[squidtone]

A few years ago I replaced my leaking '67 water pump with a FlowKooler water pump. I was very discouraged to find that the cooling performance with this new pump was very poor when compared to my old original set up. On a couple occasions in creeping traffic the car got up to 235+. The temp also crept up when cruising 60 mph or faster. Due to the "getting too hot" situation with the Flowkooler, I added an AC style shroud and replaced my original fan clutch, but this didn't help.

(By the way, my configuration is a totally stock YS 400 engine in a totally stock '67 GTO.)

Finally this past winter I bought some gaskets and took the FlowKooler pump off to see what was going on.

Firstly, I checked the distance of the rear plate to the impeller and found it to be about 0.140"...which probably did kill off some efficiency of the pump. So yes, naturally, I peened over the rear plate port opening to get it to within about 0.040" of the impeller surface. But the main point of my post is that in addition to the impeller spacing to the rear plate, I also checked the spacing of the (8 bolt style only) center divider plate relative to the impeller vanes. The center plate is supposed to evenly split the volute into two; the rear half of the casing directs pumped coolant to the passenger side of the block, and the front half directs coolant to the driver side. The vanes of the impeller should evenly sit in both halves of the casing. What I discovered is that the center plate/impeller position was biased to almost completely favor one half only. So only one side of the engine block saw any significant coolant flow from the pump.

The solution for me was to make a new divider plate that positioned the divider midway on the height of the vanes.

Conveniently, The optimum plate shape was flat (unlike the original dished divider plate) and easy to make out of 18 gauge sheet metal. The piece was carefully trimmed to fit tightly in the waterpump housing/timing cover. This change in plate shape required closing off a slot in the rear plate. I simply welded a sheet metal patch in place. With the new plate, the impeller now straddled the center divider plate evenly, and I believe equal pumping was restored for both halves of the pump cavity.

My cooling improvements are impressive . The car can sit and idle and it does not heat up. I can travel on the highway and it stays below 195 or so, at least on high 80's temperature days.

Not sure if this center/divider plate check has been mentioned before, so I thought I'd share my findings.

Pictures: The first picture is of the original dished center divider plate on the flowkooler. Note how the vanes are mostly buried in the front half of the pump cavity. The second is of the new flat divider plate which now evenly splits the cavity. The third picture shows the new configuration in the timing cover. The fourth pic is the small patch I welded to the rear plate. And finally the rear plate to impeller distance check.

Dave

[Cardo]

Thanks for sharing Squidtone. I don't have my engine here yet, but that is something i will be taking apart and going through for sure before it goes into the car. This will be helpful for sure.

[My64GTO]

Thanks Squidtone, great informative post!

After years of having a nice cool running car, I have spent the last few years dealing with a hot car after replacing my standard cast impeller water pump and timing cover with a FlowKooler water pump (and new Kaufmann Timing cover).
I had intended to spend this winter taking everything apart to see what was going on, and now I have a good idea what to look for.

Do you think one could peen the center plate so that it is midway between the vanes? Or is the only way with a new plate?

And why did you have to weld the patch to the rear plate?

[george kujanski]

good investigation.

That's one reason i think the 11 bolt design is better, 1 less variable to deal with.

George

[squidtone]

My64GTO,
In my case, the "dish" center plate would have been way too hard to pound down to get the proper fit. The offset had to be corrected by almost 1/4 inch. It was way easier to make a flat plate.
As for the welded block-off I made, I must admit this little "passage" was a head-scratcher. The function of that gap doesn't make much sense...it doesn't look like it was meant to flow coolant, but the fit was sloppy however. I'm wondering if it was simply to help mesh the two plates together. There was a tab in the dished plate to fit, but it was easier for me to just seal that gap by welding a "patch".

And yes, the 11 bolt system does seem so much easier.

[PurelyGTO68]

Fascinating! Thanks for posting. Will definitely check mine and see if I have the same issue. Easy to correct now vs later...

[GT182]

Nice work Dave.

One thing you guys need to remember.... E10 gas is a part of the problem with over heating. Find pump gas without E10 and try it. The alcohol mix burns hotter.

[MidnightAuto]

Squidtone thanks for posting this. I am replacing my water pump with a flowkooler and came upon this post. This is my first time replacing an 8 bolt water pump and the last time i did an 11 bolt was 30 years ago. so I should be classified as a newbie. Anyway, I was planning on doing the mod on the back plate if needed, but your info on the middle divider was great. I just want to make sure I understand - essentially the idea is that that plate should be positioned at 50% of the height of the blue impeller? I took my water pump off and have a pic below. Ive been having some heat problems and perhaps this is part of it. The angle may make it hard to see on the old water pump - but the divider plate is completely flush to the middle impeller. I think you are saying that is bad? This water pump does not have any logos on it, so i dont know if its stock or something else.

On a related note, when i pulled the water pump off I had to pry the middle divider off the pump itself, and it appears to have a gasket in there. I cant seem to find the idea of a water pump gasket anywhere on the web except the once on the outer rim of the pump and the timing chain cover. Are there any other gaskets to insert in the pump/divider sandwich?

