AISIN WATER PUMP PASSAGE SURFACE ROUGHness
#1
Pole Position
Thread Starter
AISIN WATER PUMP PASSAGE SURFACE ROUGHness
Aisin is the OEM supplier of water pumps to Lexus...(but you already knew that).
I observed certain passage surface areas of the pump were not polished so as to aleviate roughness of the interior casting.
To me this interior surface roughness snags flowing coolant and creates turbulence, imparting minor flow resistance.increasing with RPM.. (probably good for 1/2 H.P. -1 H.P. @6,000 RPM)
So I elected to remove the offending surface roughness before installing the water pump. This is before image of back o Aisin water pump depicting surface roughness of coolant passageways.
This is after polishing water pump passages with Dremel kit..(Impeller covered with tape to prevent ingress of abrasive dust into bearing interior area)
I removed 90% of roughness starting with Dremel 80, then 120 grit flapwheels, then with Dremel finishing buffs. Patience using the abrasives will prevent nicking or scuffing the gasket mating sutface area.
Dremel 4300 Series kit Dremel kit contents...
Highly recommend Dremel flexible shaft attachment #522 (Just like the dentist)
Dremel 80 and 120 grit flapwheels with Dremel 511E finishing abrasive buffs were acquired from supply chain partner Home Depot. It is important to let the abrasive do the work and to constantly move the abrasive element for uniform surface results.
I observed certain passage surface areas of the pump were not polished so as to aleviate roughness of the interior casting.
To me this interior surface roughness snags flowing coolant and creates turbulence, imparting minor flow resistance.increasing with RPM.. (probably good for 1/2 H.P. -1 H.P. @6,000 RPM)
So I elected to remove the offending surface roughness before installing the water pump. This is before image of back o Aisin water pump depicting surface roughness of coolant passageways.
This is after polishing water pump passages with Dremel kit..(Impeller covered with tape to prevent ingress of abrasive dust into bearing interior area)
I removed 90% of roughness starting with Dremel 80, then 120 grit flapwheels, then with Dremel finishing buffs. Patience using the abrasives will prevent nicking or scuffing the gasket mating sutface area.
Dremel 4300 Series kit Dremel kit contents...
Highly recommend Dremel flexible shaft attachment #522 (Just like the dentist)
Dremel 80 and 120 grit flapwheels with Dremel 511E finishing abrasive buffs were acquired from supply chain partner Home Depot. It is important to let the abrasive do the work and to constantly move the abrasive element for uniform surface results.
Last edited by YODAONE; 05-01-17 at 08:39 PM.
#2
That's how to do the inlet that you wanted to extrude hone. Doing that smoothing doesn't accomplish much on the pump, it would be better to just deburr the openings. The internal water jackets are not smoothed and likely have flashing in them too, and smoothing some areas may have the wrong effect as well, as it essentially reduces surface area. Analogous to sanding off the cooling fins of a motorcycle engine as an extreme example. And if it was necessary wouldn't lexus have put out a technical service bulletin to make sure to do that before installing the coolant pump?
#4
Pole Position
Thread Starter
#5
Drop in the bucket relative to the rough casting inside the entire engine cooling system. A little turbulence at the walls is desirable for increasing heat transfer, but then again you are not transferring much heat between the pump and the atmosphere.
#6
Pole Position
Thread Starter
That's how to do the inlet that you wanted to extrude hone. Doing that smoothing doesn't accomplish much on the pump, it would be better to just deburr the openings. The internal water jackets are not smoothed and likely have flashing in them too, and smoothing some areas may have the wrong effect as well, as it essentially reduces surface area. Analogous to sanding off the cooling fins of a motorcycle engine as an extreme example. And if it was necessary wouldn't lexus have put out a technical service bulletin to make sure to do that before installing the coolant pump?
It's efficiency is dictated by the impeller shape/configuration (Aisin gets an "A"), the conguration of the chambers and passages Aisin did not perfect this..too many sharp edges so "B-" and how smooth are the passages...here Aisin receives a "D".
Aisin could easily correct the passages.
I was infirmed that engine power required to drive water pump is exponential with RPM and not linear, so small improvements in flow efficiency can yield improvement.
#7
It's efficiency is dictated by the impeller shape/configuration (Aisin gets an "A"), the conguration of the chambers and passages Aisin did not perfect this..too many sharp edges so "B-" and how smooth are the passages...here Aisin receives a "D".
Aisin could easily correct the passages.
Aisin could easily correct the passages.
Trending Topics
#8
Moderator
iTrader: (7)
good idea, but gains are very negligible at best. not worth the effort in my opinion but heck if you had an afternoon to kill and perform such minor details, more power to ya! just wished the impeller would be made of metal instead of plastic. there are costs to consider and I am certain smoothing out casting imperfections was not on the top of the list either for Aisin.
