GregBBRD

Aside from pistons severely overheated on the tops, some spark plugs that have ***** of aluminum on them, and some crazy white valves, I see no other signs of overheating.

I personally think the heat was confined to the combustion chambers and water was circulating just fine.

I also believe that Mark Anderson has been driving 928's for so long that he looks more at the temperature gauge than the tach.....he's always been really reliable about catching any signs of overheating. And I'm betting that the first thing he did, when the engine hesitated or made the first noise, was to look at the gauges.

FredR

Greg,

My first instinct when you posted the pump pic was to assume that the two events are not linked - and then I thought about it a bit. Because there are no signs of damage associated with coolant flow failure may be the "curved ball". That a failure has occurred on the pump is not in doubt, but no one said the coolant flow was disrupted. When both my plastic impeller kitted pumps failed the temperature on the gauge shot up and then the alarm went off. When both pumps were removed the plastic impeller fell out of the volute completely disconnected from the pump shaft/bearing.. I have never seen a metal impeller type of pump and do not know how the metal impeller is fixed to the shaft- I assume some kind of interference/press fit but...?

What one can reasonably assume in this case is that either the shaft has migrated through the bearing or the impeller has traversed along the shaft- either way the obvious question is "why". That the impeller was able to machine into the volute wall tells us that the impeller was not slipping relative to the shaft or so I would think. That being the case coolant flow would be as expected and no abnormal deviation on the temp gauge or alarm would be indicated- i.e. nothing for Mark to have seen or noted.

The thing that strikes me is- what are the chances of two failures occurring simultaneously completely independent of each other? This would seem to be a case of "double jeopardy" if the engine failure and the pump failure are in no way linked. A possibility but not a probability. In engineering investigations one needs to look at what one sees in front of ones eyes and try to determine what best explains what we see..

I doubt that the pump has seen much in the way of run hours so why did it fail at that particular moment in time? If the impeller slips how long would it take to machine the casing? Are there any signs that might lead one to believe the migration process had been taking place for some time and thus had no connection with events that were transpiring elsewhere in the engine on race day? Are the water pumps designed for racing rpm's? Remember the power absorbed by the pump is proportional to the cube of the speed. A stock S4 trips the limiter at 6.5k rpm, the screenshot you posted earlier [post 140] of the sharktuner settings suggested a redline of 7.56k rpms. Thus if the motor was redlined at that speed the pump shaft would absorb about 60% more power than a stock S4 that is being redlined at 6.5k rpm plus the drag force due to contact- could that be a factor?

Indeed what is the design rpm of your build?- presumably the rev limiter setting [7560 rpm] shown in the ST screenshot is specified by your goodself. If the piston crown temperature was something in excess of 315C [the temperature lube oil starts to coke at] can the piston handle the heat related reduction in UTS & yield stresses? Are such piston metal temperatures beyond those designed for?


Another thought that came to mind was based on a video Louie once sent me of his motor on the dyno. What astonished me was the rate at which it built revs. Given the way that thing accelerated through the rev range I was left wondering whether the motor would need to have the limiter set lower than the design rpm's so that when it was charging the momentum alone does not cause it to "oversheet" as it were. Assuming your motors behave like that is a rev limit of 7560 too optimistic?

Whereas I have no idea what the answers are to those points they represent the kind of questions I might expect to be asked by an investigation team given the info presented.

The truth is out there!

mark kibort

why would the engine revs overshoot? momentum doesnt cause overshoot, power (force, torque) does. as soon as that fuel is cut, the RPM falls, expecially in gear. and by the way, they dyno allows for the engine to accelerate from 4000 to 6500rpm in about 4 seconds. not a crazy fast acceleration rate. once that fuel is digitally cut, there is NO more acceleration of the engine.

GregBBRD

The truth is, indeed, out there.

I'm not sure everyone wants to hear it....or would know it if you slapped them across the face with it.

If you look at the damage on the block, it is very evident that the damage is not "freshly" machined. That area is rusty....and there wasn't a drop of water left in this engine...anywhere. The holes from the broken connecting rod into the water jackets and the residual heat from that terribly high combustion temperatures took care of that!



I hate to hear engines hitting the rev limiter....it sounds awful and I've never thought that it was good for an engine. Besides that, most racing over revs are "mechanical" and no matter what rpm the rev limited shuts off the fuel and ignition, the engine is going to match the "shifting" mistake.

I move the rev limit on my engines up to the point where I'm "comfortable" with the engine going to that rpm with a stuck throttle or completely missed shift. I do not expect the driver to run the car until the rev limiter "hits".....ever. The pieces, in this engine, should be able to turn 7800 rpms....forever. I know that this engine will not still be "falling on its' face" after 7,000 rpms....so I expect that both Joseph and Mark would shift at 6800 or so.

