Seeking Water Pump Knowledge

Eric,
Davies Craig, the EWP supplier, specifies that you run without a tstat. The pump controller tries to slow down the pump enough to get water temp to the desired setpoint.
With the oversized pump, this doesn‘t work.

Ulix

Could you tell us which of the DC pumps you have on the two cars which overcool? I know DC offers a fairly wide variety of pumps of various capacities.

Are you already using one of the smaller pumps and it is too much or?

Thanks.
 
Personally if I really wanted to ensure proper cooling particularly with a water cooled turbo etc I would strongly investigate an electric pump and controller, perhaps even rolling your own controller using multiple sensors to better control the engine temperature.

I know Ulix’s system isn’t working properly but it isn’t an indicator of the performance of all possible systems.
 
I took some pictures and closer measurements of the two impeller designs that I have (AC equipped and non-AC equipped 1500).

The first four images are for the "AC" style, with indications of where I measured. These measurements may not be absolute, and may not even be relevant, but they should offer some idea of how the two impellers compare. The second four images are the same views of the "non-AC" style (same measurements were taken on both).

I'm hoping someone can confirm if one of them would generate more fluid pumping action than the other.

AC pump and housing:
AC1.JPG
AC2.JPG
AC3.JPG
ACH.JPG



Non-AC pump and housing:
NAC1.JPG
NAC2.JPG
NAC3.JPG
NACH.JPG


The dimensions for each are as follows (using the lettered indicators in the pictures):
AC pump and housing:
A) 60mm
B) 27mm
C) 35mm
D) 16mm
E) 15mm
F) 63mm
G) 37mm
H) 80mm

Non-AC pump and housing:
A) 64mm
B) 33mm
C) 32mm
D) 10mm
E) 10mm
F) 64mm
G) 40mm
H) 85mm

Both impellers have 7 blades of about the same contour and shape, but different size.
As you can see in the pictures and from the dimensions, the AC pump has a slightly smaller diameter impeller and housing but significantly taller (deeper) and longer impeller blades...compared to the non-AC pump.
Perhaps a good visual comparison can be made by looking at the oblique angles of the two (the third picture of each respectively).

I do not know to what extent the different dimensions make to the pumping properties. Is the overall diameter/size more critical than the blade surface area/size? The AC pump has larger blades but the non-AC pump is larger overall. As I understand another factor that may influence a pump's performance is the difference between the outside diameter of the impeller compared to the diameter of the housing around it. However both of these pumps have basically the same clearance (10mm) on either side of the impeller. Is it fair to make any assumptions solely from these differences, considering everything before and after the pumps are identical?

Another difference between the two versions of pumps is the sizes of the drive pulleys. As noted earlier there is some variance between examples of each type so the drive ratios are not completely consistent. But the AC pump has roughly a 120mm driver (crankshaft) pulley and a 109mm driven (water pump) pulley. The non-AC version has roughly a 130mm (crank) and 114mm (pump) pulleys. Using a simple online calculator, that indicates ratios of 1:1.1 (AC) and 1:1.14 (non-AC), and at a engine RPM of 2000 the AC pump will be rotating at 2202 RPM, while the non-AC pump will be rotating at 2281 RPM. So the non-AC pump actually spins faster than the AC pump. Although as noted, these will differ from one engine to the next due to several options of pulley sizes for all of them. If I were to guess, I imagine they were intended to turn at the same speed (pretty close to a 1:1 ratio).

Appreciate any input on this.
 
well actually you cant make those sort of assumptions...

In the post that's referred to in your "other" pump thread, there's mention of two almost identical X19's.... one that runs hot and one that does not... EXTERNALLY they are reported to look the same and be similar age etc etc... but behave differently... plus there are scores of stories about X9's that run hot (although none of mine ever have, so I do not believe there is some fundamental design flaw at work)

Again I feel it comes down to increased flow resistance... when you measure flow rates for water, you take into account the internal surfaces of the pipe / conduit you're using... and determine using whats known as a "Reynolds number" if the flow is laminar or turbulent... IMO the undercar pipes, being long enough and a significant portion of the flow path can affect the total flow of the system if the internals of these are rough and rusty, so instead of the internal surface of the pipe being smooth and promoting flow that is laminar, the flow becomes turbulent, so flow rate drops, pressure increases... so two "identical cars" but one with less care and maintenance and no anti freeze change has a crusty / rusty / powdery insides, and the other is clean and presents good clean surface area for efficient heat exchange throughout the entire system.

