did not have enough water in it. That and the fact that I had not wired my thermostatic fan up at that point. Since wiring in my 13" Pacet fan - the biggest capacity fan I could get, capable of moving 2,000 cu.ft per minute - it is pretty quick at pulling a lot of heat away from the radiator.

Mike

[NigelB]

Originally Posted by morris

I've been taking all ratings for caps as relative to atmospheric pressure. I've therefore also been assuming that a 0psi cap is simply just a cap and in no way inhibits the flow past the neck to the overflow.

I've pretty much come to the decision of fitting a 1.4bar cap on my header, a 0 to <1.4 on my rad and then a catch tank next to the header which will suck back any over spill of coolant as the system cools down.

Hi Ian,

And there in lies my conundrum!!

Practically I agree with you and I am going down the same route. And I recognise that BMWs were built to run in all climates so in the Middle East for example, the higher the cooling system pressure, the higher the boiling point of the coolant and the less chance of the engine overheating.

But from a theoretical, and an understanding perspective, and taking atmospheric pressure as the reference point, if you use a 1.4 bar cap on your header, do you think the cap will release when the cooling system reaches 1.4 times atmospheric pressure or 2.4 times atmospheric pressure.

And if you look on the SVC website they have the same header tank as you have used and they offer caps to go with it ranging from 0psi to 20psi. If that figure is relative to atmospheric (and I assume it must be 'cus why would you want a pressure cap releasing below atmospheric pressure ie. a partial vacuum) then that will give a cooling system pressure of 15psi (atmospheric) + 20psi (the highest rated cap) which is equivalent to 35psi. Now as was said previously, 35psi is well into tyre pressure territory. Could a cooling system really run at that sort of pressure...........................??

So I remain confused. I know what I am going to do practically but what do the numbers actually mean?? Is it simply a terminology thing where, when using 'bar' the quoted figure is a multiple of atmospheric, but when using psi, atmospheric pressure is used as a reference and the psi rating is relative to atmospheric.

Nigel

[peterux]

I am no expert but I think the rated psi or bar is above atmospheric pressure. So if the pressure inside the cooling system exceeds the rating of the cap e.g. 20 psi then air and coolant is ejected.
The other factor to consider is the volume of coolant versus the volume of air in the header tank.
I presume what happens is that the pressure increases because the coolant expands as it gets hot from the engine. The pressure builds up because the header tank air is compressed by the closed system. (you can't compress the coolant).
The coolant stops expanding when the thermostat opens and the radiator (and fan when stationary) cools down the coolant. Providing the coolant can be maintained at a constant temperature it stops expanding and the pressure stabilizes.
Don't forget the radiator cap only releases under a fault condition when the pressure exceeds the rating and then it dumps your expensive anti-freeze all over the road.
I chose a header tank that was similar in size to the original BMW and I'm running 1.4 bar caps on both my radiator and header tank.
(BTW, I have no idea what 0 psi caps are)

...peter

[Mike]

Originally Posted by peterux

(BTW, I have no idea what 0 psi caps are)

...peter

A filler cap.........? With no pressure release.
There will be another cap with pressure release in the system

[NigelB]

Originally Posted by Mike

A filler cap.........? With no pressure release.

But that could be a 15psi cap......................!! Let me try to explain my thinking.

Imagine a flat car tyre on its rim. There's atmospheric pressure of 15psi acting on it from the outside. But the tyre is not crushed onto the rim by that pressure because there is 15psi inside the tyre as well.

Now imagine that flat tyre is the engine cooling system. There's 15psi atmospheric presure on the outside and there must be 15psi inside the system as well othwise the hoses would start to crush. Therefore, as the coolant starts to heat up and the internal pressure raises by 1 psi the pressure inside the system is now 16 psi.

So if (for example) I need my pressure cap to release at 5psi above atmospheric, do I need a 5psi cap or a 20psi cap. Similarily, if I need a cap rated in Bar should I use a 0.33 bar or a 1.33 bar cap. But I don't think I have ever seen a cap rated at less than 1 bar.!!

But it's all academic really. I know what I am going to do. I just seems to me that the terminology in a little unclear.

Nigel

[Mike]

Did you ever watch The Young Ones , as Nigel would say slowly " H-e-a-v-y!"


A 3psi cap is + 3psi over normal atmospheric pressure, and is a relative measure.

By comparison Bar is an absolute. So a 1 Bar cap is = normal atmospheric pressure (as in 1000 millibars is normal )

Originally Posted by NigelB

But that could be a 15psi cap......................!! Let me try to explain my thinking.

Imagine a flat car tyre on its rim. There's atmospheric pressure of 15psi acting on it from the outside. But the tyre is not crushed onto the rim by that pressure because there is 15psi inside the tyre as well.