#9
Lexus Test Driver
iTrader: (1)
Besides pumping engine coolant, the water pump is not performing a heat exchange function..(.as would internal engine block passages) so surface area of pump coolant passages is not really contributing to cooling.as say the radiator
It's efficiency is dictated by the impeller shape/configuration (Aisin gets an "A"), the conguration of the chambers and passages Aisin did not perfect this..too many sharp edges so "B-" and how smooth are the passages...here Aisin receives a "D".
Aisin could easily correct the passages.
I was informed that engine power required to drive water pump is exponential with RPM and not linear, so small improvements in flow efficiency can yield improvement.
It's efficiency is dictated by the impeller shape/configuration (Aisin gets an "A"), the conguration of the chambers and passages Aisin did not perfect this..too many sharp edges so "B-" and how smooth are the passages...here Aisin receives a "D".
Aisin could easily correct the passages.
I was informed that engine power required to drive water pump is exponential with RPM and not linear, so small improvements in flow efficiency can yield improvement.
Here's one nice web page that covers most of the fluid dynamics relevant to this situation:
https://en.wikipedia.org/wiki/Darcy%...sbach_equation
(Darcy Weisbach equation, in case the link does not work)
It does carefully cover drag effects of surface roughness, specifically for pipe flow, but that would also apply for coolant flow within the water pump.
But near the end of that page, they have an especially useful paragraph, titled Practical Application:
"In a hydraulic engineering application, it is typical for the volumetric flow Q within a pipe (that is, its productivity) and the head loss per unit length S (the concomitant power consumption) to be the critical important factors. The practical consequence is that, for a fixed volumetric flow rate Q, head loss S decreases with the inverse fifth power of the pipe diameter, D. Doubling the diameter of a pipe of a given schedule (say, ANSI schedule 40) roughly doubles the amount of material required per unit length and thus its installed cost. Meanwhile, the head loss is decreased by a factor 1/32 (about a 97% reduction). Thus the energy consumed in moving a given volumetric flow of the fluid is cut down dramatically for a modest increase in capital cost."
You can be sure that any engineers working on the cooling system will know about that. It basically means that if you ever care about reducing pressure drop, you can dramatically do that (to the FIFTH power) by increasing the size of the pipe (or obviously, in this case, flow channel within a water pump). In a cooling system like this, most of the pressure drop will be at the radiator, with the very small flow channels in there; they are so small in the radiator itself because they need to maximize surface area to promote heat transfer. So basically, the Aisin designers, if they had been concerned with pressure drop at the pump, would have solved it by increasing the size of the flow channel rather than worrying about surface roughness.
BTW, the repeated dyno results are impressively repeatable - they must have been done very carefully, and with good equipment.
#10
Pole Position
Thread Starter
More specifically, the engine power for pumping will be close to quadratic with RPM.
Here's one nice web page that covers most of the fluid dynamics relevant to this situation:
https://en.wikipedia.org/wiki/Darcy%...sbach_equation
(Darcy Weisbach equation, in case the link does not work)
It does carefully cover drag effects of surface roughness, specifically for pipe flow, but that would also apply for coolant flow within the water pump.
But near the end of that page, they have an especially useful paragraph, titled Practical Application:
"In a hydraulic engineering application, it is typical for the volumetric flow Q within a pipe (that is, its productivity) and the head loss per unit length S (the concomitant power consumption) to be the critical important factors. The practical consequence is that, for a fixed volumetric flow rate Q, head loss S decreases with the inverse fifth power of the pipe diameter, D. Doubling the diameter of a pipe of a given schedule (say, ANSI schedule 40) roughly doubles the amount of material required per unit length and thus its installed cost. Meanwhile, the head loss is decreased by a factor 1/32 (about a 97% reduction). Thus the energy consumed in moving a given volumetric flow of the fluid is cut down dramatically for a modest increase in capital cost."
You can be sure that any engineers working on the cooling system will know about that. It basically means that if you ever care about reducing pressure drop, you can dramatically do that (to the FIFTH power) by increasing the size of the pipe (or obviously, in this case, flow channel within a water pump). In a cooling system like this, most of the pressure drop will be at the radiator, with the very small flow channels in there; they are so small in the radiator itself because they need to maximize surface area to promote heat transfer. So basically, the Aisin designers, if they had been concerned with pressure drop at the pump, would have solved it by increasing the size of the flow channel rather than worrying about surface roughness.
BTW, the repeated dyno results are impressively repeatable - they must have been done very carefully, and with good equipment.
Here's one nice web page that covers most of the fluid dynamics relevant to this situation:
https://en.wikipedia.org/wiki/Darcy%...sbach_equation
(Darcy Weisbach equation, in case the link does not work)
It does carefully cover drag effects of surface roughness, specifically for pipe flow, but that would also apply for coolant flow within the water pump.
