Seeking Water Pump Knowledge

Discussion in 'Workshop Forum' started by Dr.Jeff, Aug 13, 2019.

  1. Dr.Jeff

    Dr.Jeff True Classic

    Location:
    Sin City
    That's interesting, thanks for the info. Without such models here we aren't well versed in them. ;)

    I suspect the two styles of pumps/impellers offer very close to the same overall output.
     
  2. gene cooley

    gene cooley Autocrosser

    Location:
    Henrico, Va
    Assuming we are working with 1500cc (+/- ), how much HP are we trying to cool?

    Also, since we are operating almost purely on speculation about the efficiency of these pumps and have very little in hard facts, I'll throw put some more relatively useless speculation: The increased height of the "AC" impeller moves more coolant at low RPM than the non "AC" impeller and that increase? may or may carry through to higher RPM.

    Which brings about another question: How much cooling does a low compression engine operating at off boost or at minimal boost need?
     
  3. Dr.Jeff

    Dr.Jeff True Classic

    Location:
    Sin City
    There very well could be something to this. I recall when I was researching impellers for jet boats they were similar to more traditional props, in that the blade size and pitch has the same effect as the ring and pinion in a car differential. You could get a "tall" impeller for higher top speed but less pull out of the hole, or get a "low" impeller with more pull from a start but less top end. So the water pump impeller with larger vanes could offer more fluid flow at lower RPMs. So I suppose it stands to guess that the opposite would be true at higher RPMs; the smaller vane impeller with a larger diameter may outflow the other one at high rotation.

    As I've stated from the beginning, this entire thread was intended to throw around ideas about these pumps. Certainly more for discussion than scientific. But to answer your "how much power" question: one million. :eek:
     
    gene cooley likes this.
  4. Dr.Jeff

    Dr.Jeff True Classic

    Location:
    Sin City
    The following information relates to the nature of this thread in that it is about water pump performance. But it also relates to some other recent threads about the cooling system:
    https://xwebforums.com/forum/index.php?threads/thermostat-housing-and-coolant-passages.37706/
    https://xwebforums.com/forum/index.php?threads/engine-bay-cooling-options.37681/

    After some "blueprinting" of my coolant passages, T-stat housing, head/block, water pump, etc, I took a closer look at the water pump's impeller fitment in the housing (body). I'm referring to the gap between the distal end of the impeller and the tapered surface at the pump body's inlet. There is a specification for this gap in the factory manual and it has been discussed as an important aspect of pump performance. But frankly I had not addressed it until now.

    On the AC style pump that I am working on I noticed the gap seemed very large. Like 7 or 8 mm too large! This is a new pump from a well recognized maker and the pump housing (body) is also relatively new (less than a couple hundred miles), and using a stock gasket. After checking everything over I put the pump in the hydraulic press to move the impeller further out (closer to the body). It needed to be moved a lot more than I ever would have imagined. But as it got closer to the specification range I noticed something else. The gap is not consistent around the circumference of the impeller. It could be within spec on one side but over double or even triple that on the other side. In addition I noticed some irregularities in the cast iron housing (AC style pump). Mainly large casting pores and excess flash, but also a little erosion from rust. However the largest deficiency was in the actual shape of the casting; the angled surface that the impeller rides next to wasn't level or consistently even around the full circular surface. Yet another issue I also discovered is with the impeller itself. The 'blades' on it are inconsistent in shape/size. This allows some parts of the impeller to be closer to the housing than other parts. So the gap changes as the pump is rotated.

    I used some marking blue (same as used for port matching the head) and adjusted the impeller on the pump to where it just rubbed the housing. Then rotated the impeller to find the high spots. Sanded (ground) them down with a 36 grit Roloc 2" disc and die grinder. Repeated the process several more times until the impeller no longer rubbed, then set the impeller a little closer and did it all again a couple times. Eventually I was able to get the full circumference of the impeller within the specification range for the gap. Although it still isn't perfect; it is toward the high end of the range on one side and the low end on the other.

    I also did a little touch up work on the impeller blades to make them more consistent. But I did not want to remove any more material here than absolutely necessary. So most of the work was done to the housing.

    I had the thought that maybe this pump was a "bad" one. So I checked two others. They were basically the same. I also checked another fairly new pump housing and found the same irregularities in it. So I've come to the conclusion this is typical. However it may not be obvious in most cases. If you only looked at the gap where it is easily visible (through the window on the housing where it passes to the block), then you would never know that other areas of the impeller are still way out of spec. Therefore your pump may appear to be in spec when it really isn't. I still have a non-AC style pump to do the same checking on, and will be curious to see if any of this is different with the aluminum cast housing and small impeller on that style.

    Someone else (sorry I do not recall who it was) suggested that the main reason why some X's have lots of cooling issues while others don't is due to the impeller gap being in or out of spec. After seeing just how bad things can be I tend to agree this may be a major issue for some.

    This also may help to justify the use of a electric water pump as suggested in one of the referenced threads. If the stock pump housings are this poorly cast, and the standard replacement pumps have impellers this far out, then going to a completely different type of pump may be a good answer. Especially when you consider that my 'customized' water pump fitting/grinding will have to be completely redone if a new pump is installed in the same housing.
     