Now imagine that flat tyre is the engine cooling system. There's 15psi atmospheric presure on the outside and there must be 15psi inside the system as well othwise the hoses would start to crush. Therefore, as the coolant starts to heat up and the internal pressure raises by 1 psi the pressure inside the system is now 16 psi.

So if (for example) I need my pressure cap to release at 5psi above atmospheric, do I need a 5psi cap or a 20psi cap. Similarily, if I need a cap rated in Bar should I use a 0.33 bar or a 1.33 bar cap. But I don't think I have ever seen a cap rated at less than 1 bar.!!

But it's all academic really. I know what I am going to do. I just seems to me that the terminology in a little unclear.

Nigel

[Chris Cussen]

I think the values are relative to atmospheric so a 14.7 psi cap will operate at 1 bar.

The 0 psi caps I have seen are really infinity psi caps and seal the opening at the top of the filler neck (unless there is an overflow pipe in which case they are 0 psi) if you look at this item http://www.ebay.co.uk/itm/Stainless-...item2eb6535914 you will see that it has a seal in the top, but no spring loaded seal that usually goes against the bottom of the neck.



I have a filler neck in my highest hose near the radiator using a 0 psi cap as there is no overflow, and my system venting actually occurs from the filler cap in the expansion tank.

[NigelB]

Originally Posted by Mike

Did you ever watch The Young Ones , as Nigel would say slowly " H-e-a-v-y!"


A 3psi cap is + 3psi over normal atmospheric pressure, and is a relative measure.

By comparison Bar is an absolute. So a 1 Bar cap is = normal atmospheric pressure (as in 1000 millibars is normal )

Thanks Mike.

A bit of clarity at last. I've just found this and started to read it.

http://en.wikipedia.org/wiki/Pressure_measurement

But then the fog started to descend again so I stopped......................!!

N.

[atlantasportster]

KOYORAD, who build racing radiators, state that their cap is 1.3 bar, 18.9psi, raising the boiling point by 56 degrees F.
So it would seem that the ratings are based on "lbs per sq in absolute", and not "lbs per sq in guage."

[Mike]

Not sure whether to thank you guys or curse you..............

After reading your thoughts on header tanks, and pressure caps etc, I started to think about my own set up a bit more carefully - I had simply copied Chris Cunlliffe's set up, believing his worked so should mine.
Over the weekend I got my electric fan to work off the radiator thermostat, and was quite chuffed ..............until it then emptied most of the coolant on to the floor via the overflow!
All was not well.

Having given it considerable thought I have come to tthree conclusions:

The standard BMW M50 coolant volume is around 10 litres. The relative density change between water which is just above freezing, and boiling is around 4%. Therefore I need to accommodate 0.4 litre of expansion of water.

The second point I have realised is that, in making my radiator fill the nose cone as much as possible, I have seriously limited the headroom left to get my header tank above the radiator top. My current set up has the expansion tank top level with the radiator top. This can only work if I make a catch tank that will then allow any expelled coolant to be sucked back into the expansion tank when it cools down.
Thirdly , since my expansion tank is not pressurised, I can not raise the coolant temperature to the normal BMW M50 levels as intended.

I have therefore decided to iorder a new header tank, with a pressurised cap, and make sure I mount it as high as I possibly can under the bonnet at the bulk head. A standard Wilabb 1.5 litre tank should give me around 0.7 litres above the radiator top, and therefore should just cope with the predicted expansion of the coolant.

So - you have cost me £63, but possibly saved me £100s by preventing me from cooking my engine!!
A grudging thank you to all who contributed!
Mike

[AlanHogg]

Hi Mike
Just read your concerns over the fact that the top of your header tank is level with top of radiator/top hose. Mine is pretty much in same position and to date I have never experienced any cooling problems set up this way[I would struggle to get it any higher in any case] I appreciate I'm running a different engine but i wouldn't have thought your existing install would present concerns.

[Mike]

Originally Posted by AlanHogg

Hi Mike
Just read your concerns over the fact that the top of your header tank is level with top of radiator/top hose. Mine is pretty much in same position and to date I have never experienced any cooling problems set up this way[I would struggle to get it any higher in any case] I appreciate I'm running a different engine but i wouldn't have thought your existing install would present concerns.

Hi Alan
Having looked in to this I can not say I fully understand it!
Water expands when heated - my calculations suggest 4% over the temperature range of our engines. Therefore a system containing 10 litres will expand 0.4 litres. With my set up where the header is level with the radiator, it means it pushes water out the overflow to waste. So when it cools down the water level in the radiator is below full. Eventually my engine then starts to boil.
I could fix this by having a catch tank, but since my header does not appear to be pressurised, I have decided to cut my losses, and change to a pressurised header tank and raise it as far above the radiator as possible to give me more than 0.4 litres above the radiator top so that when it expands, and contracts, the water level should remain higher than the radiator top.

So the theory goes......................

Another one to let you know how I get on with!

PS - have you found an intermittent wiper relay?