Quantitative Coolant Flow Simulation
10-14-2014, 09:05 PM
#51
Ok friends!
As it turns out, the original simulations were correctly pressurized at 88kPa under 1 atm of pressure, so that's great news for the original results!
Here are the two videos of the cavitating regions forming within the water pump housing itself. Know that, even though the cavitation through the Mazmart impeller is far less, the flow through it is less as well (these are the exact same simulations that the other results are from).
Please make sure the annotations are on, without them it's going to be pretty easy to misread the videos. The stock redline of 9300 Engine RPM occurs around step 225/300 (shown in the bottom left corner).
Stock Pump
Mazmart Pump
The only possible explanation to the Mazmart part being an "improvement" over stock that I've been able to find after reviewing the results is that there was some damage to some other component (thermostat perhaps, they're generally replaced at the same time as the pump) that had failed, causing an extreme restriction. Even a severely corroded, cavitating stock impeller will flow more than the Remedy impeller under all conditions other than trying to push an extremely viscous fluid (greater than 3St maybe? I'm not great with the chemistry part of it).
As it turns out, the original simulations were correctly pressurized at 88kPa under 1 atm of pressure, so that's great news for the original results!
Here are the two videos of the cavitating regions forming within the water pump housing itself. Know that, even though the cavitation through the Mazmart impeller is far less, the flow through it is less as well (these are the exact same simulations that the other results are from).
Please make sure the annotations are on, without them it's going to be pretty easy to misread the videos. The stock redline of 9300 Engine RPM occurs around step 225/300 (shown in the bottom left corner).
Stock Pump
Mazmart Pump
The only possible explanation to the Mazmart part being an "improvement" over stock that I've been able to find after reviewing the results is that there was some damage to some other component (thermostat perhaps, they're generally replaced at the same time as the pump) that had failed, causing an extreme restriction. Even a severely corroded, cavitating stock impeller will flow more than the Remedy impeller under all conditions other than trying to push an extremely viscous fluid (greater than 3St maybe? I'm not great with the chemistry part of it).
Last edited by Legot; 10-14-2014 at 09:40 PM.
10-14-2014, 09:31 PM
#53
Great work.
Does the simulation reduce the flow when the impeller blades are partially gassed? And is the green that extends off of the impeller the bubbles getting shoved down the pipe, or is it more bubbles forming due to the flow?
I had to get my used pump rebuilt. I expect it back this weekend, or next week. I intend to swap only the pump, and hit a DE the second week in November. I will report back on the difference. Previously, on a warm day, I would stay above 7,000 RPM up a long hill and through a sweeping right hander. About a 1/4 mile down the straight, I would see temps around 210 or more. This is about 10+ degrees higher than the start of the hill (if memory serves). As soon as I lifted, or got back down below about 6,000 (braking zone turn 1), the temps would dive back down. Before my tank float finished sinking, this is where the coolant light would go on as well, and stay on until after the braking zone. My assumption was that I was getting some cavitation, that the impeller was gassed, and that flow went to hell. By start finish, the bubbles were making it through the thermostat and into the reservoir, and my have sunk float would sink the rest of the way and turn off the light.
Again, I will know more in a month.
Does the simulation reduce the flow when the impeller blades are partially gassed? And is the green that extends off of the impeller the bubbles getting shoved down the pipe, or is it more bubbles forming due to the flow?
I had to get my used pump rebuilt. I expect it back this weekend, or next week. I intend to swap only the pump, and hit a DE the second week in November. I will report back on the difference. Previously, on a warm day, I would stay above 7,000 RPM up a long hill and through a sweeping right hander. About a 1/4 mile down the straight, I would see temps around 210 or more. This is about 10+ degrees higher than the start of the hill (if memory serves). As soon as I lifted, or got back down below about 6,000 (braking zone turn 1), the temps would dive back down. Before my tank float finished sinking, this is where the coolant light would go on as well, and stay on until after the braking zone. My assumption was that I was getting some cavitation, that the impeller was gassed, and that flow went to hell. By start finish, the bubbles were making it through the thermostat and into the reservoir, and my have sunk float would sink the rest of the way and turn off the light.
