[Cooling System Upgrade] 1-BAR Radiator Cap

This is covered under Fluid Dynamics. The lower the pressure, the more likely you'd get cavitation at the impeller. Cavitation = lower total flow. In extreme and extended cases you'll even get erosion at the blades. In the case of a cooling system this also means lower heat transfer rate due to the bubbles in the system.

Cavitation as defined by Wikipedia :

"Cavitation is the formation and then immediate implosion of cavities in a liquid – i.e. small liquid-free zones ("bubbles") – that are the consequence of forces acting upon the liquid.[1] It usually occurs when a liquid is subjected to rapid changes of pressure that cause the formation of cavities where the pressure is relatively low."



Not saying this will happen in your case, but I would be cautious of reducing the designed pressure by half in a system that's been known to be weak to begin with. If you decide to test it, I would start small. Drive around town (no track days to start) and maybe avoid extended drives in elevated ambient temperatures. This is not an attack on the OP, just a little caution if you decide to proceed with the experiment.

It all depends on how hot the car gets. At 1 bar a standard 50/50 water/antifreeze mix boils at ~250 F. At 2 bar its another 35-45 degrees above that.

I dont know about BMWs specifically but anything north of 250 is pretty damn hot inst it? So it seems to me that at least form the perspective of not boiling your coolent 1 bar seems enough no?

The cap has been tested already on a dyno during various running conditions.

You guys all make very good points. One thing I'm curious about is why all other mfg (and even bmw on the newer models) operate their cooling systems at 1 BAR. My e90 and e82 had very solid cooling systems. Nothing like the e46. I'm trying to understand why the earlier models had 2 BAR systems while newer bmw's run 1 BAR.

Originally Posted by DHK

It all depends on how hot the car gets. At 1 bar a standard 50/50 water/antifreeze mix boils at ~250 F. At 2 bar its another 35-45 degrees above that.

I dont know about BMWs specifically but anything north of 250 is pretty damn hot inst it? So it seems to me that at least form the perspective of not boiling your coolent 1 bar seems enough no?

Thanks for the info. My coolant never rises above 190 degrees in the Texas summers. I'll have to wait a while to compare what my 1 BAR system will do in 100 degree temps. For now, I'm confident the vendors 1 and 1.5 BAR caps will alleviate some issues for e46 models.

Originally Posted by bryce-o

My e90 and e82 had very solid cooling systems. Nothing like the e46. I'm trying to understand why the earlier models had 2 BAR systems while newer bmw's run 1 BAR.

Electric water pumps. They never have to move at engine speed. Reduced speed equals lower required pressure to keep bubbles at bay at the impeller.

Originally Posted by cvx5832

This is covered under Fluid Dynamics. The lower the pressure, the more likely you'd get cavitation at the impeller. Cavitation = lower total flow. In extreme and extended cases you'll even get erosion at the blades. In the case of a cooling system this also means lower heat transfer rate due to the bubbles in the system.

Cavitation as defined by Wikipedia :

"Cavitation is the formation and then immediate implosion of cavities in a liquid – i.e. small liquid-free zones ("bubbles") – that are the consequence of forces acting upon the liquid.[1] It usually occurs when a liquid is subjected to rapid changes of pressure that cause the formation of cavities where the pressure is relatively low."

There is no "rapid" change of pressure in our cars, so cavitation is unlikely to occur. The pressure increases steadily as the temperature increases in the system. The question is whether the 1 bar cap will be able to hold the maximum pressure at operating temperature.

Originally Posted by SeanC

There is no "rapid" change of pressure in our cars, so cavitation is unlikely to occur. The pressure increases steadily as the temperature increases in the system. The question is whether the 1 bar cap will be able to hold the maximum pressure at operating temperature.

You didn't get what he was saying. There IS rapid change in pressure on the local level at the impeller. He is saying that by cutting the operating pressure in half , you may allow the impeller to cavitate the water it is trying to move. This reduces pump effciency greatly. What we don't know is what pressure the system runs at on a regular basis. I think OP needs to install a water pressure transducer to the system. You can't know the state of a fluid without temperature and pressure. I would go with a sensor on either side of the water pump so we can see the pressure drop.

Interesting.

Subbed for results

Originally Posted by WDE46

You didn't get what he was saying. There IS rapid change in pressure on the local level at the impeller. He is saying that by cutting the operating pressure in half , you may allow the impeller to cavitate the water it is trying to move. This reduces pump effciency greatly. What we don't know is what pressure the system runs at on a regular basis. I think OP needs to install a water pressure transducer to the system. You can't know the state of a fluid without temperature and pressure. I would go with a sensor on either side of the water pump so we can see the pressure drop.

