36 Pontiac Cooling System (or "Tales of the DPO")

[Bloo]

One thing still to be figured out if I were to convert the brass tube is why it is longer. While a bunch of stuff changed for 1937, there was no drastic change in block length. These tubes did evolve. In 1936 they added more holes.                                                                                                                                                                    

 

Six holes though, not eight because we are talking about 6 cylinder engines. I believe they are the small holes, and all of these tubes in my possession have 6 small holes.

 

 

They did only supply the later part for service, 497010.

 

 

Here is a wild theory, maybe the extra length is to locate the tube. The 35-36 tube has that little half-moon tab in front to lock it in place. The 37-up funnel shaped tube probably shouldn't be able to get loose anyway because of the taper but could it when a bunch of water hits it? Here is another interesting picture from @1964carlito's 2017 thread about his 1939 Pontiac.

 

 

This went right over my head at the time. If you look closely, it appears his old tube still has the bottom in it. I assumed the tube was shorter because the end had rotted off. I don't think so now. I think his new tube was longer than his old one. It is marked T-3B, and my 1937-54 brass tube is marked WT-3B. His new brass tube is marked 27.5 inches. My brass tube just a hair shorter than 27.5 inches but very close, and I believe we have the same part. The new longer tube must have fit.

 

Now here's this picture again from the same 2017 thread, when @1964carlito was pulling the rotten, and apparently shorter tube out.

 

 

It has the bell mouth as you would expect in 39, or anything 1937 and later, but it also has the half moon locking tab like a 1935-36!

 

I think we can infer that the tubes got longer sometime after 1939, but the new longer tube fits all the way back to 1937 despite the extra length. Maybe it would even fit in 1935-1936 with that extra length if the front of the tube was the right shape. I'll have to poke something in the hole in the block and see how deep it is.

 

[pont35cpe]

Bloo, i have a brass tube for what i thought was a `35 6cyl. It looks just like the brass one you show, same length 27 1/2", same hole layout. The far end, opposite side of seam, is folded in to meet the seam side, no solder. It has a part number 508408

[EmTee]

2 hours ago, Bloo said:

I'll have to poke something in the hole in the block and see how deep it is.

That's exactly what I was thinking.  So, what actually holds the tube?  Are there webs in the block/water jacket casting that the tube passes through?  It seems unlikely to me that Pontiac would rely on the flare at the opening to hold the tube in position longitudinally -- for exactly the reason that you found the JB Weld.  There must be something at the end of the cavity that stops the tube from going too deep...

[Bloo]

On 8/5/2022 at 1:47 PM, EmTee said:

That's exactly what I was thinking.  So, what actually holds the tube?  Are there webs in the block/water jacket casting that the tube passes through?  It seems unlikely to me that Pontiac would rely on the flare at the opening to hold the tube in position longitudinally -- for exactly the reason that you found the JB Weld.  There must be something at the end of the cavity that stops the tube from going too deep...

 

I don't have the whole answer yet, but I think Pontiac (and others who use tubes like this) have holes in the webbing every so often. Others have had difficulty getting the tube out of the hole when it rusts in place. My primary motivation for doing the evaporust thing was to make removal easier. Since it got diluted so early on I'm not sure if it helped, but the tube slid right out. I poked a brass rod down the hole today, and it is about 27-1/2". Laying a 27-1/2" tube on top of the engine, it is about the same length as the engine. These 27-1/2" tubes go all the way in and probably bump against the back, or almost. My suspicion must have been correct about the locking tab. That was only necessary when the tube didn't reach all the way to the back.

 

Edited August 7, 2022 by Bloo (see edit history)

[Bloo]

On 8/5/2022 at 12:28 PM, pont35cpe said:

Bloo, i have a brass tube for what i thought was a `35 6cyl. It looks just like the brass one you show, same length 27 1/2", same hole layout. The far end, opposite side of seam, is folded in to meet the seam side, no solder. It has a part number 508408

My 37 book doesn't know that part number. Does it have the bell mouth or is it rectangular at the front?

 

[pont35cpe]

2 hours ago, Bloo said:

My 37 book doesn't know that part number. Does it have the bell mouth or is it rectangular at the front?

