Is this bad? How bad? (Water pump)

Changed the water pump and thermostat this weekend. When removing the pump, I noticed some ridges in the surface that mates with the impeller. I haven't been able to find any good info on whether they are machined that way, or have been cut by the impeller in the old pump. The old impeller blades also look like they have been rubbing on those grooves; note the jaggedness on the edges of the blades. Car has 51k miles. The pump I replaced is the original one, judging by the gasket (was still one piece, I had to separate the halves). Thoughts?

worst case scenario: you have plastic impeller bits in your system which could potentially clog lines/passages. fix? flush. DISCLAIMER: I am not a mechanic. Above info was gleaned from this forum, the internet (tech articles) and my indy.

Thanks for the reply, Kromdom. I did think about that, but it almost looks like the blades have been slowly worn rather than broken, and I don't think the pieces would be at all big enough to identify / reassemble. There were definitely some bits in the old coolant that could easily be (or not be) the impeller shavings. I do plan to do a flush soon. I did replace all of the old coolant. I was mainly just curious whether the ridges on the block side are normally there, or are a product of the pump "failure", and whether they could negatively affect the efficiency of the new pump, for example.

I am also in the midst of water pump and t-stat replacement and the bore on mine does not look like that. Sorry.

Is your impeller metal? I don't understand how the composite impeller vanes could impart that much damage. At least it LOOKS like damage.

Quotes source posts: 12723979, 12724021

Interesting. Thanks for the pic; I guess it's definitely damage then. It appears to be the same composite as the OEM pump I replaced it with, but somehow much harder and more brittle. I wonder if some nonstandard coolant could cause hardening of the composite.

Maybe Porsche didn't machine that face properly on your engine block...there are a couple of places that look like the grains of the casting filler on the "damaged" area that I'd think would have been removed if an impeller had somehow machined away material.

Never mind.

It has machined away some of the mating surface, which will reduce coolant flow to some extent. Just keep an eye on the temp gauge.

If I may offer some non-specific advice, never having worked on a 996 water pump problem, but other water pump issues on many other cars. 1. The water flow is from the center to the ports on the outside of the pump. Due to this, I would make a cloth bushing and back flush from the center large supply port to each of those outer edge ports with a hose of modest pressure and flow. This may dislodge any FOD in the engine water jacket. Not all, but some. The same can be done for the radiator, flush from the bottom to the top to remove debris, if any. 2. The water pump scroll flange in the block is clearly damaged radially due to interference. Some interference is acceptable but too much will not allow full flow of the coolant to be driven through the system. There are two ways to resolve this that aren't too costly. Both require some modelers clay. Stick a dab of modelers clay on two of the vanes of a new pump and bolt it to the block with a seal temporarily. let the clay set, then remove the pump careful not to damage the gap defined by the clay between the impeller and the scroll flange. There is a spec avail on that gap but I would guess anything over about 50 thou (0.050") is unsuitable. The fix can be ugly. First one can take a new water pump and have the flange machined on a mill about 80% of the value of the gap. Frex; If the gap were 0.080" one would mill off 0.064" from the water pump bolt flange and just live with the lesser material on the flange so that the pump impeller to scroll gap is suitable. This is messy, costly, and weakens the housing to be sure. Also, the pulley may need to be realigned to match the driven pulley for the system. The other fix is little better. Being very careful to heat the impeller, insure it's not pinned to the shaft, and move the impeller away from the housing the 80% figure of the clay gap. Frex; If that gap were 0.060", then move the impeller out on the shaft by 0.048" to leave the correct gap between the scroll and the impeller. Once completed, drill through the impeller and the shaft together and insert a roll pin to pin the impeller to the shaft and install the pump, checking for non-interference. One of the 928 gurus has actually made a machining tool to fit on the front of the block of a damaged 928 block and machine down the scroll so that a thin shim can be fitted to the scroll area. It could be modified possibly to fit this engine, one would need to do a search on 'water pump block repair' in the 928 sub-forum and see how it goes. After the block is machined, and the shim thickness is selected, it's installed with some screws and loctight and becomes part of the block assembly permanently. There is another way, which requires the pump cartridge to be removed from the pump, and the bore step machined to move the entire assembly toward the engine, however the pulley alignment would need to be modified if this were chosen.

Quotes source post: 12724158

I agree - the machined area just doesn't look as smooth as mine. Maybe some of this "damage" was already there? The clay to measure the gap is a good idea and would give you some data to ponder this further.

The only way for the impeller to interfere with the block is for the bearing to wear, allowing the shaft to deviate. When enough play in the bearing develops for the impeller vanes to make contact with the mating surface, the wear shouldn't be even - the tension of the belt on the water pump pulley will exert force on the pump shaft, with the fulcrum being the point where the shaft passes through the bearing. I would expect wear to be present on the mating surface closer to the tensioner, with less/no wear on the side opposite the belt tensioner. OP's picture shows what appears to be a consistent radial pattern - it seems suspiciously symmetrical to me for something that is purported to be machined by a shaft that should no longer perpendicular to the block face. Maybe one of our resident rebuilders has an opinion?