A/c or non A/c pumo... not quite correct nomeclature either, as the pump for A/c X19's is used on non A/c other Fiat models (like the strada and uno)

SteveC
 
If you did a test, assuming the same head clearance of the impellers, the longer and deeper the vanes that make up the impeller are, the more coolant it can move per revolution.

The sweep of the longer vane will have more volume of coolant in it’s sweep and thus more pumping.

As was pointed out, doing a test will resolve this more effectively than any of the measurements we might make of the actual artifacts. I think you fill find that even seemingly similar parts will deliver different results.

As we say in product development, one test counters all opinions regardless of who’s opinion it might be.
 
The Meziere pump that we use in the 600/Ecotec autocrosser is rated at 20 GPM. It's adequate for cooling on 95 deg days.
Not sure if specs are available. I think it's a WP-136.
 
I completely agree on all accounts of all replies. As a former research scientist I am guilty of doing exactly what we hate the most, try and make some sort of analysis based on a few bits of information and no evidence. But honestly I am not trying to be overly scientific about it. I think someone with knowledge of pump design could look at the images and general specs of the two and say if one is likely to move more fluid than the other. And that is all I am looking for, a general assessment.


the longer and deeper the vanes that make up the impeller are, the more coolant it can move per revolution
My general understanding of impellers (which is little) agrees with what Karl said. The one with more total vane area will move more fluid (all other things being equal, which for the most part they are). This supports what I've learned in the past for boat impellers and turbocharger impellers. Granted things are never that simple, so I am certainly open to more input from those that can enlighten me on this.

The "AC" pump (Steve, I'll continue to call it that for now - for lack of a better identifier) certainly has more vane area (mathematical calculations aren't need to see this) and seems like it should provide more circulation of the cooling system. This is a factory water pump designed for this application so it stands to reason that it will work on this engine. If both pumps were mounted on the same engine (which they can be...more on that later) in the same vehicle, then the "AC" pump should offer more movement of the coolant (not accounting for possible restrictions as Steve describes). I will not make any further assumptions beyond that, mainly because that was all I was wondering. But perhaps that is why this impeller was mated to the smaller AC style pump housing, to allow better cooling for cars with AC (and their inherent hotter operating temps). The smaller pump housing is part of the redesign required to make room for the AC compressor and location of the alternator, but it could have come with the same impeller as the non-AC pump.

As for utilizing a AC pump on a non-AC engine, or vice versa, I've managed to make both work. It does not require a lot of modifications, but it does require a list of different parts (stock replacements from the other set up) to go along with the pump and housing (mostly the alternator arrangement). Is it necessary? I don't care frankly, that isn't my focus with this discussion. But it can be done. Are there better options? Always, given the money and tools anything can be done better.

Another thing that might be done with the installation of a AC style water pump is pairing it with the larger diameter crank pulley of the non-AC set up. This will rotate the pump at a slightly higher RPM (relative to either stock combination of pulleys). Note that the increase in impeller speed is minimal so there won't be any concern for things like cavatition. And likewise it might not make a lot of difference in pump output. But combining the greater impeller vane area with slightly higher rotation speed may improve cooling circulation function (if the rest of the system allows for it, as Steve has said)....regardless if it needs it.
 
If your circulating more water than necessary - if the rest of the system can not "use" the additional flow - then all you're doing is sapping more crankshaft horsepower to drive the pump. If your circulating more water (than needed to actually keep the engine cool) you can run the risk of circulating the water too fast thru the radiator, and the time taken for the coolant to pass thru the radiator isn't sufficient to allow effective heat exchange - it needs sufficient transition time.

Having built many engines that produce easily twice the factory rated power, and used the stock cooling system to keep the thing cool with no issues... I've used both the pressed steel and cast impellers and haven't noted any real difference. I haven't used the "A/C" pump as the housing is seriously heavy! as is the air con style pulley...

SteveC
 
Ulix

Could you tell us which of the DC pumps you have on the two cars which overcool? I know DC offers a fairly wide variety of pumps of various capacities.