 

It has the bell mouth. The part number is from an attached tag, no other markings.

[Kornkurt]

508708 is 37-54  6 cylinder

[pont35cpe]

5 hours ago, Kornkurt said:

508708 is 37-54  6 cylinder

I looked again, the 4 turned into a 7.

[Frank DuVal]

I fought the water distribution tube on my Graham. Here is the fight and solution:

 

 

 

Lucky the Graham uses a round tube, and a size that matches it is in ACR (HVAC) copper tube, so drill some holes and Bob's your uncle. Way easier to make the tube than getting it out.🤣 On the Graham the tube is held by a "pointy" bolt through the side of the block just rearward of the water pump that engages a hole in the tube.

 

Rock Auto also carries straight radiator hose of the new style, such as Gates Green Stripe. Not AACA correct, but can work. Helpful if your local stores (FLAPS, friendly local auto parts store) do not stock some.

[Bloo]

Did the Graham need a plug in the far end of the tube? Was your old tube good enough to get the hole locations or did you find documentation somewhere? I am now imagining I should probably measure and post the Pontiac hole locations for the day when all the NOS/NORS replacement tubes run out.

 

I must admit I have not yet gone looking for green stripe or other hose at the local stores yet. If I could put on something period correct I probably would. The lower hose doesn't come off easy like the top one does.

[Bloo]

On 8/3/2022 at 6:47 PM, EmTee said:

I'll bet the paint on the timing cover is perfect! 

 

 

I mentioned I'd love to see some original paint. Every restored 1936 I have seen has a green engine like mine. The green on mine though is definitely a repaint. The color could be right, and looks about like the color on most of the other 36s I have seen pictures of. On my car I have found little spots of it on top of sludge in several spots, so the green paint that is there is not from 1936. I found battleship gray on the transmission, lower bellhousing cover, and part of the bellhousing when I overhauled the transmission. Despite my best efforts, I have never found any trace of original paint anywhere on the block, only rust. I'm getting the front all cleaned up now.

 

 

Maybe the color is Chevrolet Orange! Oh wait, no. That's more silicone.

 

The trouble is either there isn't anything there it seems to go away with the sludge. After some careful cleaning I found 2 flecks of green paint that appear to be on the base metal, a first.

 

 

It isn't much though. It might just  be some paint that landed on a bare spot by accident. Continuing to clean the front up, with a little chemical softening and careful scraping down through the layers of sludge....

 

 

More battleship gray! It turns out that it is on that triangular thing too. Could this have been green at one time? I don't know, but it appears to be the same paint I found on the transmission. I had better less contaminated samples of that. It sure doesn't look like it could have ever been green. Maybe the block was green and all the other parts that got bolted on were gray? As it turns out, no. Here the gray is on the head, hiding in plain sight the whole time.

 

 

 

Edited August 10, 2022 by Bloo (see edit history)

[Bloo]

One reason they might have made the water tube longer: sludge. Look at how close those last holes are to the end.

 

 

Mine didn't plug or even threaten to , but there is a ball of sludge in there. I don't know if you can see it in the picture but it is there. Maybe they were trying to prevent the end holes getting plugged by making the tube longer.

 

 

I also wonder if having the wall at the end of the tube that close to the last 2 holes causes them to behave any differently? I doubt it, but still I don't see any downside to the longer tube if it will fit.

 

This is the bottom of the old tube at the front of the block where the blob of JB Weld was.

 

 

That is a fairly significant leak. I imagine the tube still got quite a bit of water, but this explains why the engine was running hotter at the back.

 

 

Edited April 7, 2023 by Bloo (see edit history)

[Frank DuVal]

21 hours ago, Bloo said:

Did the Graham need a plug in the far end of the tube? Was your old tube good enough to get the hole locations or did you find documentation somewhere?

Yes, the far end gets a plug silver soldered in. Yes, the Graham Owners Club International has the dimensions and an article on how to make it.

 

Recording dimensions and dimensions/spacing of holes is a great idea. Someday the spares will be gone, and the old ones in the block will fall apart upon removal. They usually do not come out in one piece from Graham engines.