If there is significant local pressure gradient (difference) at the impeller (which might very well be, don't get me wrong), wouldn't you expect this gradient to be larger in a system that operates at maximum 2 bars instead of 1 bar?

I signed up to calculate the pressure range at which our cooling system operates, but I need to know the volume of the empty space in the ET for max and min fill levels. Nobody has this info?

Originally Posted by bryce-o

the cooling system is presurized to 2-bar from the factory from BMW. a 1-bar cooling system will theoretically provide more reliable cooling system performance.

Good luck with your trials but initially I'm skeptical. If it's really that easy, why wouln't BMW do it in the first place? I agree with what bryce-o said. A 43-degree reduction in the boiling point is massive:

Originally Posted by fiveightandten

A 1 bar cap will lower the boiling point of the coolant by about 43º F if i'm doing the math right. That's a big jump. You may throw things out of the range of expected inputs from the ECT sensor and other elements and cause issues.

Sub'd as well, but I believe German Auto Solutions has been documenting this for some time now.

Originally Posted by SeanC

If there is significant local pressure gradient (difference) at the impeller (which might very well be, don't get me wrong), wouldn't you expect this gradient to be larger in a system that operates at maximum 2 bars instead of 1 bar?

I signed up to calculate the pressure range at which our cooling system operates, but I need to know the volume of the empty space in the ET for max and min fill levels. Nobody has this info?

The fluid viscosity and density change is negligible between 1-2 atm, so I don't see why the pressure gradient across the impeller blades would be more at 2 atm than 1 atm.

Originally Posted by SeanC

If there is significant local pressure gradient (difference) at the impeller (which might very well be, don't get me wrong), wouldn't you expect this gradient to be larger in a system that operates at maximum 2 bars instead of 1 bar?

Forget the discussion on localized pressures for a second. That's my fault in trying to go into too much detail.

By lowering the total system pressure, the lowest absolute pressure achieved at the impeller will be lower at a system with 1bar vs a system running at 2. Not saying it will be enough for cavitation, but rather saying that the potential for cavitation will be greater.

I am also interested at the results. Just throwing my thoughts in as part of the discussion.

Originally Posted by bryce-o

Thanks for the info. My coolant never rises above 190 degrees in the Texas summers. I'll have to wait a while to compare what my 1 BAR system will do in 100 degree temps. For now, I'm confident the vendors 1 and 1.5 BAR caps will alleviate some issues for e46 models.

That is simply not true, unless you've modified either the programming or the thermostat itself.

Originally Posted by cvx5832

Forget the discussion on localized pressures for a second. That's my fault in trying to go into too much detail.

By lowering the total system pressure, the lowest absolute pressure achieved at the impeller will be lower at a system with 1bar vs a system running at 2. Not saying it will be enough for cavitation, but rather saying that the potential for cavitation will be greater.

I am also interested at the results. Just throwing my thoughts in as part of the discussion.

Agreed.

Originally Posted by WDE46

This cap may work for casual driving, but what happens when you drive hard with it? You may get boiling coolant, which is totally ineffective at removing heat.

At 1 atmosphere (as in the cooling system isn't pressurized at all), the boiling point of a 50/50 mix should be around 250ºF. With a 1-bar cap, the whole system will have an absolute pressure of approximately 2 atmospheres - the boiling point will be even higher. If you're running that hot, you've got problems anyway.

I can't comment on any other potential issues, but boiling should not be an issue.

Originally Posted by WDE46

The fluid viscosity and density change is negligible between 1-2 atm, so I don't see why the pressure gradient across the impeller blades would be more at 2 atm than 1 atm.

Originally Posted by cvx5832

Forget the discussion on localized pressures for a second. That's my fault in trying to go into too much detail.

By lowering the total system pressure, the lowest absolute pressure achieved at the impeller will be lower at a system with 1bar vs a system running at 2. Not saying it will be enough for cavitation, but rather saying that the potential for cavitation will be greater.

I am also interested at the results. Just throwing my thoughts in as part of the discussion.

Good discussion

My thoughts:

When you fill the system with coolant, the system is pressurized to 1 bar ~ 1 atm with the cap open. 1 bar cap means your system can only withstand 1 bar on top of top that, i.e. up to 2 bars total. Hence, the stock 2-bar cap can withstand pressures up to 3 bars. If you had a barometer that you can stick in the ET while operating the car, it would show anywhere between 1-to-3 bars. The question is whether it is higher than 2 bars or not (so that a 1 bar cap would suffice).

By changing the cap to 1 atm, you're not changing the lowest attainable pressure, which is still 1bar = 1 atm, the pressure at fillup with the cap open. The system cannot operate below this pressure by design (unlike in the case of crankcase ventilation, in which case the system has a bit of negative pressure, i.e. vacuum). Of course, if you live 10,000 ft above the sea level, your lowest attainable pressure decreases a little bit, but changing the cap has no effect on this number.