Are you already using one of the smaller pumps and it is too much or?

Thanks.

Karl,
It‘s the 110.
There is also an 80 model which would be better.
Last night I looked into adding the t-stat back into the mix.
This will will work, it seems.
There will be some temp ranges in which both the t-stat and the pump speed are reacting to the coolant temp at the same time, but at no time should they be working „against each other“, such as the t-stat being closed and the pump running at full speed to cool things down.
 
Karl,
It‘s the 110.
There is also an 80 model which would be better.
Last night I looked into adding the t-stat back into the mix.
This will will work, it seems.
There will be some temp ranges in which both the t-stat and the pump speed are reacting to the coolant temp at the same time, but at no time should they be working „against each other“, such as the t-stat being closed and the pump running at full speed to cool things down.

Thank you Ulix.

Given the intent of the DC controller it would seem going to the lower volume pump would be a thing to try. I know limitless pocketbook :)

If you are adding a t-stat back into the mix, are you using the three way from the twin cam or are you going back to a SOHC housing of some sort. If the latter where would (does) the DC temperature sensor live?

I ask as I am interested in an electric pump solution on my 850 (I know very different application) and one of my X’s and am trying to gather more data points.

Thank you.
 
Thank you Ulix.

Given the intent of the DC controller it would seem going to the lower volume pump would be a thing to try. I know limitless pocketbook :)

If you are adding a t-stat back into the mix, are you using the three way from the twin cam or are you going back to a SOHC housing of some sort. If the latter where would (does) the DC temperature sensor live?

I ask as I am interested in an electric pump solution on my 850 (I know very different application) and one of my X’s and am trying to gather more data points.

Thank you.

During the "design development" of the 600 I had 3 different engines in the car. 1st was a 903, then came a 1300/128, then we went mid-engined using the 2.2 Ecotec. I used the same 20 GPM pump on all 3 engines, just moved it as required to match the engine and radiator location.
 
but at no time should they be working „against each other“, such as the t-stat being closed and the pump running at full speed to cool things down
Even if the electric pump was trying to circulate while the T-stat was 'closed', with the bypass design of Fiat's T-stat assembly they would not be working against each other - the coolant will just go the other route of the bypass (it is never fully closed). So I don't see any downside to using a T-stat with this type pump. Perhaps on other engines that do not have a 'bypass' type T-stat it would be a different story, which is likely the reason they recommend no T-stat is used.
 
This is certainly not a accurate assessment, but based on some reading I've done about impeller design it seems to support the impression that the "AC" pump should push more fluid (greater flow rate) than the non-AC pump. According to those sources, some of the most critical factors for flow rate are the overall diameter of the impeller (larger is better), the height and length of the vanes (again, larger is better), and the speed of rotation (more on this in a bit).

Quick measurements of the two impellers indicates the non-AC impeller is roughly 7% larger in diameter, but the vanes on the AC impeller are approximately 40% larger than the non-AC. While this is not scientific at all, it suggests what was previously said is at least partially correct (the AC version should flow more). However impeller design has much more going on than this so I'm not claiming anything to be fact by any means. But for the purpose of my curiosity (the reason for asking the question), it is close enough.

As to the impeller speed factor. As discussed earlier the relative pulley diameters (drive ratios) are actually very similar. And the pulley sizes have changed (on both types) several times (although minorly), plus if desired they could be changed for other sizes (from other models/years). Furthermore the impeller speed is continually changing with engine RPM and is subject to many factors (driving habits for example). So for the sake of comparing pump performance we can consider the impeller speed as a constant for both. The main focus here is the internal aspects of the pump (primarily impeller size/design).

As suggested earlier a test rig would provide a better answer, measuring the number of gallons per minute for each pump at a constant speed. But I do not feel like going through the task of build such a rig - for now anyway.

The bottom line is I was wrong in my prior thinking that the smaller (externally) AC pump would have a lesser fluid flow rate than the non-AC pump. If anything it may have a greater rate. ;)
 
Another thing that might be done with the installation of a AC style water pump is pairing it with the larger diameter crank pulley of the non-AC set up. This will rotate the pump at a slightly higher RPM (relative to either stock combination of pulleys). Note that the increase in impeller speed is minimal so there won't be any concern for things like cavatition. And likewise it might not make a lot of difference in pump output. But combining the greater impeller vane area with slightly higher rotation speed may improve cooling circulation function (if the rest of the system allows for it, as Steve has said)....regardless if it needs it.

