Reverse Flow/Electric Water Pump Cooling System
#62
Water boils at a whopping 2*C higher at 3 additional PSI of pressure. I don't think nucleate boiling is going to be influenced much seeing how the walls of a cylinder head can far exceed 200*C. If you want to safe guard your cooling system from nucleate boiling you just raise the flow rate through the system dramatically. This knocks the steam pockets off and re-establishes conductive heat transfer.
Higher speed flow should only induce "plug flow" as the Reynolds number of the fluid will only go up with velocity. Plug flow increases heat uptake so the faster the flow the better.
Higher speed flow should only induce "plug flow" as the Reynolds number of the fluid will only go up with velocity. Plug flow increases heat uptake so the faster the flow the better.
#63
I saw a great explanation of how that myth arose, but I can't remember it at the moment.
Joe, your explanation of more pressure head due to flow resistance = higher boiling point, is one of the reasons why Evans coolant may be a win for an electric pump application. (Plus the steam pocket GM patent infringement thing).
Another way of regulating coolant temps is to use shutters in front of the rad, like some prop planes lol. When the shutters are closed, drag would be reduced to boot.
#64
Boost Pope
iTrader: (8)
Join Date: Sep 2005
Location: Chicago. (The less-murder part.)
Posts: 33,049
Total Cats: 6,608
Yeah, it finally dawned on me that this was Rafa's concern.
Like Jason says, this is not a problem. Let's pick out one random molecule of water from the cooling system and follow it for a while. We start with a low coolant velocity. The water molecule is at, say, 150°F (a guess) when it comes out of the radiator and goes into the engine. It travels through the engine very slowly, having plenty of time to pick up heat. When it leaves the engine, it is at 220°F, so during the time it was in the engine it picked up enough heat to raise its own temperature by 70°F.
Now, let's follow the molecule again, but this time we speed things up, we double the speed of the water. Again, it goes in at 150°F, but it travels through the engine so quickly that it only has time to pick up enough heat to raise itself by 40°F, so it's at 180°F when it leaves. The water absorbed less heat from the engine.
Now, so far everything I have said is in fact true. By increasing the speed of the water, it collected less heat from the engine, and the engine was not cooled by as great an amount.
But, we're overlooking one critically important thing. When we double the speed at which the coolant passes through the engine, we also double the number of trips that Mr. Water Molecule makes in any given amount of time. So even though the water is absorbing less heat per trip, it's making more trips. As a result, the overall amount of cooling is increased, and just as importantly, the temperature gradient across the engine is decreased! All good things.
Joe, your explanation of more pressure head due to flow resistance = higher boiling point, is one of the reasons why Evans coolant may be a win for an electric pump application.
#65
Elite Member
iTrader: (22)
Join Date: Dec 2006
Location: Sunny Spanish speaking Non US Caribbean
Posts: 3,224
Total Cats: 3
Yeah, it finally dawned on me that this was Rafa's concern.
Like Jason says, this is not a problem. Let's pick out one random molecule of water from the cooling system and follow it for a while. We start with a low coolant velocity. The water molecule is at, say, 150°F (a guess) when it comes out of the radiator and goes into the engine. It travels through the engine very slowly, having plenty of time to pick up heat. When it leaves the engine, it is at 220°F, so during the time it was in the engine it picked up enough heat to raise its own temperature by 70°F.
Now, let's follow the molecule again, but this time we speed things up, we double the speed of the water. Again, it goes in at 150°F, but it travels through the engine so quickly that it only has time to pick up enough heat to raise itself by 40°F, so it's at 180°F when it leaves. The water absorbed less heat from the engine.
Now, so far everything I have said is in fact true. By increasing the speed of the water, it collected less heat from the engine, and the engine was not cooled by as great an amount.
But, we're overlooking one critically important thing. When we double the speed at which the coolant passes through the engine, we also double the number of trips that Mr. Water Molecule makes in any given amount of time. So even though the water is absorbing less heat per trip, it's making more trips. As a result, the overall amount of cooling is increased, and just as importantly, the temperature gradient across the engine is decreased! All good things.
Many thanks
#67
Elite Member
iTrader: (22)
Join Date: Dec 2006
Location: Sunny Spanish speaking Non US Caribbean
Posts: 3,224
Total Cats: 3
I see what you mean. My original thread was missing this part of Joe's quote: "But, we're overlooking one critically important thing. When we double the speed at which the coolant passes through the engine, we also double the number of trips that Mr. Water Molecule makes in any given amount of time. So even though the water is absorbing less heat per trip, it's making more trips. As a result, the overall amount of cooling is increased, and just as importantly, the temperature gradient across the engine is decreased! All good things."
Sorry for that.
Many thanks for this thread!
#69
Slowing the flow to allow a molecule to absorb heat is a complete myth, at the molecular level heat transfer is effectively instantaneous.
The same is true for the radiator; however, we maximize surface area at the expense of flow velocity.