With a 2-bar cap, the lowest attainable pressure is still 1 atm (again, pressure at fill-up with the cap open).

The difference between these two is the maximum pressures that they can operate at. With a 2 bar cap, you have a wider range of pressure gradient across the system than you would have with a 1 bar cap. That is why I said the local pressure gradient must follow. You have a pressure gradient of 2 bars vs. 1 bar across the system, why would you expect a higher pressure gradient with the latter in your system's components? You don't

Originally Posted by SeanC

Good discussion

My thoughts:

When you fill the system with coolant, the system is pressurized to 1 bar ~ 1 atm with the cap open. 1 bar cap means your system can only withstand 1 bar on top of top that, i.e. up to 2 bars total. Hence, the stock 2-bar cap can withstand pressures up to 3 bars. If you had a barometer that you can stick in the ET while operating the car, it would show anywhere between 1-to-3 bars. The question is whether it is higher than 2 bars or not (so that a 1 bar cap would suffice).

By changing the cap to 1 atm, you're not changing the lowest attainable pressure, which is still 1bar = 1 atm, the pressure at fillup with the cap open. The system cannot operate below this pressure by design (unlike in the case of crankcase ventilation, in which case the system has a bit of negative pressure, i.e. vacuum). Of course, if you live 10,000 ft above the sea level, your lowest attainable pressure decreases a little bit, but changing the cap has no effect on this number.

With a 2-bar cap, the lowest attainable pressure is still 1 atm (again, pressure at fill-up with the cap open).

The difference between these two is the maximum pressures that they can operate at. With a 2 bar cap, you have a wider range of pressure gradient across the system than you would have with a 1 bar cap. That is why I said the local pressure gradient must follow. You have a pressure gradient of 2 bars vs. 1 bar across the system, why would you expect a higher pressure gradient with the latter in your system's components? You don't

Well I think the issue is that absolute pressure does matter.

To quote the wikipedia page

The physical process of cavitation inception is similar to boiling. The major difference between the two is the thermodynamic paths that precede the formation of the vapor. Boiling occurs when the local vapor pressure of the liquid rises above its local ambient pressure and sufficient energy is present to cause the phase change to a gas. Cavitation inception occurs when the local pressure falls sufficiently far below the saturated vapor pressure, a value given by the tensile strength of the liquid at a certain temperature.

If the pressure of the surrounding fluid is low enough, then the pockets of low pressure around an impeller may be at a low enough pressure to essentially lower the boiling point such that it actually starts to bubble. How low it would have to go, I'm not sure.

Just because the atmospheric pressure is 1 atm, doesn't mean that the pressure can't drop below that when in motion. By conservation, the areas where the water is moving faster would exert lower pressure than where it's moving slower.

Of course I should mention that I'm not a physicist and my understanding could be completely off base. But if I'm understanding this properly, then I could see it being a valid concern, especially at higher RPMs.

Originally Posted by TerraPhantm

Well I think the issue is that absolute pressure does matter.

To quote the wikipedia page


If the pressure of the surrounding fluid is low enough, then the pockets of low pressure around an impeller may be at a low enough pressure to essentially lower the boiling point such that it actually starts to bubble. How low it would have to go, I'm not sure.

Just because the atmospheric pressure is 1 atm, doesn't mean that the pressure can't drop below that when in motion. By conservation, the areas where the water is moving faster would exert lower pressure than where it's moving slower.

Of course I should mention that I'm not a physicist and my understanding could be completely off base. But if I'm understanding this properly, then I could see it being a valid concern, especially at higher RPMs.

Local pressures is not the issue here. The pressure can drop locally below 1 atm no problem, but it's integral over the entire system can never drop below 1 atm (in a closed system of course, due to conservation). It if drops below 1 atm by x amount across the impeller, it should increase by the same x amount somewhere else.

Regardless, we are not talking about the pressure at the impeller or how much it can drop during operation. Question is why would a 1 bar cap would allow a higher pressure gradient across the impeller than a 2-bar cap? If the cavitation is a concern with a 1-bar cap, it could very well be a concern with a 2-bar cap. In fact, more of a concern with the 2-bar cap than the 1-bar cap. 2-bar cap has a greater tolerance than a 1-bar cap, hence by conservation, could allow even lower pressures to be attained across the impeller.

Well, that clears up my confusion about a 1 bar cap...if it does mean 1 bar above the 1 bar pressure it's at naturally.

Based on that, then, then the only advantage I'd see for a lower pressure cap is that it would blow sooner than other components would blow...namely ET.

The cap isn't changing the pressure the system runs at, it just rescues it a little sooner when the pressure rises because of some cooling system failure. I don't think it should affect its operation at all; just offers more protection...it seems to me.