[MidnightAuto]

pic

[squidtone]

Hey,
Yes there should be two gaskets. And yep, i believe the middle ate should divide that impeller evenly. I'd dig up some more photos but i'm far from home this week. (Malta) lemme know and i'll respond more next week!
Dave

[MidnightAuto]

Ok got it, thank you. I’ll post how it goes. Safe travels back....

[MidnightAuto]

Well I’ve got an update on where I am at. Painful. So I made the mods to the divider plates and bolted everything back in. I add antifreeze and the pump seems to be leaking at the bottom of the pump housing where the studs and nuts are. I thought I had torqued enough but I guess not. I start to slowly tighten those nuts more, the leak starts to lessen but does not completely go away. The nuts feel way tighter than they should be but I press on carefully and disaster strikes. The housing around the middle nut on the bottom cracks. New flowkooler worthless.

Have others had problems with the bottom of the flowkooler housing sealing? Any advice as I try this again with another pump that is in the way?

[92GTA]

I installed one of these pumps too almost 2 years ago and have fought heat problems since. I didn't make the connection because I also replaced the radiator and went to electric fans at the same time. Now after a $1,600 C&R radiator w/SPAL fans (after 2 other radiator and fan setups) it finally stays at the temperature the fans are set for. Looks like my issue this whole time is the damn water pump. I'm from the GM corporate world where the water pump just goes in and there is no "plate" like these old Pontiacs have so I knew nothing about it.

I planned on going with the Meziere electric water pump but my alternator belt gets in the way so I was going to hold off until I go with a serpentine setup. Looks like I might just pull this FlowKooler POS out and through in a stock replacement water pump!

[dataway]

I recently put a stock water pump in my new build with a clearanced plate, about .030". Looking down into the filler neck of the radiator (on test stand), I can't imagine much more water being forced through the radiator.

Not many people check their pump this way ... but if the water is a few inches below the top of the radiator you can take the cap off and see the flow through the tubes, it should be energetic. I would say a healthy pump will not dribble out, it won't water fall out, it will shoot out of each tube ... not like a nozzle on a garden hose, but more like an opened ended garden hose. It's a hard thing to quantify ... but you want to look in there and be able to say "yeah, I can imagine that is more than enough water to cool my engine"

Also people rarely mention actually feeling the radiator tanks. When the engine is running ... check the temp difference between the hot side tank and cool side ... you should easily be able to feel the difference with the engine running.

If both side are hot, there is probably a problem. Assuming you have enough water flow, then the problem not enough air flow.

If the hot side is hot, and the cool side is quite a bit cooler and the engine still over heats ... probably means enough air flow, but not enough water flow.

This assumes radiator and thermostat are in good working order.

If a person wanted to get scientific they could measure the temperature drop between the tanks with a infrared thermometer. I'm guessing somewhere in the Pontiac factory literature there is a specification for expected temp drop across the radiator at a give rpm/temp range.

[92GTA]

Originally Posted by dataway

Not many people check their pump this way ... but if the water is a few inches below the top of the radiator you can take the cap off and see the flow through the tubes, it should be energetic. I would say a healthy pump will not dribble out, it won't water fall out, it will shoot out of each tube ... not like a nozzle on a garden hose, but more like an opened ended garden hose. It's a hard thing to quantify ... but you want to look in there and be able to say "yeah, I can imagine that is more than enough water to cool my engine".

That is absolutely spot on how my C&R flows, I was really surprised myself about the strong flow of water.

From what I can gather about this thread though in regards to the FlowKooler water pump, is that the divider plate needs to be in the middle of the impeller so the water flow is directed correctly through the block.

[dataway]

On my 8 bolt I ended up tack welding the two plates together after clearancing the front one. I can see how on the design of the FlowKooler impeller that it would be necessary to balance the flow, compared to a stock design that pressurizes the whole chamber with room for the water to seek it's own path to the engine. With the divider plate setup in a FlowKooler it's almost like two separate pumps.

[92GTA]

Originally Posted by dataway

On my 8 bolt I ended up tack welding the two plates together after clearancing the front one. I can see how on the design of the FlowKooler impeller that it would be necessary to balance the flow, compared to a stock design that pressurizes the whole chamber with room for the water to seek it's own path to the engine. With the divider plate setup in a FlowKooler it's almost like two separate pumps.

That makes sense. Maybe I'll take mine apart and do something similar.

[MidnightAuto]

Thread action brought this back up for me. To close out my story, bottom line user error on my part. The middle divider plate was unaligned with bottom and as I tightened pump, it inevitably cracked the pump. So round two with a new flowkooler worked. Clearanced both plates, also added 160t and cold case radiator. When the car was running (before recently blowing a head gasket) all the cooling mods landed me never going above 170, ever. Was great. So I’m good with the flowkooler. Was my manhandling that lead to initial demise on first pump

[92GTA]

So before I go taking out my FlowKooler, I decided to buy an IR temp gun. I drove the car for 45 mins, opened the hood, and compared all of the intake ports at the heads side to side and all of my exhaust ports on the heads side to side.

Side to side, the temps were in the same range. If I REALLY wanted to stretch it, I'd say the driver's side was a TINY bit hotter than the passenger, but by literally a couple degrees.

So I'm guessing I don't have the same problem as the OP with my FlowKooler and divider plate.

Any other checks you guys can think of I could do to ensure my FlowKooler is working as it should?