Settling rate affects oil analysis

[1 FMF]

Something I inquired about not too long ago, never got a conclusive answer from members here but I've always been somewhat skeptical of oil analysis in general. According to this, high Fe or Cu doesn't necessarily mean high wear, and one should not be quick to conclude oil performance based on wear numbers without knowing all the details of the UOA.

http://www.lube-tips.com/

quote:

---

The time when an oil sample is taken is an important factor in obtaining representative and trendable oil analysis data. The optimal time to take the sample is during the machine's normal operation, because after the shutdown, all the particles begin to settle. The rate of settling is defined by Stoke's Law. Here is an example of how quickly some particles can settle four inches (assuming spherical-shaped particles):

50-micron silica particle: 12 minutes
50-micron steel particle: 2.1 minutes
50-micron copper particle: 48 seconds
50-micron chromium particle: 2.5 minutes

---

 

[jthorner]

I saw that article as well. Very interesting. Yet another big variable visa-vis oil analysis would appear to be how long the vehicle has been off *and* at what point in the waste oil drain process the sample is taken.

It would be interesting to take five samples during the course of a drain and have them all analyzed to see if the results are consistent.

John

 

[TooSlick]

I've taken samples through the dipstick tube using a suction pump and from the oil drain valve without draining that much oil first - perhaps a pint. Results have been very consistent.

The particles that determine wear metal concentrations in oil analysis are < 5 um and don't settle that quickly. I do however recommend taking a sample within 5-10 minutes of shutting off the engine, just to be consistent with the sampling technique.

TS

 

[MolaKule]

The settling time of spherical particles is highly dependent upon the oil's viscosity as per Poiseuille's and Stoke's law.

Most particles in an engine are rarely speherical so the settling time in the original post would be incorrect for a real engine/oil combo.

 

[Bio-T]

I drive to town and back, 18 miles round trip. Then with the engine still running, I drain 1 pint using the Fumoto valve. Then I drain my sample in a sterile glass jar. After I have that I shut the engine off and pour the pint back in. Fill the sample bottle, send in, and no problem. I always take enough for 2 samples in case one is lost in the mail.

 

[1 FMF]

how incorrect?

Like John said above, it would be interesting to take 5 samples during the course of draining the pan, or more even with varying times before starting the drain after the engine is off.

For me, on 5 or 10 UOA's I've ever done all being either car, boat, or truck engine listed in sig, it could be anywhere from 1 to 15 minutes before I start the drain of hot oil from the pan. And catch the sample about 2-3 seconds after starting the drain, but I've never though of being conistent here either until now so basically the sample could be caught anyhwere- out of the 1st quart or the last quart draining out of the pan from a 5 or 10 qt sump. So now, when looking at my UOA trends, I'm skeptical about the little to moderate differences in ppm wear numbers.

From people who are more knowledgeble and experienced, how critical is all this when sampling 20-50 weight motor oil from an engine?

Is there a known variation that can happen, like 2 or 3 ppm of Fe or Cu can be noticed in trending UOA's from altering sampling methods?

Or put another way to clarify my thoughts, if UOA #1 says 10 ppm Fe and the following UOA having had same parameters says 13 ppm, then even though there's an increase in Fe it's not greater than the known variation of 2-3 ppm and the conclusion would be there was no increase in wear (based on Fe ppm)? [** the 2-3 ppm I'm using is a guess soley for this example/question]