Again, I will know more in a month.
10-14-2014, 09:44 PM
#54
Yes, the reduction of flow caused by the cavitation is simulated!
The long part is the outlet, and the bubbles that you see forming down near the end only occur well above 10k Engine RPM. They're not exactly forced in that direction, but they begin to form there and their existence kind of pulsates with the change in pressure from the openings between the individual impeller blades. If the animation used more of the simulation steps, rather than one every 15 steps, the pulsing would look more violent.
The stock redline of 9300 Engine RPM occurs around step 225/300 (shown in the bottom left corner). Each step in the animation is ~150 RPM increase at the impeller or a ~125 RPM increase at the ES.
The long part is the outlet, and the bubbles that you see forming down near the end only occur well above 10k Engine RPM. They're not exactly forced in that direction, but they begin to form there and their existence kind of pulsates with the change in pressure from the openings between the individual impeller blades. If the animation used more of the simulation steps, rather than one every 15 steps, the pulsing would look more violent.
The stock redline of 9300 Engine RPM occurs around step 225/300 (shown in the bottom left corner). Each step in the animation is ~150 RPM increase at the impeller or a ~125 RPM increase at the ES.
10-14-2014, 09:54 PM
#55
Hmmm. I dunno what to tell you man, this is good analysis but it just doesn't jive with my reality. Good condition pump and thermostat came off, mazmart parts went in and I never saw 220F again hitting all the same tracks in all the same conditions. Barely break 210.
From my chem eng days, a few things come to mind.. flow restrictions elsewhere in the system, vapour pressure of coolant being different from pure water, the rotation speed of the pump being different from engine rpm (it should turn faster given the pulley sizing),... Something's missing I feel. One easy test would be to put a flow meter in the rad hose and get actual flow and compare to the simulation. It has to add up right.
Maybe it's all in the thermostat but stock and mazmart should be full open by the 210s. I would be very very cautious saying 'glad I didn't buy this'.
From my chem eng days, a few things come to mind.. flow restrictions elsewhere in the system, vapour pressure of coolant being different from pure water, the rotation speed of the pump being different from engine rpm (it should turn faster given the pulley sizing),... Something's missing I feel. One easy test would be to put a flow meter in the rad hose and get actual flow and compare to the simulation. It has to add up right.
Maybe it's all in the thermostat but stock and mazmart should be full open by the 210s. I would be very very cautious saying 'glad I didn't buy this'.
10-14-2014, 10:18 PM
#56
rev it up
Interesting conclusions, I have been running the REmedy for about 5 years now with no notable difference I simply assumed it is doing its work.
I also have a higher pressure cap which in theory increases the boiling point of the water and perhaps have an effect on cavitation
I also have a higher pressure cap which in theory increases the boiling point of the water and perhaps have an effect on cavitation
10-14-2014, 11:23 PM
#57
SARX Legend
iTrader: (46)
Hmmm. I dunno what to tell you man, this is good analysis but it just doesn't jive with my reality. Good condition pump and thermostat came off, mazmart parts went in and I never saw 220F again hitting all the same tracks in all the same conditions. Barely break 210.
From my chem eng days, a few things come to mind.. flow restrictions elsewhere in the system, vapour pressure of coolant being different from pure water, the rotation speed of the pump being different from engine rpm (it should turn faster given the pulley sizing),... Something's missing I feel. One easy test would be to put a flow meter in the rad hose and get actual flow and compare to the simulation. It has to add up right.
Maybe it's all in the thermostat but stock and mazmart should be full open by the 210s. I would be very very cautious saying 'glad I didn't buy this'.
From my chem eng days, a few things come to mind.. flow restrictions elsewhere in the system, vapour pressure of coolant being different from pure water, the rotation speed of the pump being different from engine rpm (it should turn faster given the pulley sizing),... Something's missing I feel. One easy test would be to put a flow meter in the rad hose and get actual flow and compare to the simulation. It has to add up right.
