Seeking Water Pump Knowledge

Dr.Jeff

True Classic
Has there ever been a complete primer on all of the various water pumps that will fit the SOHC? I never realized how many different variations there are for what seems like a straightforward item.

Different housings, pumps, pulleys, heater attachments, coolant return tubes, and so much more. Basically, for the majority of US X1/9s (aside from the '74), I thought there were two housings - one for AC and one without. And one pump type for each housing. It seemed all AC pumps had a bolt on pulley and non-AC had a press on pulley. But I'm finding there is so much more than that.

I'd like to get an overview of all the various combinations that were (are) available. But some specific questions are as follows.

Are all housings interchangeable solely with regard to the block (not accounting for other accessories, options, etc)?

What pulley diameters were used? And for which applications? Are there different offsets of pulleys (aside from one and two belt grooves)?

Are there differences in impellers and/or fluid volume - output capacities?

I have several ideas but need to find out more about what can be interchanged with what first. Thanks for any input.
 
I believe the coolant return port applied to all 1300s. The return pipe for the 74 only had one fitting for the heater return. I believe 75 - 78 also had the small return tube on the fitting for the water choke. My 1500 water pump looks identical to my 74 except that there is a very thin casting blocking the return port and the two holes to secure the return fitting are cast in but not tapped. The impellers and pulleys look the same.
 
I've been looking closer at the water pumps that I have. They are for the 1500, with and without AC. A few odd things were discovered:

1) The impellers are different. I've never had them exposed at the same time for a side by side comparison. The AC style impeller is about 50% thicker (blades reach deeper), but also about 5% smaller diameter. AC = 15mm deep and 60mm diameter, non-AC = 10mm deep and 63mm diameter. I assume that gives the AC impeller more fluid pumping volume/flow? This will need some more technical explanation to fully understand.

2) Naturally the pump housings are also different size, as can easily be seen externally the AC pump is smaller. But the dimensions inside are what count. The coolant return pipes are the same diameter, but the throat between where that pipe connects and the input to the impeller is slightly larger on the non-AC pump (about 3mm, which corresponds to the differences in impeller outside diameters). The inner cavity of the housing at the widest point is 75mm on the AC pump and 85mm on the non-AC pump. What effect does the size of this chamber have on the pumps performance?

3) The drive and driven pulleys are different diameters, as easily seen. But there is inconsistency between various years for them. Depending on the year, the AC pulleys are around 117 to 123mm (total OD, not belt contact area) on the crank, and 103 to 115mm on the pump. The non-AC pulleys are 130mm crank and 115mm pump (I only have one of these to measure so not sure how much variance they have). I would have to do a pulley size calculation to see how these diameters effect rotation speeds between the AC and non-AC versions. Also I understand there are several other pulleys possible for various applications.

As before, there measurements are not exact but close enough to get an idea of things. I'm hoping someone with knowledge of hydro/fluid dynamics can help explain what these difference actually mean in terms of pump performance (ultimately, the cooling system function). However I have a suspicion that I was completely wrong in my thoughts regarding the AC pumps. Judging by the smaller outside housings I thought they would provide less circulation, which a car with AC does not need. But they might in fact offer more circulation? This might be something to help with cooling on performance (non-AC) engines.
 
I can't help with the pump evaluation, but I think it won't be so simple to evaluate pump flow based on the available info.
Therefore I would recommend you just test it!
Build a simple test rig outside of the car and drive the pump at a known speed and measure volume pumped oer time.

On another note, finding the pump with the best flow is definitely worthwhile.
As I mentioned before, we have installed (oversized) electric water pumps on two high-powered X1/9s and the coolant temps are now too low!
It is hard to get the temp past 140°F with normal driving and it normally stays between 140°F and 160°F when pushing the engine.
Both cars have stock rads recored at a rad shop, so nothing special.
This shows that increasing flow makes a HUGE difference.
No fancy radiator or second fan required.
 