 

In my case the water pump started leaking, I removed that and the old "while I'm in that far" kicked in. Glad it did, what a mess. I also had to make the back of the water pump plate due to severe rust pitting.

Edited August 11, 2022 by Frank DuVal (see edit history)

[Bloo]

9 minutes ago, Frank DuVal said:

I also had to make the back of the water pump plate due to severe rust pitting.

I've got another thread going about that right now, but don't have enough info to update it yet. I may get some cut in stainless.

[EmTee]

13 hours ago, Bloo said:

Here the gray is on the head, hiding in plain sight the whole time.

So, it sounds like the engine was gray originally...?  Why the doubt?  Is the current consensus that it is supposed to be green?  Does your car have the original engine, or could it be a factory replacement?  I believe late 30s/early 40s Buick factory replacement engines were painted black, as they could be fitted to cars from multiple model years.

[Bloo]

Yes, all the restored ones seem to be green, but I don't know where the information comes from. I am fairly sure my engine is original, and fairly low miles. The block was cast in June, 1936. The transmission, bell, etc had gray paint on them too. There are a whole bunch of mid year changes in 1936 Pontiacs. I believe the factory documentation was lost in a fire. Even if this one was originally gray, and I am convinced it was, it wouldn't surprise me at all if it turns out that a bunch of them were green.

 

 

[Bloo]

I was wondering earlier in the thread why the water tube was made longer. Would having the end of the tube so close to the last holes have a detrimental effect of some kind or even a difference? I decided to blow some water through them and find out. At a lower flow the short one looks like this.

 

 

Long one:

 

 

And at a higher flow, the short one looks like this:

 

 

Long one:

 

 

So yeah, no difference. I do find it odd that the front large holes tend to squirt to the rear more than the rear ones do. The tiny holes at the back squirt more water than the front.  I have no idea why they do, but more water to the rear is undoubtedly a good thing. The hole pattern is completely symmetrical front to rear except for one of holes. The hole size does not vary front to rear. I took some measurements of the hole pattern and will post them at some point.

 

I was wondering about converting that bell-mouthed 1937-1952 brass tube to fit 1935-36. A little cutting and silver soldering could fix the shape. One problem is that even though the hole pattern is exactly the same, the 1937-1952 tube is 1/4" shorter. I figured making an extension would be tough without making the opening smaller. As it turns out though, on 1936 the front of the tube is a really  loose fit. There is plenty of room for an extension.

 

 

Meanwhile, I made a shim for the right motor mount. It sags. On this car Pontiac used two identical motor mounts (group 0.027 number 497628) at the front. 

 

 

The crankshaft is at the center of the car. The camshaft, valves, followers, springs, intake and exhaust manifolds, and all the extra cast iron in the block and head associated with having the valves on the side are all hanging out to the right side. It is no surprise one mount sagged more than the other. I really need to send these out to Steele Rubber for revulcanizing, but that will have to wait for winter. Meanwhile, here is a 3/8" shim.

 

 

 

Edited October 2, 2022 by Bloo (see edit history)

[john hess]

do you have a bore scope to look inside the tube when installed in the block ?   maybe see what's going on at the rear...... just thinkin... 

[Bloo]

No, I don't. I wish I did. I do have a long bendy thing that has a light bulb at the end. That would let me see quite a bit, but I can't find it.

 

 

 

[Bloo]

As I continue to derust, clean up and repair little parts I have been cleaning up the front of the engine. It was really buried in crud that was almost like concrete.. It's looking a little better now.

 

 

I had no old gasket where the water pump back plate meets the water tube and the block, so today I made a template. It is odd that way they put this together. The water tube is that funny trapeezoid shape, the hole in the back plate of the water pump is a rectangle that lines up to nothing in the pump, and frankly doesn't fit very well.

 

 

And the gasket itself? Round holes. Here is a picture of one that comes with a NORS water pump on ebay.

 

 

The big hole goes nowhere. If you look at the first picture in this post, you can see the shadow where it was. The smaller round hole goes over the rectangle and the trapezoid.