Maybe it's all in the thermostat but stock and mazmart should be full open by the 210s. I would be very very cautious saying 'glad I didn't buy this'.
While I agree this test is not 100% conclusive that the Mazmart pump is not worth its price of admission, I personally don't believe the pump really improves anything and is unnecessary for 95% of the people who purchase it.
Stating you had improvements after installing it doesn't really mean a lot since you changed the fluid and installed new parts. I am certain changing the coolant and installing a new oem thermostat and pump would have gotten you the same results. I had the same items and saw no difference over the stock parts. I am turbo (water cooled turbo) and in South Texas.
Hopefully he can test a S2 pump at some point.
Last edited by 9krpmrx8; 10-14-2014 at 11:29 PM.
10-14-2014, 11:43 PM
#58
While I agree this test is not 100% conclusive that the Mazmart pump is not worth its price of admission, I personally don't believe the pump really improves anything and is unnecessary for 95% of the people who purchase it.
Stating you had improvements after installing it doesn't really mean a lot since you changed the fluid and installed new parts. I am certain changing the coolant and installing a new oem thermostat and pump would have gotten you the same results. I had the same items and saw no difference over the stock parts. I am turbo (water cooled turbo) and in South Texas.
Hopefully he can test a S2 pump at some point.
Stating you had improvements after installing it doesn't really mean a lot since you changed the fluid and installed new parts. I am certain changing the coolant and installing a new oem thermostat and pump would have gotten you the same results. I had the same items and saw no difference over the stock parts. I am turbo (water cooled turbo) and in South Texas.
Hopefully he can test a S2 pump at some point.
So, take a look at this: http://www.meglobal.biz/media/produc...Global_MEG.pdf
Properties of aqueous ethylene glycol solutions starting about page 18. A few relevant points that point to water not being a good proxy for coolant in this simulation: vapour pressure of 50% glycol solution is about 20% lower than water at 180 F ( btw what temp is the simulation at?). This would make cavitation onset less severe than simulated.
Also, 50% glycol is about twice as viscous as water at 180F. I know it's not noticeable but viscosity is funny like that. This would make for a different flow profile altogether.
So I'd seriously look into validating this simulation either with better fluid parameters or with an actual flow meter.
Last edited by Loki; 10-14-2014 at 11:55 PM.
10-14-2014, 11:49 PM
#59
Loki, most of those factors could be ignored for the purposes of this simulation. It's goal was to find the best approximation of the flowrate of these pumps ever made, that can be applicable to any situation using one of the two impellers. Potential flow restrictions, properties of other coolants, and pulley sizing (duh), were all ignored because the data needs to be useful rather than perfectly tuned for a specific situation. Also, I can't afford a flowmeter that would be usable for the high temps/pressures, or high flowrates that happen in an engine (I priced them before starting the simulations, lol).
We have to remember that coolant temperature and flow are only indirectly related. Under identical external conditions more coolant flow through the radiator will always allow more cooling than less flow, and that is a fact. And I do mean identical; Ambient air temp, wind speed, humidity, road temp, everything, even time spent idling and the backpressure from your exhaust would have to be the same. There are allot of factors that could change coolant temperatures by 10F, the pump flow is only one of them.
I'm personally suspicious of a rise in oil temperatures when the car is not in motion causing the coolant temperatures to rise later on (the oil cooling setup in our cars is kind of terrible btw), but I have no data on this, it's just an idea.
The only time that I would consider the Mazmart pump is any engine going above 10k RPM, and even before that I'd go with the stock impeller on an underdrive pulley. The mazmart pump doesn't even add any "bling," it's basically just a cool upgrade to tell Honda guys you have. I'd like to hear what Paul from Mazmart has to say about this, he does have the right to defend his product I suppose (not that I'm trying to defame it), I'd also like to know if he has any reasons for going with this particular design, with these particular characteristics.
