Quantitative Coolant Flow Simulation
09-29-2014, 09:47 AM
#26
Wow those are some pretty interesting results. Good stuff.
I guess the next interesting thing to know would be what volume of flow is most effective for removing heat.
Also pretty cool to finally know how much horsepower the water pump consumes. Almost 7BHP at redline is more than I thought it would be.
I guess the next interesting thing to know would be what volume of flow is most effective for removing heat.
Also pretty cool to finally know how much horsepower the water pump consumes. Almost 7BHP at redline is more than I thought it would be.
Higher flow will also always be better at mitigating hot spots in the engine for the same reasons.
Here's something I found when I first started.
http://stewartcomponents.com/index.p...ormation_id=14
09-29-2014, 10:55 AM
#29
Holy **** that's a beautiful piece engineering! See how the vanes sweep forward at the edges? That's a design element that increases flow velocity. Those holes in the base are to equalize pressure on both sides of the base disk, and help with bearing stresses. What nice impeller! I wouldn't be able to model something that fancy without scanning the physical impeller, anyone want to donate one? 
I don't think I can model the ideal impeller, sorry FI. What do you mean model the mazmart vs stock thermostat? Do you mean modeling the two impellers using their temperatures (it would barely change the results) or do you mean cavitation inside of the thermostat (which is entirely possible and the same for both, but why)?
I don't think I can model the ideal impeller, sorry FI. What do you mean model the mazmart vs stock thermostat? Do you mean modeling the two impellers using their temperatures (it would barely change the results) or do you mean cavitation inside of the thermostat (which is entirely possible and the same for both, but why)?
Last edited by Legot; 09-29-2014 at 11:03 AM.
09-29-2014, 11:05 AM
#30
I was thinking more along the lines of flow; is there a big difference between the MAzmart and stock.
Judging by your response to 9k's pic I am thinking that it's probable that the S2 impeller is a better product. If that is the case: 1) Is it possible to swap an S2 water pump into an S1 or if that is not possible, could just the impeller be swapped?
Judging by your response to 9k's pic I am thinking that it's probable that the S2 impeller is a better product. If that is the case: 1) Is it possible to swap an S2 water pump into an S1 or if that is not possible, could just the impeller be swapped?
09-29-2014, 11:37 AM
#31
SARX Legend
iTrader: (46)
The S2 uses a whole different front cover, a swap is not really feasible because the oil filter on the S2 is located on the front cover, etc. As for swapping the impeller, I am not sure. We have a parts car R3 here but I really don't want to pull the water pump off,
09-29-2014, 11:58 AM
#32
If you want to tear it up, press out the impeller, send it to me, and wait to get it back, I can 3D scan it and adapt it to the s1 housing.
Maybe BHR or someone else in my area has one that they're willing to let me borrow, hmmmm..... I can see it now "BHR water pump $2000 msrp."
Maybe BHR or someone else in my area has one that they're willing to let me borrow, hmmmm..... I can see it now "BHR water pump $2000 msrp."
09-30-2014, 06:25 AM
#33
Registered
Join Date: Sep 2014
Location: Oahu
Posts: 12
Likes: 0
Received 0 Likes
on
0 Posts
09-30-2014, 09:38 AM
#35
This is awesome work man. Love this stuff.
I can send you before and after AP logs for when I installed mine if you're interested. Same location, same regime, ambient temp logged so can be compensated out. I saw a consistent decrease of 10-15F average and peak temp.
I can send you before and after AP logs for when I installed mine if you're interested. Same location, same regime, ambient temp logged so can be compensated out. I saw a consistent decrease of 10-15F average and peak temp.
09-30-2014, 12:07 PM
#36
FULLY SEMI AUTOMATIC
iTrader: (9)
09-30-2014, 09:57 PM
#37
Cool data..
It looks like the stock S1 starts to cavitate around 7K (flow rate levels off, line gets squiggly). Looks like it gets worse from there.
At the track, my coolant light used to go off at about 8K from foam in the coolant, it would go off as soon as the RPM dropped. So this kind of correlates.
Foam is a lousy heat transfer fluid. today, sustained running above about 7,500 results in a temp creep (1 degree every 5 seconds) at both Roebling and Sebring. Enough that I abandon sessions when it is above 85 degrees out.