AC cars do have a different pulley than non-AC cars as the drive belt layout is totally different. The housing that bolts to the block is also different.
There are at least two different steel pipes that go from the pump to the thermostat, one with a heater hose neck and one without. Probably a third one for a water choke take-off but I haven't seen any since all my Xs were either injected or early cable operated choke. The basic configuration of the pipe itself was the same.

Maybe Matt can shed some light on flow as he races Xs. If you are looking for the best flowing water pump for Xs, I'll assume electric is probably the way to go.
 
I am in the process of replacing my pump since the seal is weeping. I was wondering what specifications to use to set the impeller depth in the housing. How much space between the end of the impeller and the back of the housing? Can I adjust using thick/thin gasket on the pump end to housing? Then, the belt pulley should be inline with the alternator and the crank. It looks like the only way to adjust that is move the pressed on pulley. I have a '76 stock 1300 and it has the heater/carb tube out the top of the housing. I bought the AC/DELCO pump from Rockauto with the cast impeller, no housing and a pressed on pulley like mine has. I plan to measure the old one and use that as a template but I have no idea what impeller is in the old one until I get it out. I have seen the cast impeller, a plastic impeller and what appears to be a sheet metal one.
 
The quick and dirty way is to leave out the gasket (use sealer instead) and see if it still turns. It usually does.
Much better would be to remove some material on a mill and then use a gasket of the right thickness to minimize clearance.
 
I know we have gone over this many times, but if the pump is in good shape, the radiator is in good shape and your hoses are not kinked or something else, there should be no reason your X will overheat. I daily drove an 81 in Phoenix for two years, middle of summer and the famous 122 degree day in 92. No problems. Even my 14:! compression race car had no issues, and in that I was running only water and pump lube.

All I am trying to say is I hope you don't spend a bunch of time trying to fix a symptom when the cause still exists.
 
As I mentioned before, we have installed (oversized) electric water pumps on two high-powered X1/9s and the coolant temps are now too low!
It is hard to get the temp past 140°F with normal driving and it normally stays between 140°F and 160°F when pushing the engine.
Both cars have stock rads recored at a rad shop, so nothing special.
This shows that increasing flow makes a HUGE difference.
Is there any special about the thermostats you’re using? Naively, I’d expect that the thermostat would restrict flow to the radiator enough to allow proper warmup.
 
I am in the process of replacing my pump since the seal is weeping. I was wondering what specifications to use to set the impeller depth in the housing. How much space between the end of the impeller and the back of the housing? Can I adjust using thick/thin gasket on the pump end to housing? Then, the belt pulley should be inline with the alternator and the crank. It looks like the only way to adjust that is move the pressed on pulley. I have a '76 stock 1300 and it has the heater/carb tube out the top of the housing. I bought the AC/DELCO pump from Rockauto with the cast impeller, no housing and a pressed on pulley like mine has. I plan to measure the old one and use that as a template but I have no idea what impeller is in the old one until I get it out. I have seen the cast impeller, a plastic impeller and what appears to be a sheet metal one.
If you haven't seen this yet, take a look, skip ahead to the posts with the impeller clearance pix: https://www.xwebforums.com/forum/index.php?threads/running-too-warm.22623/#post-184709
 
We do seem to go around and around on this cooling issue. And I have yet to read a thread on this topic that conclusively "found" a universal design problem for the X cooling system.
 
Thanks to Dan for the link. It looks like 0.030" is the factory clearance for the impeller. Why did you go to 0.016"? Is there any concern that due to heat expansion or future bearing wear the impeller would contact the housing?
 
I suppose I should add a little more insight to help alleviate some of the unrelated comments. This is about finding which FACTORY pump has the GREATEST flow/circulation, regardless if YOUR car needs it or not. Some of us have special projects that DO require it, despite the perfect engineering of the stock original X1/9 and its cooling system. ;) Thanks for playing along. :)

I realized there isn't enough data given yet to evaluate the impellers I found. I will get some additional specifications for them in the hopes someone with the needed engineering knowledge can answer the question.