We have to remember that coolant temperature and flow are only indirectly related. Under identical external conditions more coolant flow through the radiator will always allow more cooling than less flow, and that is a fact. And I do mean identical; Ambient air temp, wind speed, humidity, road temp, everything, even time spent idling and the backpressure from your exhaust would have to be the same. There are allot of factors that could change coolant temperatures by 10F, the pump flow is only one of them.
I'm personally suspicious of a rise in oil temperatures when the car is not in motion causing the coolant temperatures to rise later on (the oil cooling setup in our cars is kind of terrible btw), but I have no data on this, it's just an idea.
The only time that I would consider the Mazmart pump is any engine going above 10k RPM, and even before that I'd go with the stock impeller on an underdrive pulley. The mazmart pump doesn't even add any "bling," it's basically just a cool upgrade to tell Honda guys you have. I'd like to hear what Paul from Mazmart has to say about this, he does have the right to defend his product I suppose (not that I'm trying to defame it), I'd also like to know if he has any reasons for going with this particular design, with these particular characteristics.
10-15-2014, 08:47 AM
#60
I don't think you can ignore vapour pressure if you're going to talk about cavitation. Cavitation, by definition, is fluid dropping below its vapour pressure.
If you're going to present a result, you gotta know people will challenge it
Anyway, it's where this simulation, while cool, turns into "just a cool upgrade to tell Honda guys you have" is where I personally have a concern. This simulation is not enough to draw such a conclusion at all, and you have contradicting real world data. No serious engineer would accept this as is...
BTW, I'm not saying 1 of my runs had lower temps than some other runs. I have logs on multiple days before and logs on multiple days after and there's a consistent improvement. Logs include ambient temp btw, so it can be accounted for and pressure and humidity are orders of magnitude lower contributing factors.
If you're going to present a result, you gotta know people will challenge it
Anyway, it's where this simulation, while cool, turns into "just a cool upgrade to tell Honda guys you have" is where I personally have a concern. This simulation is not enough to draw such a conclusion at all, and you have contradicting real world data. No serious engineer would accept this as is...
BTW, I'm not saying 1 of my runs had lower temps than some other runs. I have logs on multiple days before and logs on multiple days after and there's a consistent improvement. Logs include ambient temp btw, so it can be accounted for and pressure and humidity are orders of magnitude lower contributing factors.
Last edited by Loki; 10-15-2014 at 08:58 AM.
10-15-2014, 11:58 AM
#61
I agree with Loki, unfortunately this simulation is a good start but is missing a TON of important details. I have helped design flow calculation simulators for CAD applications (NX Unigraphics) in my last job working with jet engines and their fluid flow mechanics (oil/water/fuel, etc) so I can say through experience a lot of significant data is missing. Data such as:
- weight of the impeller blades themselves as well as the entire rotational assembly. Centrifugal forces are nice, but we cant forget about the orientation of the impeller and the weight.
- The chemical properties of all the metals used in the system. Each type of metal has its own atomic properties, some will absorb fluids and create a slick surface for flow, while some will create more turbulant flow
- The surface friction of the metals. Flow is only as good as the surface friction properties of all metals used. Have you factored that? This isn't an easy task once you start to factor all the mating surfaces and gusset and such
- The surface friction of the impeller bearing, as well as the torque needed to move the impeller due to the lubrication and bearing material used. Something as simple as that can throw off a simulation believe it or not
- The fluid properties through the system (water vs coolant as well as the temperatures of each). Just because a pump creates cavitation at a rotational speed with a fluid at room temperature, doesn't mean it cavitates at a hotter or colder state.
The list goes on, and i am sure you know that, but my general point is to make sure that the mazmart pump isn't discounted as a viable upgrade because of this simulation.
- weight of the impeller blades themselves as well as the entire rotational assembly. Centrifugal forces are nice, but we cant forget about the orientation of the impeller and the weight.
