Gary, I changed this topic thread title but copied your comment/question. I think I have some good data on the subject. I tracked wear and oil data on my twin Cessna for more than 3000 hours. I determined that an engine "wears" metals on a relatively constant schedule. You have to buy into that assumption or else punt this entire analysis. I measured wear numbers in terms of "ppm per hour" for each metal adjusted for the number of hours on the engine oil, adjusted for for the dilution of added clean oil and the loss of oil during the period (burned or leaked). Initially, I used A/S 15W-50 and changed oil at 50 hour intervals sometimes longer but went to straight weight and 25 hours changes. Today, I used straight weight 100 and 25 hour changes on the Legacy IO-550.
I found that after 20-25 hours the wear rates increased quite steeply. I determined from Shell that the materials in the 15w-50 get chopped and break down with use thereby moving the multigrade closer to single grade over time. I then went to straight W100 in the summer and W80 in the winter with 25 hour oil changes and the numbers improved dramatically in terms of lower wear measured as ppm of metal generated per hour. The cost was actually better since the W100 was less than half the price of 15w-50 at that time.
Today, the oil analysis folks have gone towards measuring wear rates versus straight metal numbers in the oil and that makes more sense to me. The graph below I found from 1997 only shows how the wear rates varied over the life of my airframe (100s of hours on x-scale) but I could recreate the wear per hour change with some time if you wanted to see it. However, the takeaway for me form years of tracking was:
1) 25 hours in the TSIO-520 NB was a good oil change point for wear purposes using A/S 15w-50. At 50 hours the wear rates were much higher and the increase per hour was much steeper indicating the wear was accelerating.
2) Using W100 or W100 PLUS resulted in better wear numbers per hour than 15w-50.
3) Plotting wear in ppm per engine hour (for that oil change) resulted in meaningful results that I could act on. Total numbers mean little if you can't relate to the hours or oil added and burned.
4) If you track hours and oil consumed/added and do regular oil analysis you can predict major problems in advance (see graph below).
Paul
Spruce Creek
On 2012-01-27, at 11:06 AM, Gary Casey wrote:
And why do you change oil every 25-35 hours? With a modern engine and modern oil, I don't see any reason to change oil that often. I would run it at least 50 hours and try to change it before 75. If the engine is run often, even 100 hours isn't out of line. What does the oil look like at the oil change? Can you see the dipstick through the oil? If so, it probably doesn't need to be changed yet.