......One of my mentors never put a tachometer or rev limiter in anything he built. When questioned about this, he would say, "This is a race car. If you are behind in a race, you need to go faster. Watching a tachometer isn't going to help. Push on the throttle and go faster."

ptuomov

FredR

Greg,

What are regular Porsche brake discs made of?- grey cast iron as I understand. I live in a warm climate and when I wash my cars the brake discs have the unmistakable orangy hue of ferrous oxide before the car has dried. The hotter the temperature the faster oxidation takes place. What is the pump impeller made of - grey cast iron by any chance? Such items are usually passivated to resist corrosion and the coolant contains chemicals to resist corrosion. Now look at your photos of the pump- the impeller body shows no signs of corrosion but the machined faces do- what a surprise! The crackase contains zero iron so why the discoloration?- invariably as the casing is ground out particles of cast iron will be embedded into the relatively soft alloy and surrounds. The pump, coolant and crankcase are going to be in the region of 100C. The engine grenades, the coolant takes a dump and air rushes in- how long will it take for the exposed particles of wetted unpassivated iron to oxidise at such temps? Trying boiling some water in an unseasoned cast iron pan with the lid on - I bet it will rust up in minutes.

So, just for a moment consider that the above holds and the pump surfaces rusted after engine failure- we know the impeller was grinding away at the casing but is it conceivable that whilst this was going on the impeller suddenly let go of the shaft and reattached itself after cooling down? When my plastic impeller let go one hot June day many moons ago the temperature shot up immediately and dramatically and I was pootling around in traffic- most certainly not doing 150 mph on a hot day on a banked circuit. in the middle of a shed load of loonies. The temperature suddenly races up, the driver is focussed on the task at hand, the temperature on the cylinder wall shoots up, the engine starts to pre-ignite and detonations rapidly follow. The piston metal temperature rises dramatically, the UTS/yield strength drops dramatically- how long would exposure to such scenario take before the motor grenades? Are we talking minutes or maybe even seconds? Every piston is damaged so clearly not a random build error. The fuel was seemingly down some 10 octane points.

Now put oneself in the position of the driver. Fontana is a NASCAR type oval circuit. The race was using part of the main straight, turns 1 and 2 of the banking and then dived into the infield section. Looking at Mark K's videos I would guess that by the time the cars were passing the pits they would be doing close to 150 mph. I have driven at that kind of speed but not on a race circuit and not on a banking with 20 other hot shot pilots around me. At those kind of speeds one's focus is on keeping the car between the barriers- a truly white knuckle ride if ever there was one! From passing the pits to turning into the infield the distance is just short of a mile and if the average speed were 150 mph it would take the cars 22 seconds to get from the pits to the infield turn in. Supposing the water pump failed as the car went past the pits is it reasonable to expect the driver to spot something going pear shaped on the dash panel in such circumstances? I do not know where the motor went twang but for discussion purposes let's assume for the time being this happened in the middle of turn 2. I very much doubt Mark would be looking at gauges- rather he would be more concerned for his life and trying to leave the circuit safely. Once he had come to a halt safely and said a few nice words to the big fellow upstairs, then he might take a glance at the panel [I would be heading for the bathroom to clean my shorts!]. So, could such an event take place and be over and done with in say 30 seconds? Once the coolant is lost and cool air enters the motor how long would it take for the gauges to read normal again?

Remember I am just trying to visualise a scenario that might explain all that has been noted to date..Major failures that do occur [somewhat surprisingly] invariably have a chain of events involving a number of errors with the ultimate failure being the combined effect of such. Here we possibly have the car having been stood still for some time and not fully prepared for the track, old fuel in the tank, fuel 10 points or more down on the octane it was designed to run on, a water pump of questionable provenance and an excellent driver who was not fully familiar with the car or at least not adequately briefed about it..

Whatever the scenario, there will be a credible explanation for it- hopefully you will get to the bottom of it

Rob Edwards

For whatever it's worth, the motor let go about half way up the straight after turn 6. Incidentally nearly the exact place the Zombie motor let go back in April 2012.



It attracted a bunch of attention:

GregBBRD

Look at the block. It's aluminum. That "machining" isn't new.

There's no doubt that the water pump could have contributed to the problem....that's why I posted the pictures.

Without recorded data, everything, except for the fuel, become speculation.

James Bailey

The Fontana track is 75 feet wide with a 15 foot apron or roughly 8 freeway lanes wide . The part on the oval is the easy part They let total novices in rental "NASCAR" race car lap at 155mph in the driving experience after a few minutes of instruction. Admittedly the old very brown 1980 was not doing that speed speedo was pegged at 85 But fun track !! very technical on the infield section.

mark kibort

yes, it is all speculation, BUT, first i would want to know HOW the pump failed or show some evidence that it did fail? grinding away at the block is not evidence. im near certain that the pump was functioning properly . if it failed, the impeller would show evidence, and it didnt. heads would be warped, head gasket would have blown, ... none of this happened.

great theory, but its all based on the pump failing to flow. sure it was grinding the block, but that doesnt show failure of flow

sure this could be a root cause, but there would have to be a pump failure. do you think the pump failed to flow? certainly woudnt stop flowing... and by the way, the car just came off the straight and banking, running near 155mph, so the engine gets some cooling from the air flow.around the engine and particularly accross the radiator. so any flow is cooled substantially.

I agree.. ive only hit the rev limter, (90% ) of the time , FOR YOU GREG. there was a time when you were saying i was easy on the engine, so i hit the rev limiter occasioally to show the engine is running at near redline on all shifts. IRONICALLY, you said you have never watched my video. anyway, i used to always blame many of the blown engines on guys carelessly hitting the rev limiter. doing some quick math, it seems the engine under max stress, having a nasty change in force, might cause undue stress on the engines mechanical components......mainly on the bearings.......still could be an issue, but ive only hit the rev limiter on your behalf about 2-3 times per race. and i always put them on the videos for you .... maybe you can watch one so i wont ever have to do it again!

FredR

Gents,

Each and every investigation into an engineering failure follows a pattern. First step is to gather information, second step is to evaluate the information and the third step is to piece together what may have happened. It should be reasonably obvious that a degree of speculation is involved in this process- if we knew the answers there would be no need for an investigation would there! Once one has a chain of events the investigation team then has to analyse every detail for consistency and credibility..

The "scenario" I have put together is meant to test the gaps in knowledge- what is credible and what [if anything] is not. Given that Mark A is a very experienced hard charger who knows the 928's characteristic behaviour as well as anyone, had he perceived all was not well he would have pulled over immediately but seemingly this was not the case. This tells me that whatever went wrong, it did so rapidly and catastrophically. Had octane rating been the sole issue could the damage we have read about have happened?.If there was something obviously wrong in turns 1 and 2 would Mark have been going round those bends at 160 mph or whatever? Three turns and about 20 seconds later the motor "grenades" somewhat spectacularly by the sound of it. Is that the time frame everything went pear shaped?

It would seem that pre-ignition took place and it would also seem pretty certain that detonation was taking place [that the EZ system could not contain]..Did this fatal damage take place in a matter of seconds and if so,why? Tuomo has opined that if coolant flow was disrupted the combustion chamber temperature would take off. Mark K has opined that the heads would have been warped had this happened but what if the event was all over and done with in a matter of a few seconds?

Greg is of the opinion that the damage to the water pump volute happened some time ago- presumably because it presents a dull finish [?]. If that was the case then on race day by extension, the impeller had to have been running clear of the casing or it would have been "polished"- is this a realistic proposition?

Asking questions is the relatively easy bit- knowing the answers is the difficult part!

FredR

Jim,

I would think the track is that wide to allow for the different trajectories the cars come out at when not spot on the dime upon entry. Most of the crashes seem to happen coming out of the bends- I wonder why?

As a Brit somewhat fond of our version of motor racing, it is not difficult to understand why the infield section is more interesting to drive but no matter what you say the boys who go round those bankings at 160 mph get my attention never mind the NASCAR loonies doing 200 mph or whatever it is they do- huge gonads needed to do that on a daily basis.

mark kibort

155? novice in a NASCA? really? any video of this? but yes, during the straight there is plenty of time to check approaching shift RPM and to check guages. once you hit the #1 marker, you are pretty busy carving around that first turn. ... then , 150mph to a 40mph left turn takes full concentration.

GregBBRD

Did the pictures of the failed pump and the block only show up on my computer? Go back to Post #390 and see if the pictures of the water pump and the block are on your guys computers.....

A quick glance at the water pump tells you that the impeller has moved at least .100". (There should only be a very slight "air gap" between the outer edge of the impeller and the water pump casting.)

The corrosion on the block tells you, instantly that the damage is not current.

FredR

Further to my post No 299 above I was trying to find a bit more info about the "super knock" condition and came across the attached extract.: It is not difficult to visualise that the gudgeon pin somehow exerted a tremendous force on the rod to crack it, the piston bosses counter absorbing the same force and let go first on No 7. The interesting bit is the comment about the engine life if this happens and thus whether something changed to create an environment for this to happen- logically something somewhere getting progressively hotter until onset?

As I was re-reading this thread I then picked up on something I had not really thought about when Mark introduced what had happened to this poor car. When the engine grenaded there was a significant fire presumably ignited by flames emanating from No7 that presumably set fire to the oil mist in the engine and continued to spray fuel into the fire until the motor stopped completely- did this cause the oil on the underside of the piston to scorch rather than heat input from the combustion chamber?