- The chemical properties of all the metals used in the system. Each type of metal has its own atomic properties, some will absorb fluids and create a slick surface for flow, while some will create more turbulant flow
- The surface friction of the metals. Flow is only as good as the surface friction properties of all metals used. Have you factored that? This isn't an easy task once you start to factor all the mating surfaces and gusset and such
- The surface friction of the impeller bearing, as well as the torque needed to move the impeller due to the lubrication and bearing material used. Something as simple as that can throw off a simulation believe it or not
- The fluid properties through the system (water vs coolant as well as the temperatures of each). Just because a pump creates cavitation at a rotational speed with a fluid at room temperature, doesn't mean it cavitates at a hotter or colder state.
The list goes on, and i am sure you know that, but my general point is to make sure that the mazmart pump isn't discounted as a viable upgrade because of this simulation.
10-16-2014, 02:48 AM
#62
well, I never worked nor I know much about flow chemistry. But speaking from personal experience. I'm having results same as Loki ---- With the pump, things are cooler, will stay cooler for longer. regardless of what I do. compare to stock pump. so I bought my 2nd pump not that long ago for my 2nd Rx8 for this reason.
10-16-2014, 10:13 AM
#66
Registered
iTrader: (2)
Join Date: Oct 2007
Location: Pacific Northwest
Posts: 16,684
Likes: 0
Received 239 Likes
on
109 Posts
I've been following the thread, and I've got a couple questions that I don't expect immediate answers to, just food for thought:
1) The Mazmart pump is charted to flow less across a given RPM range, as well as cavitate less. If the Mazmart pump was spun faster to flow more, would it also hit the same caviation at the same flow rates as the stock pump? In other words, if you were to do a pulley size change on the Mazmart pump so that it flows as much as the stock pump, would it replicate the stock pump's caviation?
2) More flow = more heat transfer through the radiator. That's pretty well understood by everyone here I think. However, how much flow does the RX-8 actually require in various conditions? How much flow is needed to remove X heat? The Mazmart may flow much less, but if it flows enough for an heat abusive track car, then the fact that it flows less isn't actually a problem. I'm not saying that it is more than sufficient, I'm saying that we aren't really discussing how much flow is the baseline minimum needed, and it's a very key point to understand. This will also change depending on what radiator is in use. A more efficient radiator will shed the same amount of heat at a lower flow rate, all else being equal.
And a side comment:
If the Mazmart pump has sufficient flow for track use (which means certainly enough for street use), then there IS a point to using it, since it has less parasitic draw on the engine, so more power to the ground. There is a tangible gain there, as long as you aren't compromising your cooling system in the process.
1) The Mazmart pump is charted to flow less across a given RPM range, as well as cavitate less. If the Mazmart pump was spun faster to flow more, would it also hit the same caviation at the same flow rates as the stock pump? In other words, if you were to do a pulley size change on the Mazmart pump so that it flows as much as the stock pump, would it replicate the stock pump's caviation?
2) More flow = more heat transfer through the radiator. That's pretty well understood by everyone here I think. However, how much flow does the RX-8 actually require in various conditions? How much flow is needed to remove X heat? The Mazmart may flow much less, but if it flows enough for an heat abusive track car, then the fact that it flows less isn't actually a problem. I'm not saying that it is more than sufficient, I'm saying that we aren't really discussing how much flow is the baseline minimum needed, and it's a very key point to understand. This will also change depending on what radiator is in use. A more efficient radiator will shed the same amount of heat at a lower flow rate, all else being equal.
And a side comment:
If the Mazmart pump has sufficient flow for track use (which means certainly enough for street use), then there IS a point to using it, since it has less parasitic draw on the engine, so more power to the ground. There is a tangible gain there, as long as you aren't compromising your cooling system in the process.
Last edited by RIWWP; 10-16-2014 at 10:15 AM.
10-16-2014, 10:28 AM
#67
SARX Legend
iTrader: (46)
Yeah I think the problem with most claiming improvements is that they are replacing old tired hardware and coolant and then citing performance improvements. It's like guys with high mileage ignition coils replacing them with a new BHR kit and then claiming gains. But we all form our own opinions in different ways so..... But yes these test are somewhat inconclusive. But I don't see any data at all from Mazmart either, I would like to see that if it is available.