 

I might make more than one gasket, so I made a template out of office paper. I started by cutting a round hole the same size as that shadow on the block. I have a plastic thing I use for making electronic drawings that has patterns for drawing circles of various sizes. Then I lined it up to the block and cut the "gasket" using a ball peen hammer.

 

 

That's partly right, but that hole at the lower right and the area around it have nothing to do with the water pump. Using a pump body and a back plate I was able to get the shape of the gasket on the bottom 3 sides

 

 

The rust pitting on the pump's rear plate gave up the size of the other hole. Lined up for least interference, it looks like this:

 

 

And then if it loses about 1/4 inch on the top, because the rusted out back plate suggests the gasket did not quite reach the top of the machined face on the block, it looks like this, just needing to be cut out.

 

 

It fits and matches the ebay picture.

 

With the front of the car all apart, I took the opportunity to remove a couple of useless appendages. The first was an overflow system or maybe just a catch tank that is rusted out and had been long disconnected when I bought the car. I think it is made out of some kind of salvaged can. I left it there figuring I might someday solder a new bottom in it and use it, but I don't think it is needed and I don't think that is ever going to happen. It isn't completely out yet either. The bolts were draggy, so I put more oil on them and I'll get it the rest of the way out in the morning.

 

 

The second was an electric fuel pump. Getting rid of it was instrumental in making the car run right several years ago, but I left it hanging there disconnected, thinking I might need it someday if I couldn't make the mechanical pump work in hot weather. The mechanical pump works fine in hot weather. Later on when I had the toe boards out, I removed the wiring.

 

 

Look at those holes down and to the left.

 

 

That is exactly what it looks like. Someone apparently drilled holes to mount the pump without checking to see if there was any space for the pump where they were drilling.

 

 

And why do the bolts holding it on have two different size heads?

 

 

Because apparently someone drilled it oversize, and then forced a bolt in without cutting any threads.  

 

Edited August 15, 2022 by Bloo (see edit history)

[EmTee]

18 hours ago, Bloo said:

I do find it odd that the front large holes tend to squirt to the rear more than the rear ones do. The tiny holes at the back squirt more water then the front.  I have no idea why they do, but more water to the rear is undoubtedly a good thing.

My explanation for your first observation is the water at the front of the tube is traveling at higher velocity and the water 'leaking' out of all of the holes is following Newton's first law of motion.  The plug at the end forces the water velocity to zero at the end of the tube, so that water just squirts straight out the hole.

 

As you noted, it's interesting that the hole sizes along the tube are constant.  Your garden hose test seems to validate that design decision, as the outflow seems to be fairly even along the length of the tube.  My guess is the holes are small in relation to the volume of water entering the tube.

 

So, are you going to make your gasket with two round holes, like the Ebay one, or are you going to cut yours to match the trapezoid shape of the water tube opening?

[Bloo]

Round. The pattern came out like this.

 

 

I'll probably cut 2 gaskets from this, one for now and one to put with the spare pump. I guess I should have checked with Best Gasket and see if they have it, but it was the weekend, and I sort of doubt it because their 1937-48 gasket is all over Ebay from multiple resellers. The only place you see usually this 1933-36 one is included with NOS/NORS/Reman. water pumps, and more often than not wadded into a ball.

 

There is a fork thing up on top of the grille shell that holds the "silver streak" trim. I thought I had a before picture of it but I don't. It thwarted me when I was trying to take the hood off. Between @aristech's pictures and @pont35cpe's pictures and advice, the hood should have just come off after I unbolted it. It sure seemed like I had missed a bolt or screw somewhere. Nope. It was just solidly rusted in place, stuck for the ages to the hood trim. The fork thing is permanently attached to the shell, so now it soaks in evaporust while I continue to clean up the front of the engine and a million screws.

 

 

 

 

Edited August 15, 2022 by Bloo (see edit history)

[EmTee]

Don't forget to buy plenty of this... 

 

[Bloo]

Out of the bath...

 

 

No more scaly rust on the fork thing...

 

 

A little black paint to keep it from coming back:

 

 

 

[EmTee]

Wow - that came out nice!  How long did it soak?