I bought the AC/DELCO pump from Rockauto with the cast impeller, no housing and a pressed on pulley like mine has. I plan to measure the old one and use that as a template but I have no idea what impeller is in the old one until I get it out.
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):
BM-water-pump-56812_t2.jpg

1) Overall outside diameter of the impeller portion (red arrows).
2) Diameter of the input area (yellow arrows).
3) Height of the impeller blades at the outside (blue arrows).
4) And at the inside (white arrows).
5) The number of blades.
6) Maybe the length of the blades if possible (not indicated on the diagram).

There may be more that's required. Hopefully one of our engineers can tell us exactly what is needed. But as much of this as you can get will be greatly appreciated.

A absolute value for the impellers is not needed, only some inclination of which one should offer the most pumping action between the available choices of stock impellers. So I believe the measurements do not need to be exact for this purpose.
 
Thanks to Dan for the link. It looks like 0.030" is the factory clearance for the impeller. Why did you go to 0.016"? Is there any concern that due to heat expansion or future bearing wear the impeller would contact the housing?
As I recall the linked discussion about the impeller gap includes this info, but I will say it again just in case. You should be able to move the position of the impeller on the pump's shaft to adjust the clearance. Some may be more difficult to move than others, and be careful not to damage any of the components. While playing around with a used pump, the cast iron parts broke when I applied too much pressure. I believe the tighter the clearance the more efficient the pump. But I'm sure too close will cause issues with interference, etc. Using gaskets to change the position of the impeller also changes the position of the belt pulley, as you noted. However I've found most of the new pumps I've purchased had the belt pulley in the wrong position anyway. So you might have to adjust that either way. Because I have a AC application, the pulleys are not pressed on but bolted to a flange on the end of the pump's shaft. So I have not tried to move the pulley on the pressed on style that you have.
 
Jeff, how good is your math?

what you are wanting to work out from your pictures above is known as a centrifugal pumps "velocity triangle" google it and be prepared for some math.

next you will need to determine the system flow losses, that can get pretty tricky to do theoretically.

real world measurement of pump flow and pressure is needed to determine the pumps efficiency.

In centrifugal pumps one thing that has a marked effect on pump efficiency is what is known as "neck ring clearance" ... the impellers on our sohc water pumps don't really have a true "neck ring" (inlet clearance) and the closest is the impeller to body clearance, the closer the better (until it touches) is most efficient

I'm not a fluid engineer, but my full time job for almost 15 years was pumping large volumes of water in the mining and resources industry, so I have a very good working knowledge of how it works, plus I have what's known as a "Certificate 4" in pneumatics and hydraulics which puts me at about second year of a three year engineering degree for those disciplines.

IMO the pump generally won't be the problem with any pumping volume issues, it will come down to increased resistance to flow from some other restriction being the limiting factor, but you're welcome to go thru all the math to determine this for yourself.... with fluid pumping (talking about 1.0 specific gravity water) if the rate of flow exceeds 2.4metres per second thru whatever sized pipe / conduit you have, the resistance to flow goes exponential on the scale past this value... so simply thinking you can pump more volume thru the same sized pipe (i.e. by speeding up the pump for example) doesn't always work ... as wherever in the system has the smallest cross sectional area will be the limit to the systems flow restriction.

SteveC
 
Is there any special about the thermostats you’re using? Naively, I’d expect that the thermostat would restrict flow to the radiator enough to allow proper warmup.

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.
 
This is the stock AC WP impeller from my '87. I cannot affirm it is the original from new unit, however the PO said he never changed it. When I first replaced the WP about 10 years ago, the one I got with the whole housing (before I knew you could just do the pump & manually adjust the impeller clearance) has the nasty plastic impeller. I say nasty cos it doesn't look like it was effiecient at all; fat blades & different contour. I removed at at some point & used a GMB pump that had the cast iron impeller that matched the form of the original I had removed (since I tend not to throw things out, I still had it then)

IMG_2949.jpg


You can see the wide gap of the plastic impeller. I wouldn't mess with pressing that back & forth on the shaft to adjust, just as likely to stress crack the impeller.
IMG-2930.jpg


GMB-supplied replacement (doesn't come with gasket, the one shown was one I incorrectly ordered) - used at this point, I transferred it to the 1600.