I will have a Remedy pump in (looks a lot like the one on the page before this) before DarkSide at Roebling, where I expect mid-80s in the afternoon. I will see if the temp issue goes away. If my coolant bottle still works, I will plug in the sensor and see what happens.
My thought is that if the pump is moving 90 gallons a minute of liquid with a remedy, that is possibly better for cooling than 110 gallons a minute of foam? But that is only if 90 gallons a minute is sufficient to cool the engine. That is a lot of water. That would fill a 15,000 gallon swimming pool in under 3 hours. Or, fill a gallon jug in 2/3 of a second. Or, a 55 gallon barrel in about 35 seconds. Again, I will know more in about 6 weeks.
Or, am I completely off base. Fluid Mechanics in 1984 did not cover this all that much.
It looks like the stock S1 starts to cavitate around 7K (flow rate levels off, line gets squiggly). Looks like it gets worse from there.
At the track, my coolant light used to go off at about 8K from foam in the coolant, it would go off as soon as the RPM dropped. So this kind of correlates.
Foam is a lousy heat transfer fluid. today, sustained running above about 7,500 results in a temp creep (1 degree every 5 seconds) at both Roebling and Sebring. Enough that I abandon sessions when it is above 85 degrees out.
I will have a Remedy pump in (looks a lot like the one on the page before this) before DarkSide at Roebling, where I expect mid-80s in the afternoon. I will see if the temp issue goes away. If my coolant bottle still works, I will plug in the sensor and see what happens.
My thought is that if the pump is moving 90 gallons a minute of liquid with a remedy, that is possibly better for cooling than 110 gallons a minute of foam? But that is only if 90 gallons a minute is sufficient to cool the engine. That is a lot of water. That would fill a 15,000 gallon swimming pool in under 3 hours. Or, fill a gallon jug in 2/3 of a second. Or, a 55 gallon barrel in about 35 seconds. Again, I will know more in about 6 weeks.
Or, am I completely off base. Fluid Mechanics in 1984 did not cover this all that much.
10-01-2014, 11:21 AM
#38
Sure thing Charles, just let me know and I'll send them over 
Green, the cavitation doesn't really cause foam. Cavitation is the formation of vapor bubbles in a liquid due to ultra low pressures. In all liquid, if the pressure is low enough it will turn into a gas regardless of the temperature. Since, in an impeller, the low pressure would only be behind the blades, once the liquid/vapor "foam" leaves the area of low pressure the vapor bubbles instantly implode and turn back into liquid.
It would be possible for foam to be created due to the low pressures at the impeller if the coolant was somehow really well oxygenated, but once all of the coolant flows through the impeller it would get trapped at the highest point in the system (the throttle body). Also, liquids hold significantly less gas at higher temperatures, so that's unlikely.
The formation and rapid destruction of vapor bubbles due to cavitation is what causes the reduction of flow since the volume of the fluid/vapor mixture sucked through the impeller in is effectively larger than what's being pumped in or out of the pump housing.
I have an animation that demonstrates what I'm talking about really well for both pumps, but it's going to take a day or so to render and upload them both.
Green, the cavitation doesn't really cause foam. Cavitation is the formation of vapor bubbles in a liquid due to ultra low pressures. In all liquid, if the pressure is low enough it will turn into a gas regardless of the temperature. Since, in an impeller, the low pressure would only be behind the blades, once the liquid/vapor "foam" leaves the area of low pressure the vapor bubbles instantly implode and turn back into liquid.
It would be possible for foam to be created due to the low pressures at the impeller if the coolant was somehow really well oxygenated, but once all of the coolant flows through the impeller it would get trapped at the highest point in the system (the throttle body). Also, liquids hold significantly less gas at higher temperatures, so that's unlikely.
The formation and rapid destruction of vapor bubbles due to cavitation is what causes the reduction of flow since the volume of the fluid/vapor mixture sucked through the impeller in is effectively larger than what's being pumped in or out of the pump housing.
I have an animation that demonstrates what I'm talking about really well for both pumps, but it's going to take a day or so to render and upload them both.
Last edited by Legot; 10-01-2014 at 11:23 AM.
