When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Cooler, better...looking, radiator caps? Glycol % vs. psi info
These caps seem to fit better, into the filler vs. on top of it, and possibly seals or reduces the leakage of the leaks-when-new Porsche tank I have on my '88. Fits well on the URO Mercedes 380SL tank on my supercharged '89.
Stock is 100 kpa. I ordered a 120 accidentally only to find that I should be using that one because I am running 33% glycol (G05 with 2 bottles of Water Wetter) for theoretically better heat transfer. Less glycol requires a higher pressure cap to maintain the same boiling point. Also, don't freak out like I did if you see what look like oil globules in the tank when using Water Wetter...the globs are most likely the WW. (There is a diesel WW which may not do this.) FYI: a fully drained 928 cooling system takes 3 gallons on the nose.
They are sold by multiple companies but appear to be all made by one. There are no part numbers or company info on any of them just the kpa rating. Below is the list I compiled from Mercedes and BMW part numbers, plus the companies that box them.
Kpa/Psi | part number | Company
100/14.5 | 17111100848 | CRP
URO tank (FYI Roger stocks these) - note position of the cap is more to the rear vs. the stock tank and the overflow tube faces forward. One small hose inlet so you need a 'T' in place of the soon-to-break pressure sensor. No level sensor - can just be left unplugged just like the deleted pressure sensor.
Your concerns about what happens if the coolant is diluted although understandable, are something of a "non issue". The operating pressure in the system is determined y the level in the coolant reservoir. The vessel has a volume of circa 2 litres and is filled to the central seam give or take a little thus the air space volume is circa 1 litre. The cooling system is very well designed and on a hot day with the ac running will typically hit 98C when cruising and a little more if worked hard. At this temperature the coolant inventory will expand by ciurca 0.5 litres and thus as per dear old Mr Boyle the pressure will double from 15 psia to circa 30 psia [15psig] which of course is the design pressure of the system- any higher and the pressure cap would lift and blow off pressure until it reseats itself at about 28 psia [13 psig].
Needless to say diluting the mix improves the heat carrying capacity of the coolant and to have an excursion in pressure of this nature can only happen if and when something drastic has happened like the water pump throwing its impeller - then the coolant boils and the system cannot cope with such a happening so arguably better that it blows off at the stock pressure rather than a higher pressure that could conceivably over stress the system.
In reality fitting a higher pressure rated cap buys you a few seconds more in a non viable situation before the thing blows and nothing more than that - just not worth it IMHO.
On a separate note I got one of those URO tanks from an outfit in San Francisco for about 30 bucks or so- there is no nozzle for a level switch that arguably is not really needed and there was only one nozzle for the small return lines that was also oriented a bit differently but nothing too problemmatical to work around. I pressure tested it at ambient conditions and it seemingly held 15 psig. Would they stand up to normal qworking conditions long term- who knows but at the price well worth a punt. That the US$500 stock items can fail even when new is downright scary!
Carl's aluminium tanks although not completely optimal for level determination, when they were available at around $100 a pop or whatever looked a real bargain to me- scope for a new project for you maybe?
"The operating pressure in the system is determined y the level in the coolant reservoir."
I think you mean you REACH a pressure sooner, just a little sooner..less air volume will reach cap pressure sooner, bleed off coolant if it's too full, or just push and and pull in air..over and over.
"The operating pressure in the system is determined y the level in the coolant reservoir."
I think you mean you REACH a pressure sooner, just a little sooner..less air volume will reach cap pressure sooner, bleed off coolant if it's too full, or just push and and pull in air..over and over.
I mean exactly what I wrote- the steady state system pressure is determined by compression of the air space due to expansion of the coolant as it heats up and the volumetric increment of expansion at normal operating temperature is pretty much 500cm3 as per my calcs and the operating temperature only shows small changes under normal circumstances. Of course if and when the system goes unstable the temperature will increase and on our later models the digi dash will show an alarm- anyone who ignores such alarm and does not shutdown the engine immediately deserves all they may get attrition wise.
If and when the coolant reaches its saturation temperature at the cylinder wall interface it will start to boil and then all bets are off. The sudden increase in volume due to steam generation will blow the pressure cap open- fitting a higher pressure rating cap will only buy a few seconds before that blows.
I mean exactly what I wrote- the steady state system pressure is determined by compression of the air space due to expansion of the coolant as it heats up and the volumetric increment of expansion at normal operating temperature is pretty much 500cm3 as per my calcs and the operating temperature only shows small changes under normal circumstances. Of course if and when the system goes unstable the temperature will increase and on our later models the digi dash will show an alarm- anyone who ignores such alarm and does not shutdown the engine immediately deserves all they may get attrition wise.
If and when the coolant reaches its saturation temperature at the cylinder wall interface it will start to boil and then all bets are off. The sudden increase in volume due to steam generation will blow the pressure cap open- fitting a higher pressure rating cap will only buy a few seconds before that blows.
What you are saying to the laymen in here, including me, is that the pressure relief cap does nothing, because pressure is determined by the air space. That's the way it reads, to us.
