Why so much variation on oil analysis?

[peter-h]

I have seen big variations.

For example nickel has gone down from 2.5ppm to 0.9mm and that is from a 29hr sample to a 46hr sample.

Obviously one should not complain but what could be the physical explanation?

Iron has also gone down by 1.5x, copper is inline with the oil age, chromium has gone from 5.5 to 3.2.

Assuming the metals get into the oil mainly from bearing / sliding surface wear (which is not an unreasonable assumption!) why do they vary so much?

 

[ScottM]

Are these decreases after an engine maintenance event?

I noticed that when I did the top overhaul to my engine that copper was real high for the first several oil changes, but settled into a relatively flat amount after 100 hours. I was told that this was due to copper coming off of the new parts and is normal.

 

[Clark1961]

Sampling technique and dilution will have major effects on the oil sample quality. Basically, if the oil is not sampled exactly the same way each time then the results will be very different. The dilution aspect should be obvious.

 

[Bob Noel]

Was there any variation in the number of engine hours between analysis?
Was there any variation in the type of flying? cold vs warm?

 

[peter-h]

I take reasonable care to keep it consistent.

The engine is run for a few mins, till 90F shows on oil temp, and then I drain the oil.

I let it run for 1 minute into the jerrycan, and then I fill the little bottle from the flow.

Curiously other metals are more consistent. Is nickel a lot heavier than the others?

There was no recent maintenance. Rebuild (SB569A crank swap) ~ 300hrs ago.

 

[ScottM]

How much variation in the numbers are you seeing?

 

[peter-h]

It is what I posted in my first post.

 

[ScottM]

It is what I posted in my first post.

Doh!

 

[Bob Noel]

I have seen big variations.

For example nickel has gone down from 2.5ppm to 0.9mm and that is from a 29hr sample to a 46hr sample.

Obviously one should not complain but what could be the physical explanation?

Iron has also gone down by 1.5x, copper is inline with the oil age, chromium has gone from 5.5 to 3.2.

Assuming the metals get into the oil mainly from bearing / sliding surface wear (which is not an unreasonable assumption!) why do they vary so much?

How many quarts were added for the 29 hr sample and for the 46 hours sample?

 

[Steve Foley]

The engine is run for a few mins, till 90F shows on oil temp, and then I drain the oil.

A difference in ambient temps would cause a difference in the running time to bring oil up to 90F.

I have no idea if that would make a difference.

 

[peter-h]

How many quarts were added for the 29 hr sample and for the 46 hours sample?

A good point but it is pro rata, more or less.

Even when the analysis is corrected for this, and normalised for oil life, there are still big variations.

Maybe one should fly the aircraft before draining the oil, but then you end up with a very hot engine which is impossible (initially) to work on, which is why one doesn't do it that way round.

 

[Clark1961]

Maybe one should fly the aircraft before draining the oil, but then you end up with a very hot engine which is impossible (initially) to work on, which is why one doesn't do it that way round.

Actually, that is how I do it. I do have a quick drain and attach a hose prior to stabbing the quick drain on the fitting. Gotta wear a long sleeve shirt and gloves to avoid leaving strips of flesh on the exhaust manifold.

Using this method and taking the sample from the middle third of the stream I do get fairly consistent samples. If I do anything else then I'll generally get shifts in the amounts of most metals. The one time I took a sample from the drain bucket a couple of days after draining the oil the results were really crappy.