IMG_3007.jpg


Increasing the rad core dimensions & slightly reducing the pulley OD, running Evans waterless, 1600cc, raised comp., with an idle of 850rpm & 12ºBTDC gave me a coolant temp that rarely hit the 190º mark. Cruising, the coolant temp stays around 175-180º (less outside of the NY summer months). High load situations would bring temp closer to normal op. For me heat soak was the factor that would push the engine temp up to and/or slightly over normal operating temp, so I added the bay fan to alleviate that. This past year I didn't even have to look at the gauge, which was a change from the past.
 
Dr Jeff,
I will make the measurements you suggest along with the impeller to housing gap on the old pump. I did not have overheating problems but my pump of over 30 year age was leaking. I am out of town till next week so won't be working on this till then. Thanks to all for all the excellent information on this issue.;)
 
what you are wanting to work out from your pictures above is known as a centrifugal pumps "velocity triangle" google it and be prepared for some math. Next you will need to determine the system flow losses, that can get pretty tricky to do theoretically. Real world measurement of pump flow and pressure is needed to determine the pumps efficiency.
This is exactly why I posted another thread asking engineers to please look at the information in this thread. I realize from prior experiences that this is beyond my level of understanding, so I need input from those that can digest this and offer a better understanding of it.

However (hopefully) it may not be necessary to do all the math and compute the velocity triangle in this case. I really only want to get a feel how the two basic impellers (original Fiat AC vs non-AC) compare in general terms - i.e. which one would move more fluid at a given RPM. Just looking at them I think the AC impeller does, but I know enough to not bank on casual observations and what I think.


it will come down to increased resistance to flow from some other restriction being the limiting factor
Excellent point, makes perfect sense. For now I'm going with the following assumptions:
1) Other aspects of the cooling system are being modified, which will likely change this to some degree.
2) It would be the same for both styles of impellers, so it is something of a constant factor having the same influence on both impeller options.
3) The factory used both impellers, so it might be fair to assume the total system is capable of handling either impeller effectively. In other words, the point of greatest resistance is sufficient for either impeller. So choosing the better (larger flow) of the two is best.


Perhaps one way to describe the goal here might be to find which combination of stock components (pump housing, pump impeller, pulley sizes, etc) will maximize the cooling system flow. Earlier Ulix stated his experience with using a electric pump made a significant difference -
finding the pump with the best flow is definitely worthwhile
Based on his findings with a much larger pump, it seems the remainder of the cooling system is sufficient to work with a greater pump volume/flow, providing a cooling benefit.
 
This is the stock AC WP impeller from my '87.
Thanks Huss. I did not say it earlier, but all of the impellers I have are the cast iron type like you pictured. I believe one of mine to be a official Fiat factory item, as it was in a Fiat parts box and has all of the Fiat markings on it. The other two may not be genuine factory items, but I'm assuming them to be the same design as original ones (which is why I'm asking others to offer up any dimensions they can to help identify variations that may exist).

I have never been a fan of the stamped steel (flat blade) impellers, nor the plastic ones. Granted plastic will not corrode like cast iron, but I've seen them with broken blades that were in worse condition than rusty old iron ones.

The cast iron ones you posted look like my AC type impellers. Judging from the gasket difference, I'd say yours is a AC style pump, so that makes sense. I only have one non-AC pump, and the impeller on it is significantly different in dimensions. So I'd like to do two things; verify it is the same as all (most) non-AC impellers, and determine how different it acts compared to the AC design.

As I stated before, I previously had the impression the AC pump would provide a lesser flow than the non-AC pump. That was based simply by looking at the outside of the housings with the AC style being much smaller. However I might have been wrong about that. After looking closer to the internal aspects, it might be the other way around (AC type flows more)?
 
Back
Top