Im saying that up until that point it is, because more or less air space determines the rate of pressure increase but not a peak pressure, the cap releases pressure will maintain that peak after that. Just how the words flow...to us...
Either way, too full just annoys the cap, too low may take overheating to allow the cap to do anything (Engine generating active steam)
What you are saying to the laymen in here, including me, is that the pressure relief cap does nothing, because pressure is determined by the air space. That's the way it reads, to us.
Im saying that up until that point it is, because more or less air space determines the rate of pressure increase but not a peak pressure, the cap releases pressure will maintain that peak after that. Just how the words flow...to us...
Either way, too full just annoys the cap, too low may take overheating to allow the cap to do anything (Engine generating active steam)
jeff,
Your first sentence is correct. The operating pressure is determined by the free air space volume in the upper half of the expansion tank and the amount of expansion of the coolant as it heats up.
Your second sentence is not correct- the rate at which the pressure builds up is determined by the rate at which the coolant heats up and that typically takes about 3 minutes. When the tank is filled correctly the air space in the tank will pressurise very close to 15 psig.
If the tank is overfilled the pressure cap will lift and blow off pressure however when this happens the cap will likely not fully reseat until the pressure drops to about 90% of the cap set pressure due to the hysteresis effect. If and when the system overheats the pressure will increase slightly until such time as the coolant boils and once that happens the system will be out of control with steam pissing out from under the bonnet..
With regard to your third sentence overfilling can cause the cap to lift . Under filling [i.e. low level] will cause the system to operate at a lower than optimal pressure wherein the coolant would boil at a lower temperature were it ever to be exposed to such.
The only problem I have seen with the cooling system was when my water pump threw an impeller- the temperature rose rapidly to the red line, the digi dash alarm went off and I immediately shut the motor down- the only damage being the hole in my wallet
Under filling [i.e. low level] will cause the system to operate at a lower than optimal pressure wherein the coolant would boil at a lower temperature were it ever to be exposed to such.
Wow, I never knew this or even thought of it. Interesting.
Wow, I never knew this or even thought of it. Interesting.
Whoah there, if cavitation from steam pockets is -not- a design problem, then this is not a problem.
Pressure has zero effect on coolant heat transfer.
Pressure can impact LOCAL heating where steam cavitation is a problem.
"If and when the coolant reaches its saturation temperature at the cylinder wall interface it will start to boil and then all bets are off. The sudden increase in volume due to steam generation will blow the pressure cap open- fitting a higher pressure rating cap will only buy a few seconds before that blows."
In an otherwise functioning 928 cooling system, this never happens. I've run on a warm (80s) day with a cut in a radiator hose, that only under pressure leaked at all, and I made it 2hrs home. A 1 atmosphere (cap off) cooling system on a 928 functions just fine, however it's absolutely not ideal for this and other reasons.
If you generate enough friction (rings at 10,000 rpm) or enough HP (exhaust valve seat and port energy being pushed out) I'm sure you can start to have this form of cap-off activity.
You appear to be going off on a number of seemingly irrelevant tangents albeit it is difficult to see why you appear to be somewhat confused- hopefully the following observations will help you better visualise the reality of the situation we are talking about:
1.The only thing the coolant is doing is expanding. It has no direct action on the cap venting pressure.
The coolant is acting like a piston to compress the captive air pocket and that action determines the operating pressure of the system. If the expansion tank is overfilled it will then generate a static pressure that will cause the pressure cap to lift and vent the pressure until the cap reseats. When this scenario happens, after the engine is shut down and the system cools, the subsequent contraction will then start to pull a partial vacuum and the vacuum breaker built into the cap will thus open and allow air into the system.
2. The volume of coolant will only get so hot, and will expand the same, every single time. But it needs a place to go...the air space in the tank.
That bit you clearly understand
3. A tank nearly full (with LESS air space) can reach cap vent rate, before a coolant tank that is say, mostly air space.
An completely filled expansion tank most definitely will blow the pressure cap. A tank that is underfilled will never make optimum pressure under stable operation.
4. I can make a 13psi cap vent at or below normal operating temperature all day long, just provide no air space in the tank.
No, you cannot make a cap vent all day long by filling the tank. If you do fill the tank as the liquid coolant heats up it will immediately blow the pressure cap given the coolant is non compressible and this will happen in a number of very small steps until the total expansion volume of circa 0.5 litres is ejected. When the motor is stopped and the coolant returns to ambient temperature it will try to pull a vacuum and the cap vacuum breaker will open. The engine will thus have lost 0.5 litres of coolant. The next time the engine heats up it will push the air out again but should not lose coolant . Thus why the tank should be filled to the midway seam so that the 1 litre of air can compress down to 0.5 litres and create a stable system without ever blowing the pressure cap. Thus why filling the expansion tank with coolant would be a dumb idea.
Wow, I never knew this or even thought of it. Interesting.
Originally Posted by FredR
Jeff,
You appear to be going off on a number of seemingly irrelevant tangents albeit it is difficult to see why you appear to be somewhat confused- hopefully the following observations will help you better visualise the reality of the situation we are talking about:
1.The only thing the coolant is doing is expanding. It has no direct action on the cap venting pressure.