10-01-2014, 12:59 PM
#39
Registered
Legot you seem to have a good grasp on what cavitation realy is. Us navy has studied this to death and back again then studied some more as cavitation from a subs prop at speeds is a realy bad thing. Back to cars cavitation isn't just at the pump your low and high pressure observation is right on. There is a high pressure zone on the back side of the thermostat and low pressure on the front side the more you pump the higher the high pressure zone gets. Not to mention all the twists and runs of the cooling system it's self creating turbulence slowing flow rates in the hole system. I'd say the mazmart pump is better for the fact it's flow rate is lower alowing high pressure zones to decrease as well as turbulence in the system in turn getting better flow through out the hole system. Your data is grate but because you don't look at the grater picture your conclusion is incorect. I discovered all this years ago when I built my 350 small block and slapped a high flow water pump on thinking more pump flow would be better I was greatly wrong over heated constantly because of what I just mention it wasn't till I went to a more stock lower flow pump did my heating problems go away. With all the fins in the coolant chamber of the engine more faster flow is going to create cavitation as well loosing flow through the block but lower that flow through the block will decrease cavitation in the block and increase flow. You realy need to do a model of the hole thing if you realy want true flow numbers and cooling ability of bouth stock and mazmart pumps.
10-01-2014, 01:16 PM
#40
@ Legot Yes, same issue with boat prop. I miss that boat....
If I am not foaming though, why would the float sink and trigger the dash light, only to have it go back out a few seconds after I dropped below 7,000 RPM (Engine)? I cannot run that part of the test, my coolant float is all the way dead, I checked today.
Or, given the flow resistance, once the pump cavitates, the bubbles stay in the housing, collect, screw up the flow, and enough made it out through the thermostat to sink the float...
I have no doubt that the bubbles will collapse back into the fluid. It is just when that will occur. And where.
Either way, I think it is freaking cool that you start to predict cavitation about the same place I have what I think are cavitation related problems. The question that really needs to get asked is if under-driving the water pump will help cure the issue? Looking back, I assume you are showing impeller RPM, right, which is close to double crank RPM? (have pity on the poor EE that barely passed Fluids).
If I am not foaming though, why would the float sink and trigger the dash light, only to have it go back out a few seconds after I dropped below 7,000 RPM (Engine)? I cannot run that part of the test, my coolant float is all the way dead, I checked today.
Or, given the flow resistance, once the pump cavitates, the bubbles stay in the housing, collect, screw up the flow, and enough made it out through the thermostat to sink the float...
I have no doubt that the bubbles will collapse back into the fluid. It is just when that will occur. And where.
Either way, I think it is freaking cool that you start to predict cavitation about the same place I have what I think are cavitation related problems. The question that really needs to get asked is if under-driving the water pump will help cure the issue? Looking back, I assume you are showing impeller RPM, right, which is close to double crank RPM? (have pity on the poor EE that barely passed Fluids).
10-01-2014, 01:17 PM
#41
Registered
Forgot to add too is did you account for the viscosity of the coolant? That can play a big part in foaming and cavitation. Straight water has a different viscosity than water with coolant wich can change flow rates as well. Grate question green I hadn't even thought of that he did say at engine speed so is that the rpm of the engine or impeller?
Last edited by niteshade247; 10-01-2014 at 01:21 PM.
10-01-2014, 02:46 PM
#42
Driving my unreliable rx8
Yeah and the imploding bubbles from cavitation can remove metal they cause a lot of problems but I would not expect a foam without a surfactant in the system. Water doesn't foam without soap. A little shake test with some coolant might show if it foams or not.
The increased Viscosity should reduce the Foaming action. It will also slow the flow rate. I would imagine it would decrease cavitation also.
9300 engine rpm is 11125 Water pump RPM. His chart showed Water Pump RPM.
The increased Viscosity should reduce the Foaming action. It will also slow the flow rate. I would imagine it would decrease cavitation also.
9300 engine rpm is 11125 Water pump RPM. His chart showed Water Pump RPM.
10-01-2014, 06:15 PM
#44
Sorry, if this first make much sense, I'm on a bus.
