When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
WOW, now there is you smoking gun. maybe he didn't see the temps rise. or the light come on? are the lights still enabled ? it just looked like the engine was running fine and then it grenade. if it was pinging, wouldn't it be running funny? maybe just enough flow to keep the heads cool, but not he engine core? could this be the culprit? maybe!
that is ugly.... was this a rebuilt ,or a laso, or a geba??
Originally Posted by GregBBRD
Here's something new....removed the water pump.
Not pretty. The rebuilt 928 International supplied water pump has a severe case of impeller migration....and the steel impeller did a really masterful job of machining the crap out of the front of the block.....perhaps the worst one I've ever seen.
The portable lathe. I love aftermarket water pumps and water pumps with steel impellers. Job security at its best! Everyone needs to use these pieces!
The corresponding volute surface is machined at least .100". Although I have a tool to repair this, I think this block is worn too far for my tool and inserts to repair.
The obvious question....did this cause or be a part of the overheated cracked pistons and cracked rod?
A few things make me think that this is not the direct cause.
1. And perhaps the most important....Mark Anderson has been driving 928's for eons. And he has had them overheat...for eons. My guess is that he watches the water temperature gauge more than the oil pressure gauge. Mark also reported (his is "heresay"....I got zero first hand information about the failure) that the gauges were completely normal as he turned the engine off.
2. Overheated pistons generally "skuff" the skirts and the cylinders. The piston skirts and cylinders, in this engine look fantastic.
3. There is no sign of a compromised head gasket, warped head, or warped block, as is common with an overheated engine caused by a coolant problem.
Of course, could an engine with low octane and elevated water temperature be more prone to pre-ignition and knocking?
Certainly. For the impeller to migrate, it has to spin on the shaft....the two things are absolutely going to occur.
However, that would mean the driver failed to see the water temperature going up.....
is it spinning on the shaft? seems like it would , if the pump impeller would to be acting like a lathe
WOW, now there is you smoking gun. maybe he didn't see the temps rise. or the light come on? are the lights still enabled ? it just looked like the engine was running fine and then it grenade. if it was pinging, wouldn't it be running funny? maybe just enough flow to keep the heads cool, but not he engine core? could this be the culprit? maybe!
that is ugly.... was this a rebuilt ,or a laso, or a geba??
is it spinning on the shaft? seems like it would , if the pump impeller would to be acting like a lathe
928 International rebuilt pump.
Impeller currently tight.
In order to do the "lathe" action, the impeller has to be fairly tight.
In order to do the "lathe" action, the impeller has to be fairly tight.
this means the bearing failed and migrated , or did the impeller move outward on the shaft? usually if it starts to migrate, it stops spinning or spins intermittently. this looks like it was a fully attached impeller.
this means the bearing failed and migrated , or did the impeller move outward on the shaft? usually if it starts to migrate, it stops spinning or spins intermittently. this looks like it was a fully attached impeller.
note to self.. never use rebuilt......
No, it migrated on the shaft. Easy to see. The shaft is now down inside the impeller.
A far better note to self:
Never use a pump with a metal impeller!
When things go wrong (and all water pumps fail), it usually ends up damaging the block.
Porsche didn't change from metal impellers to plastic impellers 30 years ago because the world ran out of the material to make metal impellers!
No, it migrated on the shaft. Easy to see. The shaft is now down inside the impeller.
A far better note to self:
Never use a pump with a metal impeller!
When things go wrong (and all water pumps fail), it usually ends up damaging the block.
Porsche didn't change from metal impellers to plastic impellers 30 years ago because the world ran out of the material to make metal impellers!
Yes, i dont have the experience that you have , but i have quite an experience with these water pumps. plastic porsche OE impellers that just come loose off their inserts. (spins and then the car overheats. and it does it intermittently so its near impossible to nail down), this might be find for street cars, but the high continous forces over time , are too much for that knurling/plastic cast design. then the leaking bearing, thats always fun, then the loose bearing and howling noise, then the entire bearing moving out hitting the block (scots failure). my last faiure , was not catastrophic, but i did see the migration of the impeller on the shaft for the first time and the tiny tiny knick on the block. (saved in time after that overheat incident at t hill a few years ago) so, i have had my share of failures.... fortunately , none hurt my block , and only minor damage to scots block which didnt need a repair.. thats rebuilts, laso, porsche OE, and recently , the Geba. its a crappy little design and too bad its the weak link on our cars!