The coolant is acting like a piston to compress the captive air pocket and that action determines the operating pressure of the system. If the expansion tank is overfilled it will then generate a static pressure that will cause the pressure cap to lift and vent the pressure until the cap reseats. When this scenario happens, after the engine is shut down and the system cools, the subsequent contraction will then start to pull a partial vacuum and the vacuum breaker built into the cap will thus open and allow air into the system.
Thats what I said too.
Originally Posted by FredR
2. The volume of coolant will only get so hot, and will expand the same, every single time. But it needs a place to go...the air space in the tank.
That bit you clearly understand
3. A tank nearly full (with LESS air space) can reach cap vent rate, before a coolant tank that is say, mostly air space.
An completely filled expansion tank most definitely will blow the pressure cap. A tank that is underfilled will never make optimum pressure under stable operation.
That's what I said too...
Originally Posted by FredR
4. I can make a 13psi cap vent at or below normal operating temperature all day long, just provide no air space in the tank.
No, you cannot make a cap vent all day long by filling the tank. If you do fill the tank as the liquid coolant heats up it will immediately blow the pressure cap given the coolant is non compressible and this will happen in a number of very small steps until the total expansion volume of circa 0.5 litres is ejected. When the motor is stopped and the coolant returns to ambient temperature it will try to pull a vacuum and the cap vacuum breaker will open. The engine will thus have lost 0.5 litres of coolant. The next time the engine heats up it will push the air out again but should not lose coolant . Thus why the tank should be filled to the midway seam so that the 1 litre of air can compress down to 0.5 litres and create a stable system without ever blowing the pressure cap. Thus why filling the expansion tank with coolant would be a dumb idea.
I think you took that too literally, bad word choice on my part, I meant if you did that, it would instantly push water out, over and over, if you refilled it as before and ran the car again.
Whoah there, if cavitation from steam pockets is -not- a design problem, then this is not a problem.
Pressure has zero effect on coolant heat transfer.
Pressure can impact LOCAL heating where steam cavitation is a problem.
"If and when the coolant reaches its saturation temperature at the cylinder wall interface it will start to boil and then all bets are off. The sudden increase in volume due to steam generation will blow the pressure cap open- fitting a higher pressure rating cap will only buy a few seconds before that blows."
In an otherwise functioning 928 cooling system, this never happens. I've run on a warm (80s) day with a cut in a radiator hose, that only under pressure leaked at all, and I made it 2hrs home. A 1 atmosphere (cap off) cooling system on a 928 functions just fine, however it's absolutely not ideal for this and other reasons.
If you generate enough friction (rings at 10,000 rpm) or enough HP (exhaust valve seat and port energy being pushed out) I'm sure you can start to have this form of cap-off activity.
Jeff,
We have established in your previous post that you understand the concept of what happens.
Now it seems you need to differentiate between what happens under "normal" steady state operation and irregular "unsteady state". The case you stated above was an "irregular" but "steady state" operation- i.e. something was wrong but it did not impact the way the system operates.
The system design is such that it can operate acceptably with no pressure cap to seal the system but in doing so one would lose a significant margin of operational safety. Pure water boils at 100C at atmospheric pressure- assuming my engine is typical it runs with the temperature in the water bridge at 98C. In theory this tells us we could run with pure water and no pressure build up but there would be little to no safety margin. The stock 50% coolant elevates the boiling point as does the pressure build up when filled and sealed correctly. Due to the coolant boiling point elevation you had a part margin of safety thus why you were able to get back home. Had your water pump thrown its impeller you would be lucky if you got half a mile down the road- that is an unsteady state out of control event.
I mean exactly what I wrote- the steady state system pressure is determined by compression of the air space due to expansion of the coolant as it heats up and the volumetric increment of expansion at normal operating temperature is pretty much 500cm3 as per my calcs and the operating temperature only shows small changes under normal circumstances.
If I understand the dynamics properly... the 928 cooling system takes almost 12L to fill, assuming 2.5% expansion of water for every 100F then at 250F there needs to be room enough for about 0.75L expansion (6.25% of 12).
If I understand the dynamics properly... the 928 cooling system takes almost 12L to fill, assuming 2.5% expansion of water for every 100F then at 250F there needs to be room enough for about 0.75L expansion (6.25% of 12).
Ken,
No idea where you got your expansion coefficents from but I used data from my design files and considered it applicable in the range 20C to 100C. I also allowed for some expansion within the engine and rounded number calculated to 0.5 litre thus generally use the term "circa" when talking about the volume displaced. I also considered the working inventory to be 12 litres.
This thread seems to be degenerating into various sides arguing (and agreeing on) the same point using different language,
I have a suggestion:
Degrees fahrenheit has no place in a debate about boiling temperature of water in a cooling system. . Use degrees C where 100C or Celsius is boiling point and 0C is freezing. Fahrenheit is very good for describing a hot summers day and that is all.
Stick to PSI for pressure where it relates to Automobiles.
Various modern units like KPa, Pascals, etc have no place in conversation between educated gentlemen.
11-09-2022, 02:49 PM
Cooler, better...looking, radiator caps? Glycol % vs. psi info