Last edited by 9krpmrx8; 10-16-2014 at 10:46 AM.
10-16-2014, 11:02 AM
#68
Yeah I think the problem with most claiming improvements is that they are replacing old tired hardware and coolant and then citing performance improvements. It's like guys with high mileage ignition coils replacing them with a new BHR kit and then claiming gains. But we all form our own opinions in different ways so..... But yes these test are somewhat inconclusive. But I don't see any data at all from Mazmart either, I would like to see that if it is available.
I still remember when I first had the pump changed, I noticed cooler running temp. that's what most (if not all) buyers cared about, this is an actual, real life gain. their design might not be perfect, but seriously, it's a ton better than the Bullshit original pump/impeller that's pretty much the same **** for the past 30 -40 fooking years.
Last edited by nycgps; 10-16-2014 at 11:04 AM.
10-16-2014, 11:30 AM
#69
SARX Legend
iTrader: (46)
I sold my Mazmart pump with only 20,000 miles on it (installed with the new engine) only to have the new owner (a mod here) send me a video of it leaking thru the weep hole when he installed it. I only pulled it because I was doing some aesthetics on the front cover and then decided it didn't really offer any benefit (being turbo I rarely wind it out to redline) and just installed a new OEM pump back on it. And maybe it was leaking before I removed it and I just didn't notice since the Evans is clear.
I know a couple of others have had failures or needed rebuilds as well, not sure what the failure rate is but it was still pretty disappointing to me to find that it leaked with such low mileage on it. The new owner is having it rebuilt by Mazmart.
I know a couple of others have had failures or needed rebuilds as well, not sure what the failure rate is but it was still pretty disappointing to me to find that it leaked with such low mileage on it. The new owner is having it rebuilt by Mazmart.
Last edited by 9krpmrx8; 10-16-2014 at 11:33 AM.
10-16-2014, 11:57 AM
#70
There have actually been significant design changes to rotary water pumps over the years, but no-one really cares. I'd love to see the accurate transient flow rate data and cavitation models for these pumps using real coolant that everyone seems to have on hand.
Paimon, all of those things are accounted for. Weight and orientation won't make an actual difference but they're factored in, the properties of all the materials are factored in (I would have thought that to be obvious too but whatever). We don't really care about the bearing friction because it would be frivolous and difficult (mentioned that earlier). Lastly, of course the temperature is taken into account, I'm not an idiot. It's even mentioned earlier in the thread once or twice.
NYC, you don't know what you're talking about, but that's ok because that's the reason this thread exists. There are exactly two variables that make this so apparently unsatisfying for some of you. These are the losses associated with flow through the radiator and peripheral coolant passages (which I will never test), and the fact that it doesn't use the most commonly used coolant (water is actually the most effective coolant, and there are good reasons to simulate it instead of some mix of stuff).
In order to get you guys you stop complaining I'll do 4 more simulations. I'll even make them dead simple, so if anyone argues it will be very easy to explain where any misunderstanding happens. They will be static flow at 9300 engine RPM in both pure water (again), and in a 50/50 glycol mix, each lasting for less than a second. Transients are obviously too confusing and unconvincing for the internet.
Riwwp, I might try that cavitation vs flow idea with the mazmart pump, but it's more trial and error than anything and that's a real pain. With the very gradual increase in cavitation around the inner portion of the blades, I'd expect the mazmart pump to reach speed above 20k RPM at the impeller before random, violent surging starts to happen. Surging coming from that impeller would be way more violent thanks to the design of the lower disk.
Paimon, all of those things are accounted for. Weight and orientation won't make an actual difference but they're factored in, the properties of all the materials are factored in (I would have thought that to be obvious too but whatever). We don't really care about the bearing friction because it would be frivolous and difficult (mentioned that earlier). Lastly, of course the temperature is taken into account, I'm not an idiot. It's even mentioned earlier in the thread once or twice.