I see no reason that the metal impeller has to move on the shaft.. thats a terrible press fit. it should be Pinned! in my professional life, this was a common failure of shafts and press fits under load. Pinning them was the answer, in many cases, we could flat the shaft, but that was only for rotational slipping. (which is actuallly a bad thing for our application as we would want the impeller to slip when it started to fail or it would migrate to the block. )
That shaft is super hard and brittle, since it is part of the bearing assembly. Makes pinning or welding problematic.
In my entire 928 "life", I've seen two plastic impellers spin on the knurled piece...and they were both from very "early" (1989) black impeller versions. Since the impellers got superceeded, I've yet to see one fail.
I can live with that failure rate.
I just replaced my 5th "aftermarket" low mileage water pump in a month, yesterday.
That failure rate is good for the bottom line, in my shop, but terrible for the clients.
Yes, i dont have the experience that you have , but i have quite an experience with these water pumps. plastic porsche OE impellers that just come loose off their inserts. (spins and then the car overheats. and it does it intermittently so its near impossible to nail down)
With most other cars, plastic impeller water pumps are tossed in the garbage in favor of metal. They key difference is none of these other cars suffer from a design where the metal impeller will damage the engine block when it fails. My local import repair shop has a box of busted up plastic impellers behind the counter to show customers what happens (not from 928's).
Sadly as 928 owners we need to pick the lesser of two evils:
1. Water pump failure which leads to an overheating engine (plastic)
2. Water pump failure which leads to an overheating engine & destroys the block (metal)
If the plastic impeller fails and you lose an engine because you didn't notice the temps rising, that's on the driver since the same could be true when the metal impeller pump fails.
You say you have "quite the experience" with plastic impeller pumps. Exactly how many have you installed & seen failed? Considering you re-use so many parts as long as they are not broken, I have a difficult time believing you've bought a plethora of water pumps.
With most other cars, plastic impeller water pumps are tossed in the garbage in favor of metal. They key difference is none of these other cars suffer from a design where the metal impeller will damage the engine block when it fails. My local import repair shop has a box of busted up plastic impellers behind the counter to show customers what happens (not from 928's).
Sadly as 928 owners we need to pick the lesser of two evils:
1. Water pump failure which leads to an overheating engine (plastic)
2. Water pump failure which leads to an overheating engine & destroys the block (metal)
If the plastic impeller fails and you lose an engine because you didn't notice the temps rising, that's on the driver since the same could be true when the metal impeller pump fails.
You say you have "quite the experience" with plastic impeller pumps. Exactly how many have you installed & seen failed? Considering you re-use so many parts as long as they are not broken, I have a difficult time believing you've bought a plethora of water pumps.
I have a lot of experience with "water Pumps" not just plastic.. the only plastic water pump i used was the one that failed on the Holbert car (spinning on the shaft, intermittently, until it totally failed at laguna ) I bought quite a few new pumps and only re-used the pump if they were recently changed or felt good. any clicking or noise, movement, i would toss them. (or if they were really old)
to answer your question specifically........... i remember about 6 or 7 , 10 or 11 including the ones for cars i was supporting locally. (scots, gregory's 6 liter, my old '79 spare racer beater, and Petty's)
Ill have to check my records for the actual number, but based on memory:
Failures from new:
1. the plastic impeller OE,
2. the initial WP in holber car OE, leaked on installation, leaking one after a few seasons, then
3. replacement laso that we think caused the overheat and made lots of noise, and now
4. currrently on 3rd or 4th season of the Geba. There were a few before that. especially on my older 928 '84. there was one that i remember making a lot of noise and the impeller was almost ready to fall out of the housing.
. and i seemed to remember an old one just changed at a timing belt change for new , 6.
I have a log book.....ill have to check to get an accurate total.
I find it hard to imagine how anyone would be looking at the dash panel when driving round a 150 mph banking in the middle of a bunch of adrenalin junkies whose sole mission is to cream the chap in front of him!
Given it was a hot day if the impeller decides to take a walk how long will it take to seriously over heat with the motor spinning at or close to max rpms? That Greg says there is no sign of heat damage in the piston skirt area is somewhat baffling if the water pump failure was the incident initiator given one might expect the cylinder walls to toast first.
I have had two plastic impeller failures- one on the original pump and the other on a Devek rebuilt pump- ironically both failed at about the same mileage and the same age [80k km and 10 years]. Both case the bearing and seal were perfect - if it happens again, now that I do my own TB/WP jobs I might even try to stick the impeller back on just to test how it fares. If NASA can stick tiles on the Space Shuttle one wonders if there is a high tech adhesive that can support this duty.
I find it hard to imagine how anyone would be looking at the dash panel when driving round a 150 mph banking in the middle of a bunch of adrenalin junkies whose sole mission is to cream the chap in front of him!