Quotes source posts: 12724569, 12724597, 12725085

Definitely something to consider. I wish I had left it apart, then I would measure the clearances for sure. However, at the time it only seemed mildly interesting. It wasn't until later that I realized it could be significant wear. (In a hurry to get the job done, I guess) I'm not sure it's worth pulling it again at this point just to measure that clearance, especially if I can't find a spec for acceptable range. I found another thread here https://rennlist.com/forums/996-foru...with-pics.html where the OP's photos show similar rings. It doesn't look like those rings were seen as much of an issue. (Although to be fair, that poster had a full pump failure that the discussion centered around). Looks like both cars are '99s, although I imagine that's purely coincidence.

Is it staying cool? My thought is that if the engine seems to be cooling ok, then I would just let it go and move on since it's already put back together. That's probably not the Porsche politically correct answer though.

I wouldn't worry about the concentric rings. I had similar rings in my early 99 996, and Flat6 Innovations told me that those rings appeared to be from machining, not from the impeller. It doesn't appear that the impeller was chewed up much at all. There is a (very small) chance that some of the plastic bits could be in a dangerous place for the cylinder heads. I was told that the water passages in the 996 cylinder heads are extremely small in some places, and that some of these impeller bits get stuck in the head, creating hot spots that eventually lead to cracks. The cracks allow intermix of coolant and oil, causing the cylinder walls to deteriorate rapidly. It's expensive enough just to repair the cylinder head cracks, but it's practically a total loss if the cylinder walls also have to be rehab'd (it can be fixed, but cost-wise...). Drive on - just watch out for intermix on a regular basis. Monitor the coolant reservoir, check the oil fill cap for coolant residue, and when you change the oil, inspect it for any coolant contamination.

The edges of the impellers look just like the ones of the water pump removed from my Boxster at 172K miles. I didn't see the block as the work was done at the dealer but there was no mention of any possibility of any impeller debris in the cooling system. I want to believe the impellers are a bit "tall" to if not make contact with the block to come darn close so the water pump is as efficient as it can be. A bit of contact would have the impellers worn to fit the contour of the block surface in no time.

KC/Macster/JMJ, thanks for the responses. I feel a bit better about it. I'll watch it, but so far no signs of intermix and cooling is about the same as before.

Hope Porsche machined the rest of the block better than that.

Quotes source post: 12729480

Yeah, it's really surprising to see that's the way it was made, but I guess functionally that's fine. You don't really want water flowing ACROSS that surface anyway, so a rough surface kind of makes sense.

Quotes source post: 12729557

If I may disagree sir, any water flowing across a course medium is prone to either cavitation and/or flow disruption. The lowest head pressure will be maintained with a very smooth surface. I doubt this kind of surface left the factory that way. There is a whole area of study in fluid dynamics concerning surface friction and laminar flow.

Quotes source post: 12724569

This is getting way too technical for me My head is going to blow up

Quotes source post: 12729599

Sorry, Engineers. What are ya gonna do?

Quotes source post: 12729602

[Image-only post; no author-written text. Original image references:] https://rennlist.com/forums/images/smilies/beerchug.gif

Quotes source post: 12729596

Per this (very traditional) water pump design, flow disruption along the outer edge of the impeller is exactly what you want. Any increase in flow along this area is very bad, as it would reduce pressure and flow.

Quotes source post: 12729878

Well, ok. Not what I learned as an intern engineer at Rocketdyne designing and building cryogenic turbine pumps down in SoCal.

Quotes source post: 12729939

You're talking apples and oranges. I don't think you "see in your head" what's actually happening with the flow of coolant. You do NOT want flow across this surface. In a perfect world, the impeller would seal up against this surface (obviously not possible, so it's minimized), that's why the distance between the impeller and the surface is kept at an absolute minimum. It's also why a water pump bearing failure immediately results in interference between the impeller and this surface (and subsequent impeller damage - and in the case of metal impellers, also damage to the block). Many people have experienced damage to this surface (for many years, because this design has been around a long time). This creates a problem because it increases the ability for coolant to flow along this surface, so a lot of effort and money is spent on repairs or modifications to restore the gap to a minimum by either machining the block and fitting a spacer ring in its place, or machining the housing of the water pump to compensate for the amount of material taken off the face of the block.

I'm out, do what you want. Your car. Buh-bye.

Quotes source post: 12729995

Ahh, should have taken note of your profile description of "shameful thread killer". That's exactly what you were trying to accomplish here, uttering complete nonsense based on experience from your "intern" job.

OMG, just drop the freaking engine already. Its probably gonna blow anyway.