Last edited by fluke; 06-07-2009 at 05:15 PM.
#70
Btw, with all these talks about the coolant system mods, has anyone tried plumbing in a swirlpot and reroute the the paths similar to what guys do in rally turbo setups or at least to what the factory routing for something like an Impreza STI?
When you have a watercooled turbo, after 10-15 minutes of track driving coolant going through the turbo starts boiling hard - that introduces bubbles in the coolant which hurt the coolant performance big time. The system needs a swirlpot extracting the bubbles from the boiling coolant, etc.
The Miata coolant system is routed marginally ok for the stock anemic N/A setup. Coolant reroute helps things a bit, but that routing has never been made with high hp and turbo boiling coolant in mind ...
When you have a watercooled turbo, after 10-15 minutes of track driving coolant going through the turbo starts boiling hard - that introduces bubbles in the coolant which hurt the coolant performance big time. The system needs a swirlpot extracting the bubbles from the boiling coolant, etc.
The Miata coolant system is routed marginally ok for the stock anemic N/A setup. Coolant reroute helps things a bit, but that routing has never been made with high hp and turbo boiling coolant in mind ...
#71
Moderator
iTrader: (12)
Join Date: Nov 2008
Location: Tampa, Florida
Posts: 20,664
Total Cats: 3,013
Btw, with all these talks about the coolant system mods, has anyone tried plumbing in a swirlpot and reroute the the paths similar to what guys do in rally turbo setups or at least to what the factory routing for something like an Impreza STI?
When you have a watercooled turbo, after 10-15 minutes of track driving coolant going through the turbo starts boiling hard - that introduces bubbles in the coolant which hurt the coolant performance big time. The system needs a swirlpot extracting the bubbles from the boiling coolant, etc.
The Miata coolant system is routed marginally ok for the stock anemic N/A setup. Coolant reroute helps things a bit, but that routing has never been made with high hp and turbo boiling coolant in mind ...
When you have a watercooled turbo, after 10-15 minutes of track driving coolant going through the turbo starts boiling hard - that introduces bubbles in the coolant which hurt the coolant performance big time. The system needs a swirlpot extracting the bubbles from the boiling coolant, etc.
The Miata coolant system is routed marginally ok for the stock anemic N/A setup. Coolant reroute helps things a bit, but that routing has never been made with high hp and turbo boiling coolant in mind ...
#72
Btw, with all these talks about the coolant system mods, has anyone tried plumbing in a swirlpot and reroute the the paths similar to what guys do in rally turbo setups or at least to what the factory routing for something like an Impreza STI?
When you have a watercooled turbo, after 10-15 minutes of track driving coolant going through the turbo starts boiling hard - that introduces bubbles in the coolant which hurt the coolant performance big time. The system needs a swirlpot extracting the bubbles from the boiling coolant, etc.
When you have a watercooled turbo, after 10-15 minutes of track driving coolant going through the turbo starts boiling hard - that introduces bubbles in the coolant which hurt the coolant performance big time. The system needs a swirlpot extracting the bubbles from the boiling coolant, etc.
#75
Exactly, and that's like a constant source of bubbles straight into the engine. And I was told the coolant going through the turbo always starts boiling - no way around it (talking about performance driving with turbo in boost after every corner of course. street driving where the driver is in boost once in a year doesn't count)
#78
From what I was told - nope. Otherwise racers and OEM wouldn't have to waste time and money installing these swirlpots, but just dump to the radiator instead.
Exactly, and that's like a constant source of bubbles straight into the engine. And I was told the coolant going through the turbo always starts boiling - no way around it (talking about performance driving with turbo in boost after every corner of course. street driving where the driver is in boost once in a year doesn't count)
Exactly, and that's like a constant source of bubbles straight into the engine. And I was told the coolant going through the turbo always starts boiling - no way around it (talking about performance driving with turbo in boost after every corner of course. street driving where the driver is in boost once in a year doesn't count)
The problem occurs whe they are present in the intake stream of the water pump, the water pump creates a lower pressure condition at certain points and causes the extremely hot water to cavitate. this is why I don't dump the turbo to the pump inlet. Water pumps don't pump water w/ air in it very well, this is the main issue.
#79
Boost Pope
iTrader: (8)
Join Date: Sep 2005
Location: Chicago. (The less-murder part.)
Posts: 33,049
Total Cats: 6,608
At worst, these bubbles will collect at the top of the upper radiator tank and be pushed out into the expansion tank when the radiator cap vents. At best, they will simply collapse into liquid form as soon as they enter the radiator.
I'd also point out that if the coolant is in fact boiling inside the CHRA, then the cooling of the turbo is being compromised just the same as the cooling of the head is compromised when localized boiling occurs there. This suggests a need for increased coolant flow through the turbo, decreased temperature of coolant entering the turbo, or some other means to prevent or reduce this boiling as much as possible. I have no issue with boiling in the turbo after shutdown, but it shouldn't be happening during operation.