[Bloo]

2 days in hot weather, and I pulled it out once and rinsed and brushed it off after a day. You can speed the reaction up a little by getting the settled crud out of the way.

[pont35cpe]

Oh yeah!! Cleaning all those nut/bolts, it`s going to go back together, a lot easier than it came apart.

[Bloo]

I am back from the 2022 Pontiac Flathead Reunion, and so it is time to bring this thread up to date. I mentioned that homemade looking coolant overflow tank that looked like it might have once contained motor oil or olive oil or something. I did get it removed before I left (finally) and decided it was completely beyond help.

 

 

I found out on my trip that this is a real Pontiac part. These were highly recommended for anyone running alcohol antifreeze. Odd really, because the whole point of Pontiac's cross flow radiator with the chamber at the top was to condense alcohol antifreeze.

 

Anyhow, it is pretty well shot.

 

 

 

Here's another typical old car problem, radiator mounting brackets broken loose from the tanks. No radiator shops anymore, so it was down to the welding supply to get some killed acid flux. I used a piece of sheet copper to keep the heat away from the core and worked quickly. Looks like it came out OK.

 

 

 

This anti-squeak shim is missing on one side of the radiator shell, tilts it ever so slightly. I'm not sure what to make this out of. It's 1/8". I'm thinking a piece of VCT from the hardware store.

 

 

The front sheetmetal on this car is bolted with nuts in places, but the blind holes have d-shaped nuts riveted into the sheet metal. You can sort of see what they are like in this picture of the pan under the grille. It is broken and missing some metal giving you a better view.

 

 

One was missing.

 

 

A drive across town, and a whole bunch of help from @37_Roadmaster_C resulted in this nicely machined replacement. Now all that's left is to heat it up with a torch and beat the snot out of it with a ball peen hammer.

 

 

Challenge accepted.

 

 

That's much better.

 

 

This was all assembled with some funny pointy bolts. I broke the tips off 2 of them getting the car apart. I might have to make a couple or 3 of these if I can ever figure out for sure how many I need. The heads are undersize, so it's not quite as easy as it looks.

 

 

 

 

[Frank DuVal]

What's vct? Google was no help.

 

Never heard of killed acid flux. Never had a problem soldering those tabs/brackets back in the day. Biggest issue is cleaning between the parts. Next is controlling the heat. Back to the flux, good old Nokorode or such is not on your list because it has Vaseline in it? Works well for tinning car bodies. 

[Bloo]

I've never seen nokorrode work quite as well as plain killed acid, but I have performed some radiator patch miracles long long ago with it before I knew about killed acid. It is great stuff. It has the same active ingredient, zinc chloride.

 

Killed acid recipe: take some pool acid (muratic) and dissolve zinc in it until it won't take any more. it is the classic radiator repair liquid flux. If I had any extra zinc though, I think I would put it in my zinc plating bucket. The liquid flux I bought is Harris "Stay Clean". It has a second metal dissolved in it in addition to zinc, but seems to work about the same.

 

VCT is Vinyl Composite Tile (supermarket floor tile). I bought one today to cut that shim from. It might be great, or it might just break up, I don't know. I'm open to ideas for things to make that anti-squeak shim out of if the VCT won't work. Something that won't hold water would be good. @37_Roadmaster_C used to have some nearly indestructible green gasket material that thick, and years ago he made a gasket for my 1913 Studebaker water pump out of it, but he doesn't have any more. NAPA's thickest doesn't even come close.

 

[EmTee]

9 hours ago, Bloo said:

I don't know. I'm open to ideas for things to make that anti-squeak shim out of if the VCT won't work.

I have some leftover red rubber sheet stock that I bought awhile ago at Ace hardware.  If I recall correctly, the package included pieces that were 1/16" and 1/8" thick.

 

[Bloo]

That looks like it would do nicely, and there is an ACE right down the street. I'll try to cut the VCT this afternoon probably and if it doesn't work, I'll go grab some of that.

 

What was this thread supposed to be about? Oh yeah, water tubes. As you might recall I have a 1937-up brass tube, but it is flared at the front and 1/4" too short for proper hole alignment.

 

 

Surgery time!

