Interesting Oil Analysis (Blown Motor)

[insite]

for grins, i decided to submit an oil sample from my locked up 2.5L to blackstone for analysis. This is the motor i blew at the track two years ago. i strained the oil through a sieve to get any big chunks of metal out of it. here is the text of their result:

This oil may have been in use a long time, but wear metals don't look any worse off for it. Universal averages show typical wear levels for this type of engine afer about 4,800 miles on the oil. This oil was in use a little longer than that, but wear metals were nice and low, which is a good sign that all is well mechanically in this Boxster. Physically, the oil looks fine. The additive packatge was nice and strong and the trace of fuel isn't a concern. No coolant or other harmful contaminants were present. Air and oil filtration were fine. Nice report at 140,000 miles.

so what does this mean? to me, it means that none of the aftermarket M96 preservation techniques would have done anything to prevent this. the oil didn't suffer any viscous breakdown from high temps. the bearings were not starved for oil. the sump didn't run dry. the motor didn't overheat. there was no d-chunk or coolant intermix.

this failure was sudden and fatigue driven. this supports our initial guess: this was a rod bolt failure (or blackstone didn't actually analyze the oil, but i've used them many times w/ good results & trust their work implicitly). teardown this summer will confirm the failure mode.

[paulv]

Agree that it was a fatigue-related failure. I'm sure that you know, but most people don't know, is that UOAs are beneficial for trend analysis were metals are wearing. Considering that your UOA looked OK, even trending the data wouldn't have predicted the failure (which you already know).

Regards,
paul...

[Jake Raby]

When was this UOA pulled??

Even if a rod bolt broke the collateral damage from that would have created wear metals that were elevated, even if it was sudden.

There are lots of caveats to oil analysis, I have learned these over the time we have been developing our oil, beginning in 2006. One of the most important parts of gaining accurate UOA results is to calibrate the gear with the use of a virgin oil sample from the oil thats being tested. We have no issues with this since our UOA goes back to the company that creates the oil for us.

Blackstone does not develop these engines.. They probably do not know exactly what a good, bad or horrible M96 sample really is. I do see some elevations in this sample from normal, but they are not horrible.

Blackstone has never impressed me.

Looking at that oil, based on some of it's properties it looks like Royal Purple, is that what you were running?

[insite]

the oil was mobil 1 0W-40. the engine blew in march '09. i drained the oil & collected the sample about 2 weeks ago.

blackstone doesn't develop these engines, but i bet they see a LOT of samples from M96's. it would be interesting to find out what they use for their 'universal averages', as i know they choose different values based on the source of the submission sample.

that having been said, if an engine is wearing heavily, i don't think it takes a rocket surgeon to point out that fact w/ an oil analysis. the materials in the M96 are pretty standard, no?. the only oddity might be the lokasil.

if we had overheating of the oil & metal to metal contact as the result of viscous collapse, i'd think we would see some signs of this in the analysis, possibly impairment of the polymer package that allows multi-viscosity performance (lower flash point, reduced viscosity).

d-chunk obviously would have shown intermix.

[chamilun]

does anyone even have any evidence that changing the oil at 5k intervals is beneficial? Ive read the LN engineering articles and there is little proof.

Every oil analysis Ive ever seen (mainly in BMW forums) shows, the synthetics (generally Mobil 1) look as good at 10k as they do at 5k.

To me, the oil is not a culprit in these things going bust.


as an edit, my MY99 was totaled last Friday. Car pulled out right in front of me. Had it not been for the incredible build quality of the Boxster, I'd be toast

[Flavor 987S]

Originally Posted by insite

blackstone doesn't develop these engines, but i bet they see a LOT of samples from M96's. it would be interesting to find out what they use for their 'universal averages', as i know they choose different values based on the source of the submission sample.

Just call Blackstone, and ask them. They will tell you the number in thier database sample size, at the time of the testing.

For example, when I checked last year for my cars, they had about 80 samples for the 987 Boxster and about 40 993 Turbos. Not huge number (I was expecting more) but few people purchase this service, and there were very few 993 Turbos shipped to the USA in '96 & '97 (about 1,800).

[Jake Raby]

I am sure they see quite a few samples, but not as many as one would think.. I just sent out 14 samples last week to our lab, all but one from an M96 engine.

More than likely the engine did break a rod bolt, or it snapped the crank, something happened quick. Pull a sample from it now and send it in.

The older M1 from around 2009 had more moly in it and zinc, thats why i thought this was a modern sample of Royal Purple, I didn't look at the date.

[insite]

yeah, this oil was 2.5yrs old, but the sample was just taken 2wks ago. possible, too, that this gave a lot of particulates a chance to settle & drop out of solution. levels may have been higher had i taken a sample shortly after failure. regardless, i will dissect this thing this summer.

Originally Posted by Jake Raby

More than likely the engine did break a rod bolt, or it snapped the crank, something happened quick. Pull a sample from it now and send it in.

The older M1 from around 2009 had more moly in it and zinc, thats why i thought this was a modern sample of Royal Purple, I didn't look at the date.

[Mike_Yi]

Originally Posted by insite

yeah, this oil was 2.5yrs old, but the sample was just taken 2wks ago. possible, too, that this gave a lot of particulates a chance to settle & drop out of solution. levels may have been higher had i taken a sample shortly after failure. regardless, i will dissect this thing this summer.

Would the metals have settled out of the oil in the last 2.5 years and therefore not shown up in the sample?

[insite]

Originally Posted by Mike_Yi

Would the metals have settled out of the oil in the last 2.5 years and therefore not shown up in the sample?

some would have, but we'd still expect to see elevated metals & other contaminants.

[Flavor 987S]

Insite:

FYI.

Blackstone claims 40 (forty) 2.5L Porsche engine samples in your data. And they throw out all data that's "really messed up from raced engines" so that data is not included in your universal averages.

They are now up to 85 (up from 80 in late 2010) Boxster 987S engines (3.2L) and 50 Turbo engines for the 993.

[insite]

Flav -

thanks; good info. this is certainly statistically significant.