Given it was a hot day if the impeller decides to take a walk how long will it take to seriously over heat with the motor spinning at or close to max rpms? That Greg says there is no sign of heat damage in the piston skirt area is somewhat baffling if the water pump failure was the incident initiator given one might expect the cylinder walls to toast first.
I have had two plastic impeller failures- one on the original pump and the other on a Devek rebuilt pump- ironically both failed at about the same mileage and the same age [80k km and 10 years]. Both case the bearing and seal were perfect - if it happens again, now that I do my own TB/WP jobs I might even try to stick the impeller back on just to test how it fares. If NASA can stick tiles on the Space Shuttle one wonders if there is a high tech adhesive that can support this duty.
Yep, in the banking , you are not able to really look , unless you plan early. even then it is nerve racking to take your eyes off the action. however during the straights, especially long straights, its pretty much routine. however, if you dont catch it, the red warning light will grab your attention, as im always looking at the tach for near redline shifts and a red light would catch my attention. was his red warning light still working? if so, and it didnt go off, that wasnt the cause of failure, however, maybe it started detonating due ot the extra heat . the heads could have been VERY hot even without the warning light coming on. the entire thing is still very puzzling!
Yep, in the banking , you are not able to really look , unless you plan early. even then it is nerve racking to take your eyes off the action. however during the straights, especially long straights, its pretty much routine. however, if you dont catch it, the red warning light will grab your attention, as im always looking at the tach for near redline shifts and a red light would catch my attention. was his red warning light still working? if so, and it didnt go off, that wasnt the cause of failure, however, maybe it started detonating due ot the extra heat . the heads could have been VERY hot even without the warning light coming on. the entire thing is still very puzzling!
No cylinder wall or piston skirt damage virtually eliminates the water pump as a point of failure.
The water pump is probably just an interesting data point.
No cylinder wall or piston skirt damage virtually eliminates the water pump as a point of failure.
The water pump is probably just an interesting data point.
im thinking the same... the clearance is in the negative numbers, so the pump is pumping (along with a little aluminum) . the friction would create some heat too! but its a non issue. I wonder why that couldnt be heard? was the pump bearing loose axially, so maybe it only contacted at high flow and pressure levels? anyway, that is some point of interest.
I agree the pump had nothing to do with the failure.
If anything, the super-tight clearances between the impeller and block made water circulation even more efficient, and why no scuffing on the piston skirts.
On this occasion maybe I should be the elephant in the room!
A pump failure can happen any time and so can an engine failure. When a water pump failure occurs it is probably all over in a matter of seconds- the impeller migrates on the shaft but it is tight enough to mill the volute by 0.1 inches or whatever. If a water pump typically fails every 60k miles and we assume the car runs at an average of 60 mph that is a failure every run 1000 hours. Now how long does a racing engine last? - for purposes of discussion let's assume 100 hours. If we assume the water pump can fail in a matter of 60 [?] seconds the probability of failure is 1000x3600/60. If we assume failure of the engine can take place in a matter of a few [6?] seconds then the probability of failure is 100 x 3600/6. Multiply the two numbers together and we get a probability of the two occurring at the same time of one in 3.6 billion.
Still think the two events are not connected? Taking the blinkers off and looking at the big picture is there anything that could reasonably explain why a water pump and an engine should grenade at the same time?
On this occasion maybe I should be the elephant in the room!
A pump failure can happen any time and so can an engine failure. When a water pump failure occurs it is probably all over in a matter of seconds- the impeller migrates on the shaft but it is tight enough to mill the volute by 0.1 inches or whatever. If a water pump typically fails every 60k miles and we assume the car runs at an average of 60 mph that is a failure every run 1000 hours. Now how long does a racing engine last? - for purposes of discussion let's assume 100 hours. If we assume the water pump can fail in a matter of 60 [?] seconds the probability of failure is 1000x3600/60. If we assume failure of the engine can take place in a matter of a few [6?] seconds then the probability of failure is 100 x 3600/6. Multiply the two numbers together and we get a probability of the two occurring at the same time of one in 3.6 billion.
Still think the two events are not connected? Taking the blinkers off and looking at the big picture is there anything that could reasonably explain why a water pump and an engine should grenade at the same time?
I see where you are going on this, but this pump looks like it was going bad for some time.. (relative to the race engine use) and it didnt really fail functionally.. as swaybar says, with the gap so low, it was more than functional , unless it was slippping rotatioally, which is unlkely. that coupled wth the fact that even with catastrophic failure, the water has to heat up which happens pretty quick. the water warning light would have lit up, but more likely, this kind failure has the pump working and being effective in cooling the engine.
08-01-2018, 07:55 PM