    Conda, kmead, Ulix and 2 others like this.
  5. Conda

    Conda Daily Driver

    Location:
    Cedar Rapids, IA
    When i did my water pump, i just slapped it in there. I remember glancing at the gap and thinking it was kinda large but I didn't know any better (still kinda a newbie to this stuff). I find my pump works, but flow rate is very low at idle which I've seen others say before. Right now normal driving temp is at around 195 on the gauge and gets as high as 215 or so when stopped. Every now and then reving it will give it a shot of colder water from the radiator and reduce temp below 190 but not very often.

    I plan on pulling the pump again (oh joy) and adjusting the gap and seeing if I get more effective cooling results. I think lack of flow is my internal engine heat issue, as the radiator is brand new and fans function just fine. Radiator moves a lot of heat off the car when engine rpm is over 3000, below that cooling is sketchy at times.

    Wont be able to test any of this until i replace my failing brake master cylinder. :(
     
    Dr.Jeff likes this.
  6. Ulix

    Ulix True Classic

    Location:
    Stuttgart, Germany
    I agree.
    My cooling system runs so cold it has trouble reaching operating temp with the electric water pump and a recored stock rad.
    So the rad is NOT the weak link, it is the pump.
     
    myronx19 and Dr.Jeff like this.
  7. Dr.Jeff

    Dr.Jeff True Classic

    Location:
    Sin City
    Adding a little more to this water pump impeller gap thing.

    Today I did the same task on a non-AC style water pump. For the most part it had all of the same issues as I found on the AC style pumps that I did the other day, but not quite as severe. The housing casting (aluminum as opposed to iron for the AC pump) was very porous and the working surface (angled circular face that the impeller rides against) was not precise. That surface was uneven and inconsistent height around its circumference. And the impeller blades were also not very precise. However this pump was not quite as bad as the prior AC pumps were; the aluminum housing had less pitting (erosion), the surface unevenness was a little less severe, and the original gap (before any tweaking) wasn't as large (but still beyond spec). And being aluminum, the face of the housing was very easy to sand down. I did it the same way; adjust the impeller so it actually has a slight interference fit with the housing. Find the high spots (did not have to use the blue marking dye due to the soft aluminum showing obvious scratches where the impeller blade tips touched it), and sand them down. Move the impeller a little tighter and repeat the process until things are pretty even all the way around. I found one shortcut. Rather that try to get an exact fit through many repeated trials, it was easier to get it close (but still a little rubbing) without a gasket between the pump and housing. Then when the gasket is added the clearance was enough to eliminate the rubbing. So overall the non-AC pump process was quicker and easier than the AC ones were, but still very necessary.

    The factory spec for the gap is .030". By making the interfacing surfaces smooth and consistent I was able to achieve a .010" to .012" gap all the way around. According to all of the technical information I learned about impeller pumps (see my earlier posts), the tighter the gap the better the pump's performance. And I believe that is even more evident at lower RPM's where the pump is less efficient overall.

    One more note. Initially before starting the adjust process, I thought it might be a real hassle to move the impeller to achieve the desired movement on its shaft. However by using a hydraulic press it was fairly easy. Once the rig was set up I could gently advance the press to get very small increments of movement. But it was mostly trial and error to achieve the final result. I should say that these are all new pumps, and things might be a lot more difficult if you attempt to adjust a old rusty pump (impeller frozen on the shaft).
     
  8. Dr.Jeff

    Dr.Jeff True Classic

    Location:
    Sin City
    It will be interesting to see what your "before and after" findings are.
     
  9. kmead

    kmead Old enough to know better

    Location:
    Michigan
    So how did you set up the press to do the adjustment? Some pics would be nice...

    Any worries that the run out in the bearing under side load from the belt will cause you to touch on the housing with such a tight set up or am I misunderstanding what is going on?

    Any images of the resultant work? Meaning what does the impeller now look like, what does the housing look like. The images of your water outlet area project were very good and helped to visualize the problem and your solution.
     
  10. Dr.Jeff

    Dr.Jeff True Classic

    Location:
    Sin City
    You are understanding correctly and yes there is such a risk. I won't know if the belt tension (or normal wear) has any effect on it until I can mount everything. But it would be easy to get more clearance if necessary. So I'll start with it tight and go from there. The main reason for making it that tight initially was to enable me to find where the mating surface (on the housing) needed to be leveled out and made concentric to the impeller. Then I decided to leave it there and see how it works. With the constant flow of coolant across it I'm not overly concerned if there is miniscule kissing of the impeller blades, that will self clearance quickly. But more than that would require readjustment.

    As for pictures, it really looks just like a stock pump in every aspect. The surfacing of the housing/impeller-blade interface cannot be seen from looking at it, only by measurement. And all of the equipment that was used has been put away (overcrowded shop requires things to be shuffled around for different tasks). So unfortunately no pics. Basically support the back side of the impeller and press the shaft from the other side. They are a press fit and will move with the pressure of a hydraulic press.
     
  11. kmead

    kmead Old enough to know better

    Location:
    Michigan
    Ok you are excused :)
     
    Dr.Jeff likes this.

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice