geordi wrote:
Replacing the valves is a good idea b/c while I don't have the official results yet, preliminary answer is that the valves aren't failing b/c of EGR or temperature... It is simple cyclical failure. They are opening and closing something like 3x10^9 times (something like 300 million times) and apparently that is just too many times for the stem to put up with. I wouldn't have thought it would be enough violence on the valve b/c it should just be spring pressure... But that is what the testing has shown at a molecular level.
Criminy Jim! It has to be more than that. 3x10^9 works out to 10-12 thousand highway miles, and that's with no idle time.
(I'm not attacking you and I'm pretty sure we share the same interest here, so hear me out, please)
We know that this isn't the only engine that's suffered valve head separation. There seems to be a couple things that stand out to me, but in CRD's case, I'm only aware of what I've read on the forums.
First would be the frequency of failure. We know the Liberty's popularity was not the highest in 2005/06, and the CRD was a relatively small percent of those produced. Yet, this failure seems to be mentioned with nearly the same frequency as big block chevy's, an engine with many years in production and much higher numbers.
Is this because it really is a much higher percent failure, or because of how surprised people are at the relative expense of the repair?
You can buy and install a whole Reman long block big block for the price of just one one CRD head. Then you potentially have Piston, rings, liner, gaskets, tappets/lifters, timing belt, water pump, thermostat, filters, etc. And if you're not doing all the work yourself, you have a small fortune in labor cost on top of all that.
This tends to make people much more vocal about any said failure.
Second are, (and this is all speculation, not having actual metallurgical data), specific materials, and specific design.
Are they failing due to flaws in materials? I tend to doubt this a bit, because failures due to materials are more like early OHV design failures.
Us old timers are more familiar with cracks/chunks coming loose from valve heads, low quality valve guides wearing their stem holes oval, and such. Improvements in both materials and design (Sodium valves, sintered metal seats, much higher machining accuracy) have made failure rates much lower than in our day.
This R428's valve seems to separate mostly at the stem-head union. To me, this lends more credence to the design theory.
Are the valve springs perhaps a great deal stronger than they need to be, slamming the valves closed with much greater than necessary force?
We already know that this engine was designed when available lubricants and fuels had higher levels of Zinc, Magnesium, and Sulfur, who's purpose was to provide some cushion-type function, for both valve-seat contact, and valve-guide contact.
Does lacking these elements create increased stress during valve closing events?
Was there simply not sufficient fudge factor, (overkill) designed in to compensate for natural statistical variations?
Is the head-stem union too small diameter to handle this shock over billions of events?
I wonder how much available valve guide thickness there is, to create a valve with a larger stem and same head?
Such a solution would add weight to the valve, so would doing this just create different problems, for, say, valve seats?
The more I drive this vehicle, the more I like it, and the CRD is a major part of what I enjoy. If there is some way to significantly lower the chances of catastrophic failure, it's something I'm interested in, and I'm pretty sure you share that feeling. But to do more than take random shots in a dark room, for me, I need data. This is the hope I have in your friend.