Easier to use a smaller OD WP pulley, IMO. For me, the pulley sizing was critical in terms of circulation at idle under heat soak conditions (discussed HERE).
 
As a little follow up, I was able to have a great discussion about this water pump comparison with someone that actually knows about impeller design (which I don't). They pointed out that the overall diameter has a relatively larger influence on the flow rate than the vane size. That's something I'd read but did not realize to what extent. In our example the vane size seems to be much greater a difference (increase) between the two than the diameter, however when considering the degree of influence each factor has they may offset one another to some extent. So the difference between the two in terms of flow rates might be fairly minor. This might come down to doing some proper 'scientific' testing. But to do so would not be easy due to several variables that need to be kept constant and the required measurement equipment to be accurate.

I know this has vacillated back and forth somewhat as the information develops, but I still have the overall impression that the "AC" pump (especially when combined with a selection of stock pulley sizes to maximize rotation speed) will offer a slight flow advantage (independent of all other factors).

For me it is enough to make the decision not to switch from my existing AC style pump to a non-AC pump. Previously I thought there might be an advantage to going the other way around, but no longer. Would I convert from a non-AC to a AC pump? I suppose that comes down to a whole bunch of other considerations beyond the scope of this discussion, so I guess "it depends".

Thanks for playing along.
 
Peter, if possible please take a few measurements of the impellers on your old and new pumps before installing it. I believe the needed specifications are as follows (this is a generic impeller image):


I am including 2 pictures with measurements you requested. I am replacing a stock pump that was leaking with the ACDelco one purchased at Rockauto.

ACDELCO WP1.jpg

ACDELCO WP2.jpg


I had no cooling problems prior to this. The old pump has identical measurements, other that some construction differences, to the replacement pump. Both depths from flange to end of impeller are within a mm of each other. The gap between impeller and case was .060" for the old pump and the new. I know this is larger than the ones in the previous discussion but I didn't want to risk breaking the new impeller by pulling it. Interesting to note, the old impeller was threaded onto the shaft, the new one was pressed. It seems my earlier pump design is very different from "lookingforjoe" pictures and measurements of his later pumps. I have the earlier pump with pipes from the case to the heater and carburetor choke. My case also has signs of impeller scraping from a previous pump.
 
I am including 2 pictures with measurements you requested.
Thanks Peter. Those measurements are close to the ones I found, so it helps confirm the design differences compared to the other ("AC") pump. I'm sure various manufacturers will have minor differences in the impeller. I suppose one could even be custom made like the upgraded billet impellers for turbochargers or jet-boats, but that would require some very advanced impeller design knowledge.

That is interesting about the impeller screwing on, I've never noticed one like that before. And yes, there are several different designs of exterior housings; with or without jackets for the heater, carb, etc.. Naturally those differences coincide with other changes elsewhere, to relocate the same connections in various places. Plus things like the mounting boss for the alternator on the AC pumps (which are cast iron instead of aluminum). The very early AC pumps are radically different with a long shaft extending out the back. But as far as I know all of them are interchangeable, at least in so far as they all have the same mount to the block. So if you wished to change something on your engine you could pick the pump that best matched those requirements.
 
The very early AC pumps are radically different with a long shaft extending out the back.
Talk about timing, Dan just posted about these pumps. Apparently they were used on AC equipped X's till '78? Here is what they look like (from Dan's post):

early x ac water pump 2.jpg


You can see what looks like the regular type water pump shape to the right end of them.
 
FYI the Uno Turbo uses the "A/C" style of pump as stock... though this is likely for more axle clearance, as the alternator mounts higher, and the Uno T uses equal length driveshafts, so has an axle mount where the usual alternator mount is.

And yes diameter of the impeller is the thing centrifugal pumps are measured by, well that and the inlet / discharge sizes... the cross section of the impeller cavity is seldom mentioned in pump literature. What is always given is a "pump curve" showing where the pump is most efficient (too fast or too slow efficiency drops)

SteveC
 
Back
Top