 

 

 

I drilled some holes and made some cuts. It is amazing how fast the brass work hardens, about halfway through the cuts the tinsnips stop working right. Another thing I discovered the hard way, I knew I would probably be annealing the brass a bunch of times, but I didn't take this seam seriously enough.

 

 

So I needed to silver braze it like this, which I didn't do initially because I knew I would be heating the brass up a bunch of times to anneal it. The picture you see here with the dot of silver is AFTER the seam fell apart and I spent almost an hour putting 6 or 8 inches of it back together.

 

 

I took some pie shaped pieces of brass out with the snips while forming the tube, annealing it multiple times. If I were doing it again, I would remove less, as it was almost too much that I removed, and I wound up having to straighten the cuts with a thin Dremel cutoff wheel anyway, cutting 2 thicknesses of brass at once.

 

 

A little .015" brass for the extension. I am going to make this have a locking tab like the original did. Since this water tube is longer at the rear, the tab at the front might not be strictly necessary, but the original 1936 tube relied on it, and there is a clearance notch in the front of the block for it. I will put the tab on and that will lock the tube solidly in place.

 

 

(there are more pictures of this job, but I have misplaced them. Probably coming soon).

Edited December 18, 2022 by Bloo (see edit history)

[EmTee]

15 hours ago, Bloo said:

It is amazing how fast the brass work hardens, about halfway through the cuts the tinsnips stop working right.

Well, I can appreciate how a seemingly 'simple' task can get complicated fast!  It looks like you've conquered the worst of it.  In the end, it will be good to have the brass tube in there so that there's no worry about future replacement.

[Bloo]

Agreed. Brass is forever. I have seen postings where Packard guys pulled their brass tube out, washed out the sludge, and put it back in! It was a little extra work to convert this but I am glad I did it.

 

All done!

 

 

Not pretty, but not horrible and definitely functional.

 

 

It would look a little cleaner If I had not let the front of that seam fall apart. If anyone attempts this, I strongly suggest tacking the seam with silver ahead of time, maybe in a couple of spots. Never mind you will be annealing it several times while you reshape the front, and that will remelt the silver. It never tried to move or come apart after I tacked it.

 

 

There is plenty of room for the sleeve, and the tube fits nicely, a little closer than the original.

 

 

 

 

 

Edited October 4, 2022 by Bloo (see edit history)

[Bloo]

Well... That's done (whew). Lets talk about water pumps.....

 

Belt alignment is the next thing to figure out. I have 3 water pumps, and all 3 had the fan flange at slightly different heights. Which is right? One thing I was not able to find while doing research at the AACA Library and the Pontiac-Oakland Museum on my recent trip was the specification for flange height on these 1933-1936 water pumps. I'm about to assemble the first water pump, so I need to know. This is an aluminum dowel tied in the crankshaft pulley with a bungee cord.

 

 

2.334" to the edge of the rod.

 

 

From here I can calculate my way from the water pump mounting face to the center of the balancer pulley, and then from the center of the fan pulley to the front of the flange. I probably did it before I went on the trip, but I am not finding any notes. I may need to re-do the rest of this.

 

[EmTee]

Are you sure that's your car you're working on?  I didn't recognize your engine without all of the silicone, grease and the sexy new paint job! 

 

 

I have the same issue with the water pump on my '38 Buick.  Whoever rebuilt it put the fan flange about 1/16" too close to the pump body.  The fan belt wants to climb up the forward pulley flange.  I'm going to make a shim to fit between the pulley and the pump flange for now.  Next time I have the pump off (hopefully not soon) I'll assess the possibility of correcting its position.

[Bloo]

HAHA that paint job is just on the front. Cleaning all the scrunge off left bare metal, and we can't be promoting rust.... I did consider cleaning up and repainting the whole thing, but what you see here won't even be visible with the car assembled.

 

38 Buick has a "modern" pump with a cartridge bearing, doesn't it? I think you should be able to just support the flange and press the shaft, without actually getting into the pump. I might use some loctite just because I am paranoid about the fan coming off.

 

Edited October 4, 2022 by Bloo (see edit history)

[Bloo]

I finally got around to putting a water pump together. It is the one with the "modern" ball bearing. Remember this?