But near the end of that page, they have an especially useful paragraph, titled Practical Application:
"In a hydraulic engineering application, it is typical for the volumetric flow Q within a pipe (that is, its productivity) and the head loss per unit length S (the concomitant power consumption) to be the critical important factors. The practical consequence is that, for a fixed volumetric flow rate Q, head loss S decreases with the inverse fifth power of the pipe diameter, D. Doubling the diameter of a pipe of a given schedule (say, ANSI schedule 40) roughly doubles the amount of material required per unit length and thus its installed cost. Meanwhile, the head loss is decreased by a factor 1/32 (about a 97% reduction). Thus the energy consumed in moving a given volumetric flow of the fluid is cut down dramatically for a modest increase in capital cost."
You can be sure that any engineers working on the cooling system will know about that. It basically means that if you ever care about reducing pressure drop, you can dramatically do that (to the FIFTH power) by increasing the size of the pipe (or obviously, in this case, flow channel within a water pump). In a cooling system like this, most of the pressure drop will be at the radiator, with the very small flow channels in there; they are so small in the radiator itself because they need to maximize surface area to promote heat transfer. So basically, the Aisin designers, if they had been concerned with pressure drop at the pump, would have solved it by increasing the size of the flow channel rather than worrying about surface roughness.
BTW, the repeated dyno results are impressively repeatable - they must have been done very carefully, and with good equipment.
#11
Pole Position
Thread Starter
4" impeller stats H.P consumed vs RPM
3" Impeller: H.P. consumed vs. RPM.. Considerably more H.P. consumed increased RPM vs flow.. Any impediments to flow results in power consumption...
I understand Meziere uses CMC milling on internals so these figures reflect smoother surfaces than unfinished rough internal casting of internal flow surface area typical of OEM. ..
Both Meziere and Edelbrock technical support agree that rough internal flow surfaces do not promote efficiency and are in fact impediments to flow.
#12
Pole Position
Thread Starter
A water pump manufacturer shows data that flatters their product - I'm shocked.
The surface texture doesn't mean much, you're changing the thickness and characteristics of the boundary layer, which is more turbulent and much slower than the main stream. Even with a mirror polish, the Poiseville flow will lag far behind.
I do it anyway.
The surface texture doesn't mean much, you're changing the thickness and characteristics of the boundary layer, which is more turbulent and much slower than the main stream. Even with a mirror polish, the Poiseville flow will lag far behind.
I do it anyway.
For those with 1998-2000 LS400's, check out this water inlet modification using 2001-2006 LS430 water inlet housing to pick up a few ponies
https://www.clublexus.com/forums/ls-1st-and-2nd-gen-1990-2000/802539-ls430-water-inlet-housing-in-ls400.html
A few images from post; This image depicts why Lexus modified the water inlet housing in LS430 over 1998-2000 LS400.....
1998-2000 LS400 water inlet housing...you can't see through to the other side..and coolant has to navigate around two 90 degree bends....requires extra engine power...even at low RPM
The LS430 water inlet housing is much more efficient,..you can see through to the other side although mass produced and still room for improvement...some sharp edges that will Dremel away. The boss at near lower right does not have the LS400 minor coolant tube..but it is there to accept a barbed brass nipple after drilling and tapping..
Last edited by YODAONE; 05-02-17 at 09:33 PM.
#13
Besides pumping engine coolant, the water pump is not performing a heat exchange function..(.as would internal engine block passages) so surface area of pump coolant passages is not really contributing to cooling.as say the radiator
It's efficiency is dictated by the impeller shape/configuration (Aisin gets an "A"), the conguration of the chambers and passages Aisin did not perfect this..too many sharp edges so "B-" and how smooth are the passages...here Aisin receives a "D".
Aisin could easily correct the passages.
I was infirmed that engine power required to drive water pump is exponential with RPM and not linear, so small improvements in flow efficiency can yield improvement.
It's efficiency is dictated by the impeller shape/configuration (Aisin gets an "A"), the conguration of the chambers and passages Aisin did not perfect this..too many sharp edges so "B-" and how smooth are the passages...here Aisin receives a "D".
Aisin could easily correct the passages.
I was infirmed that engine power required to drive water pump is exponential with RPM and not linear, so small improvements in flow efficiency can yield improvement.
#14
And when the thermostat is closed there is next to zero HP required to turn the coolant pump. Only time hp is used is when its open. There is no hp savings in smoothing that water pump casting. And funny stuff "flow efficiency" what is the flow with the thermostat closed? Come on. And if the pump is causing a power loss go with an electric one.
In the end unless you are building a race car all this is a huge waste of everything, especially to drive to work or to the store.
In the end unless you are building a race car all this is a huge waste of everything, especially to drive to work or to the store.
Last edited by dicer; 05-03-17 at 10:42 AM.
#15
An impeller pump like a vacuum cleaner will unload when the inlet or outlet is blocked just try it. If the fluid is not moving it can't be pumped. So the only resistance is the impeller trying to move in the water. So make it a tesla blade less and there will be almost no loss there either.