I did the simulation using pure water, I think I mentioned that somewhere. The reason being that pure water is the same for everyone, while there are several coolant/coolant mixes that someone could be using each is probably more viscous and less likely to cavitate (improving flow efficiency and increasing power consumption across the board). Also, it was my understanding that allot of racing leagues only allow water as coolant.
Shaking the coolant in a container to see if it foams won't show you anything. Cavitation and foaming are not the same thing. Foam is created when air is forcefully mixed into something, it cannot happen inside this pump. Cavitation is when a gas is created from the original liquid due to low pressure. It would be impossible for any air bubbles to collect anywhere aside from the peripheral hoses, the flow is way too high.
About the rest of the system having an effect the pump, yes, it does. However it's much less of an effect than you'd expect. The radiator is designed to be minimally restrictive on the system, as are the two main radiator hoses. The peripheral coolant hoses are at the mercy of the flow through the two rad hoses (I think every one is a venturi), and will not significantly effect flow through the pump. I'm estimating accuracy of the flow rate to be in the 0% to +15% range between the simulation and the real world.
Cavitation is actually really well handled within the pump housing, sure it's not great that it happens all, but there are some measures against it. For one, the weird lump on the outlet side is a measure to keep pressure at the impeller higher than the outflow pressure. This high pressure region before the true outlet completely eliminates the possibility of cavitation outside the pump itself.
I have exactly one guess as to why people "see" improvements from the mazmart impeller, but I have to review the data.
The two graphs that have to do with rpm are showing the Engine rpm, that's why they both say Engine RPM.
I did the simulation using pure water, I think I mentioned that somewhere. The reason being that pure water is the same for everyone, while there are several coolant/coolant mixes that someone could be using each is probably more viscous and less likely to cavitate (improving flow efficiency and increasing power consumption across the board). Also, it was my understanding that allot of racing leagues only allow water as coolant.
Shaking the coolant in a container to see if it foams won't show you anything. Cavitation and foaming are not the same thing. Foam is created when air is forcefully mixed into something, it cannot happen inside this pump. Cavitation is when a gas is created from the original liquid due to low pressure. It would be impossible for any air bubbles to collect anywhere aside from the peripheral hoses, the flow is way too high.
About the rest of the system having an effect the pump, yes, it does. However it's much less of an effect than you'd expect. The radiator is designed to be minimally restrictive on the system, as are the two main radiator hoses. The peripheral coolant hoses are at the mercy of the flow through the two rad hoses (I think every one is a venturi), and will not significantly effect flow through the pump. I'm estimating accuracy of the flow rate to be in the 0% to +15% range between the simulation and the real world.
Cavitation is actually really well handled within the pump housing, sure it's not great that it happens all, but there are some measures against it. For one, the weird lump on the outlet side is a measure to keep pressure at the impeller higher than the outflow pressure. This high pressure region before the true outlet completely eliminates the possibility of cavitation outside the pump itself.
I have exactly one guess as to why people "see" improvements from the mazmart impeller, but I have to review the data.
The two graphs that have to do with rpm are showing the Engine rpm, that's why they both say Engine RPM.
10-02-2014, 10:06 AM
#45
Question: wouldn't you need to take into account (or did you) the cooling system pressure? When fully warmed up the system runs what, 15psig static all around?
Also for the coolant mix, wouldn't the vaporization characteristics be significantly different from pure water, the vapour pressure would be higher, so harder to cavitate? All of this is rusty for me, but, well, I was thinking in the shower.
Also for the coolant mix, wouldn't the vaporization characteristics be significantly different from pure water, the vapour pressure would be higher, so harder to cavitate? All of this is rusty for me, but, well, I was thinking in the shower.
10-02-2014, 10:19 AM
#46
Question: wouldn't you need to take into account (or did you) the cooling system pressure? When fully warmed up the system runs what, 15psig static all around?
Also for the coolant mix, wouldn't the vaporization characteristics be significantly different from pure water, the vapour pressure would be higher, so harder to cavitate? All of this is rusty for me, but, well, I was thinking in the shower.
Also for the coolant mix, wouldn't the vaporization characteristics be significantly different from pure water, the vapour pressure would be higher, so harder to cavitate? All of this is rusty for me, but, well, I was thinking in the shower.
I'm not sure about the vaporization properties of coolant vs water, I'm terrible with chemistry.