 

 

 

To me it was the lowest priority of the three, but the others have not come together yet. I got a kit for this online somewhere by matching up the parts I took out to the pictures.

 

 

Yeah, this is exactly the same stuff I took out, except for that brass thing. Oldcarmanualproject answered that. It is a slinger introduced in 1940 meant to keep water away from the bearing if the seal leaks. It's on page 14 of the PDF. http://www.oldcarmanualproject.com/manuals/Olds/1940/Service/40Olds07-Engine.pdf . The Oldsmobile pump is different of course, but after 1938 or so most American car water pumps used a cartridge bearing about like this, and continued to do so for decades. The slinger installs with .025 clearance to the front of the bearing. The impeller here is literally identical to the one I took out. I don't have exact specifications for this pump, so I'm left to guess. Here is what it measured before I took it apart.

 

 

From everything I have read, I suspect this is way too loose everywhere, but the important one is the impeller face. As I understand it, and I only know enough to be dangerous, pump efficiency decreases in a logarithamic way as the clearance increases. If you look at other makes, like Packard, who had a nearly identical water pump and water distribution tube design, .015" is (apparently) about ideal on an old packing pump with curved blades, but a little bit looser might be OK. When Packard went to straight blades and a cartridge bearing for 1937, they specified .040". When Pontiac went to straight blades and a cartridge bearing in mid 1937, they also specified .040" for the new pump design.

 

This measures .067", and it's actually worse because the angle on the front of the impeller doesn't quite match. The Early Times Chapter water pump article says:

 

Quote

<snip>Just because an impeller is new, does not mean that it is any good.  A poor casting is a poor casting, period.  Design tolerances vary, but I prefer to see no more than about .040” maximum between impeller and housing, you can easily check it with a wire-type spark plug gapping tool.  Again, position the impeller as close as possible to the inner housing.<snip>

 

I'm seeing a pattern here. The new impeller is identical to the old one, and it is crap. You can't just fix it by pressing it on further. The seal surface is too high. The angle doesn't quite match. Additionally, it is thinner overall and slightly smaller in diameter. It might be possible to braze on the face of the blades and machine it....

 

When I drove through Iowa on the way back from the Flathead Reunion this year, I stopped at @Kornkurt's place. Among other things, I spotted this:

 

 

Yes that's a NOS GM impeller for this ball-bearing pump! I pressed it on flush with the end of the shaft and the clearance was .039". You can accurately measure it too, because the angle is correct. The rest of the less critical clearances were tighter too. What's not to like?

 

When I last posted I was measuring the distance from the center of the fan belt to the water pump mounting face on the block. I couldn't find a specification. The idea was to calculate how high the pulley mounting flange should be pressed on, because all 3 of the water pumps were at different heights when I took them apart. It's pretty simple math really, but after screwing up and getting it wrong 3 times, and I still don't understand how, I just measured the error and fixed it. The correct measurements go as follows, and assume gasket thicknesses of .015" (pump to back plate) and .030" (back plate to block).

 

The flange height is 4.545" *when assembling a pump*. This is the measurement from the rear face of the water pump casting to the front of the fan-and-pulley mounting flange, without any back plate or gaskets. This is the spec you need because the back plate will be removed when you are pressing things. Again, it assumes the gaskets will be .015" and .030" when they are added.

 

If you measure an assembled pump with back plate, it should be 4.685". (4.545" + .015" + .125" = 4.685"). This includes everything except the back-plate-to-block gasket, which wouldn't be installed yet on an assembled pump.

 

The total distance then from the front of the block to the front of the flange is 4.715"  (4.685 + .030" = 4.715").

 

When I took this one apart, it was 113mm (4.449") from the back of the casting to the front of the flange. It should have been 4.685". Off by .096". Not horrible, but certainly not great.

 

 

So this pump gets put on for now, so I can get the car inside for winter (35 miles away). Meanwhile I am still working on the other two pumps. @Bhigdog made me a repair sleeve for the late 1936 needle bearing pump's body. That is the one I am really trying to restore. It is correct and might even be original to the car. More on that is coming as soon as I can get back to it.