NYC, you don't know what you're talking about, but that's ok because that's the reason this thread exists. There are exactly two variables that make this so apparently unsatisfying for some of you. These are the losses associated with flow through the radiator and peripheral coolant passages (which I will never test), and the fact that it doesn't use the most commonly used coolant (water is actually the most effective coolant, and there are good reasons to simulate it instead of some mix of stuff).
In order to get you guys you stop complaining I'll do 4 more simulations. I'll even make them dead simple, so if anyone argues it will be very easy to explain where any misunderstanding happens. They will be static flow at 9300 engine RPM in both pure water (again), and in a 50/50 glycol mix, each lasting for less than a second. Transients are obviously too confusing and unconvincing for the internet.
Riwwp, I might try that cavitation vs flow idea with the mazmart pump, but it's more trial and error than anything and that's a real pain. With the very gradual increase in cavitation around the inner portion of the blades, I'd expect the mazmart pump to reach speed above 20k RPM at the impeller before random, violent surging starts to happen. Surging coming from that impeller would be way more violent thanks to the design of the lower disk.
10-16-2014, 01:59 PM
#71
I would be careful assuming we don't understand and complain. I think we all want to get to the bottom of this (and other rotary water pump changes, that sounds super interesting), but the bottom is not the first conclusion that pops out. It has to actually be accurate.
RIWWP has a super valid point btw, how much flow is really needed?
And I can contribute my logs, FWIW, just need to dig them up and normalize, and I'm willing to buy a flow meter if it's not $$$ and take actual measurements, for what it's worth.
RIWWP has a super valid point btw, how much flow is really needed?
And I can contribute my logs, FWIW, just need to dig them up and normalize, and I'm willing to buy a flow meter if it's not $$$ and take actual measurements, for what it's worth.
Last edited by Loki; 10-16-2014 at 02:10 PM.
10-16-2014, 08:10 PM
#73
NYC, you don't know what you're talking about, but that's ok because that's the reason this thread exists. There are exactly two variables that make this so apparently unsatisfying for some of you. These are the losses associated with flow through the radiator and peripheral coolant passages (which I will never test), and the fact that it doesn't use the most commonly used coolant (water is actually the most effective coolant, and there are good reasons to simulate it instead of some mix of stuff).
In order to get you guys you stop complaining I'll do 4 more simulations. I'll even make them dead simple, so if anyone argues it will be very easy to explain where any misunderstanding happens. They will be static flow at 9300 engine RPM in both pure water (again), and in a 50/50 glycol mix, each lasting for less than a second. Transients are obviously too confusing and unconvincing for the internet.
In order to get you guys you stop complaining I'll do 4 more simulations. I'll even make them dead simple, so if anyone argues it will be very easy to explain where any misunderstanding happens. They will be static flow at 9300 engine RPM in both pure water (again), and in a 50/50 glycol mix, each lasting for less than a second. Transients are obviously too confusing and unconvincing for the internet.
what does that mean to me? the pump works well. and I guess I don't need to go to school to know what works and what doesn't. and don't assume ppl know nothing (eg: H2O has higher heat carrying capacity than glycol)
TBH, what you did was cool, but it does not change the fact that something is missing in your simulation, and at this moment it STILL does NOT 100% reflect real life scenario.
If that means "complaining" to you, then I got nothing else to say.
Last edited by nycgps; 10-17-2014 at 04:26 AM.
10-16-2014, 09:52 PM
#74
@ Legot,
I think what you have done is grand. Now, part of that is because I recognize the work that went into it, and part of it is because it actually lines up pretty well with my observations and assumptions and folk data based on the factory pump, and the remedy pump. I see weird cooling behavior above 7,500 RPM. The temps go up quickly, quicker than they should given the power I am putting down. And, if I short shift, there is no real issue at all and I am only putting slightly less power down. My assumption is that the pump is slipping and ventilating. The little tiny bubbles are dorking up the ability of the pump to grab water and send it some place. Instead of a viscous fluid like water, when those stupid bubbles are there you get a very very low viscosity fluid that is mostly a gas. That gas will just run around the blades, as opposed to get moved by them. Same thing occurs when a boat prop ventilates. In my old one, I had to sit and wait for all the little bubbles to go away before I could go anywhere (200 hp, 17 ft, 27 inch prop, water did not stand a chance).
