Mongo

I know but from a rebuild perspective these nitwits out there don't learn from the mistakes... then again these rebuilds could be coming from the damn Hunan provence across the Pacific where they even put lead in pet food.

worf928

At least it was a plastic impeller.

I've got big box of failed rebuilt pumps. Many have autopac stickers.

GregBBRD

I've seen you offer up this solution, before.

How does one drill the hardened shaft?

I've tried this.....and a carbide drill won't touch it....as a matter of fact, the carbide drill just burns up.

Mongo

Wow. I didn't realize how hard the shaft was. I have drilled through hardened steel (gunsmith) with carbide bits using old motor oil as a coolant with zero failures to my drill bits. I'm curious to know what the hardening specifications are for that shat if it cannot be drilled and pinned.

Cosmo Kramer

Doesn't Ed pin the impellers on the Guardian Pump? I wonder how he drills it?

OBehave2.0

When I used metal impellers I did pin the impeller. I used carbide tipped drill and cutting oil. The trick is to grind the small area of the shaft surface and drill the impeller prior to pressing on impeller. Then use the long carbide drill to penetrate the shaft using the drilled hole in the impeller as a guide.
It is a delicate process and at times became quite expensive as it sometimes required 2-3 drills to do it and at $14.00 a pop, plastic impellers from Porsche became the better alternative, and the more preferred choice among customers.

z driver 88t

Are there discussions on why 928 water pumps are so prone to impeller migration? I'm guessing its because all the cores are so old at this point and have been rebuilt so many times that the interference fit is no longer as tight as it should be? Maybe the shaft has actually worn?

When I came to 928s from the Nissan world all the Nissan guys used standard Autozone or Advance Autoparts rebuilt water pumps on Z cars and 240SXs. Every once in a while one might leak and need replacement, but I never saw the impeller move like so many 928 rebuilds seem to be doing.

Mongo

Place both a Nissan pump and a Porsche pump together; I am curious to see what differences there are aside from the makes the pump is applied to. Perhaps it is not a 'new pump' using an old design we need, but a 'new design' for a 'new pump'.

ROG100

The belt on a 928 apply's a lot more pressure onto the WP pulley. So a poor quality seal and bearing package does not stand up well.

docmirror

The impeller and interior shaft of a water pump is submerged in coolant all the time. The strain of the pulley belt combo on seal would allow coolant to exit the wetted area, and has nothing to do with the shaft to impeller fit. Also, the coolant is always under pressure, so there's no way to get air into the system once it's filled unless it overheats and boils over. Which would cause a complete coolant failure. But - even in that case of coolant vaporization, the coolant in the water pump housing would be near the last place it would vaporize. The interior of the engine will see vaporization first. Maybe that's cooking off the impeller, but by that time, the whole thing is compromised, including likely warped heads, damaged rads, etc.

docmirror

Here's my theory. TL/DR at the bottom of this treatise.

The castings made for injection-molded impellers must provide a press-on fit for the shaft of the molded part. The metal shaft is not rebuilt, and there is clearly going to be a variance in the quality, and diameter of the shafts used for rebuilding. I seriously doubt those shafts are being rejected and replaced during a 'rebuild'. Because the company doing the injection molding does not want to reject a high number of parts, they make the shaft size of the impeller juuuuuuuuuuust a slight bit loose.

Additionally, the quality of injection mold plastics can vary wildly. You have seen this if you put a cheap plastic toy in the sun for very long, it will get soft and melt. The standards for mold plastic can be very rigid($$$$$), or very loose($). So, injection mold companies try to get away with the lowest grade of plastic media for the mission of the part. In this case, the part must endure temps of 260F for many hours, and also allow a low temp of 32F for even longer times without loss of rigidity or swelling, or deformation. That's not a problem for injection mold media, it can be done, but it's more costly to use quality media, so they begin to skimp on the specs, and then that skimpy spec part is mated to a slightly undersize shaft which has been spun once before and - viola! we have a candidate for a failure.

So, for the TL/DR it comes down to cheap companies, using cheap products, and cheap labor to make cheap stuff. Maximize profits. It only hurts when the company that MAKES the product is affected. And that - rarely happens.

edit: Here is a brief discussion of the various media and the resulting qualities. http://www.nicoletplastics.com/resou...aterials-guide

jcorenman

Has Porsche ever used a plastic impeller without a metal bushing?

GregBBRD

No.

GregBBRD

The shaft is part of the bearing cartridge, so it is made to very specific and very tightly controlled tolerances. (Doubtfully, no one here has the tools or a contolled environment test room to accurately measure the differences.)

Because the shaft is an integral part of the bearing, it is always replaced in a "rebuild".

GregBBRD

The problem is quality control on rebuilds (and cheap new pumps).

Impellers and pulleys get larger each time they are pressed off and on.

Cheaply made new parts can't have the tolerances that high quality new parts have.....they would cost more!