 

Edited November 19, 2022 by Bloo (see edit history)

[Bloo]

I mentioned earlier that it is impossible to get the radiator out with the radiator shell still on the car, despite what the manual says. I suspect that maybe it is possible on a 1935 and the changes were never noted. Maybe. Two important changes occurred between 1935 and 1936 that could have affected that. First, the 1936 hood extends forward over the radiator area. Underneath that, the radiator neck projects up through a hole. You can see the radiator neck and cap sticking through the hole in this internet photo of a 1936 Deluxe Eight.

 

 

The other thing is that the headlights are bolted to the side of the radiator shell. The rear headlamp bracket stud goes through the radiator brackets. The radiator brackets are clearanced in the front-rear direction, but not in the up-down direction. Sooner or later, those headlight brackets have to come off.

 

When replacing a water tube on a 1936, there is just no way you are going to remove the radiator and the silver streak (or grillle) and just pull the tube out the front of the car like you can on many other years of Pontiac.

 

There are 2 possible avenues. One is to remove the whole nose as an assembly, and that is probably the less work, especially on reassembly. The other is to remove the radiator shell from the car, leaving the fenders still mounted.

 

Of note here is the fact that if you are going to take the radiator out too, there is no advantage I can see to taking the whole nose off. You will still have to remove the shell from the nose to get the radiator out.

 

To come out, the radiator has to come up to disengage itself from the two bolts at the bottom. That is assuming you can even get them loose, because the straight-slot heads are completely blocked. It would then have to go down to clear the neck where it goes through the hole in the picture above. The harmonic balancer would need to be removed and out of the way. It is blocked by the headlight brackets. When you remove those, you find out that it is also blocked by the support structure for the fenders.

 

All of that is a waste of time. It doesn't move near enough, and it doesn't come apart. You don't need to take the radiator out to do a water tube. You can either take the nose off, or take the radiator and shell assembly off to get clearance for the water tube. The radiator and shell assembly, once unbolted, slides straight up toward the sky to come out. If you are taking the radiator out also, you might as well just take the radiator shell off of the car, take the shell to the bench, and disassemble it.

 

The headlight brackets had a gasket underneath originally. Not much to work with here.

 

 

A pattern emerges.

 

 

A scan of this pattern next to a ruler exists. I'll edit it in later if I can find it.

 

I wanted to add redundant grounds for the headlights while this was all apart. One reason I never did it was that the wires should probably run between the radiator and the grill and I didn't particularly want them whipping around in the breeze and banging. With the shell off and the radiator out, now is the time. How do you keep it from blowing around? Solid wire! My dad used wire like this to ground electrical panels decades ago.

 

 

The tabs are made of brass sheet.

 

 

If I were doing this again, I would use copper for the tabs and the little tubes to solder wire to at the ends. I already had the brass here so I used it. Silver soldering this was really tough. I was expecting some trouble because the wire was so much heavier than the brackets and tubes, but due to the extreme heat-sinking of the copper it wasn't just tough, but about impossible to get the wire hot enough without melting the brass. Not recommended. If the tabs and tubes were copper, it would still be a little tough due to the disparity in thickness but not so bad. Also if everything was copper it could probably be soldered with phos-copper instead of silver, which is cheaper by a lot. Anyhow it worked. Then I soft-soldered some wire to the tubes at the ends and painted it all black.

 

 

The wires at the right end of the pic were left long because I didn't know how long they would need to be.

 

 

The ends with the ring terminals go to the headlight brackets. Ideally these would go all the way to the bulb sockets, but getting there without drilling unwanted holes is a problem. This isn't perfect, but it is pretty good. At the other end, the first piece of metal these encounter when they pass into the engine compartment is the front engine mounting plate, and that is what they will attach to. On a 36 Pontiac, that plate doubles as the generator mount. There is no better location, as the generator is where the current comes from when you are going down the road.

 

Edited November 27, 2022 by Bloo (see edit history)

[critterpainter]

Just a quick note regarding the green stripe Gates radiator hoses.  Bobs Automobilia DOES use this hose (or at least did as of 4 years ago when I was there last).