A post here mentioned that an under-driven factory pump might do the same thing, and it would explain why people with pulleys do not have heating issues. The pump actually performs well at a lower speed.
I also think that your tool (not your work) may be letting you down a bit. Autodesk is a good tool, but you are on the fringe of where you need to start tracking molecules and that much computing horsepower is hard to come by. I know people that have done it (lots smarter than I am) and they say it is a PITA, especially if any cavitation occurs. They also use supercomputers for the work. The part about the blades just beating the gas as opposed to moving it is something I learned from them.
I do not believe anyone disputes your product. We are just trying to map the new data into the observations we have made. I can tell you the factory pump fits my observations almost exactly. Sustained high RPM, temp spike. And, previously, unfortunately before I had temp gauge, sustained high RPM, lift to brake, coolant light goes on. That would be when the lower RPM allowed the vanes to get a bite of the coolant and toss the gas/coolant mix through the system / thermostat / hose and into the coolant tank, where the float would sink in the foam. The light usually went out after several seconds.
I will happily share my observations with the Remedy. I got it back today. I have new coolant, but the old coolant had not been in there that long. I am using the same thermostat, so the test is reasonable. If it works out as expected, we will have a simulation, as well as data from a few sources that say the factory pump craps on itself above about 7,500 RPM. We will also have an assumption that the Remedy does not crap itself at higher RPM that will have at least one test case.
The fact that the series II pump vanes look like the love child of the 2 series 1 options is another interesting data point.
Gotta run
I think what you have done is grand. Now, part of that is because I recognize the work that went into it, and part of it is because it actually lines up pretty well with my observations and assumptions and folk data based on the factory pump, and the remedy pump. I see weird cooling behavior above 7,500 RPM. The temps go up quickly, quicker than they should given the power I am putting down. And, if I short shift, there is no real issue at all and I am only putting slightly less power down. My assumption is that the pump is slipping and ventilating. The little tiny bubbles are dorking up the ability of the pump to grab water and send it some place. Instead of a viscous fluid like water, when those stupid bubbles are there you get a very very low viscosity fluid that is mostly a gas. That gas will just run around the blades, as opposed to get moved by them. Same thing occurs when a boat prop ventilates. In my old one, I had to sit and wait for all the little bubbles to go away before I could go anywhere (200 hp, 17 ft, 27 inch prop, water did not stand a chance).
A post here mentioned that an under-driven factory pump might do the same thing, and it would explain why people with pulleys do not have heating issues. The pump actually performs well at a lower speed.
I also think that your tool (not your work) may be letting you down a bit. Autodesk is a good tool, but you are on the fringe of where you need to start tracking molecules and that much computing horsepower is hard to come by. I know people that have done it (lots smarter than I am) and they say it is a PITA, especially if any cavitation occurs. They also use supercomputers for the work. The part about the blades just beating the gas as opposed to moving it is something I learned from them.
I do not believe anyone disputes your product. We are just trying to map the new data into the observations we have made. I can tell you the factory pump fits my observations almost exactly. Sustained high RPM, temp spike. And, previously, unfortunately before I had temp gauge, sustained high RPM, lift to brake, coolant light goes on. That would be when the lower RPM allowed the vanes to get a bite of the coolant and toss the gas/coolant mix through the system / thermostat / hose and into the coolant tank, where the float would sink in the foam. The light usually went out after several seconds.
I will happily share my observations with the Remedy. I got it back today. I have new coolant, but the old coolant had not been in there that long. I am using the same thermostat, so the test is reasonable. If it works out as expected, we will have a simulation, as well as data from a few sources that say the factory pump craps on itself above about 7,500 RPM. We will also have an assumption that the Remedy does not crap itself at higher RPM that will have at least one test case.
The fact that the series II pump vanes look like the love child of the 2 series 1 options is another interesting data point.
Gotta run
