The info in the table in SAE 650866 could be seen in a few different ways without knowing exactly how the testing was done. For instance, did they do each of those dust size ranges one test at a time with the same total number of dust particles in each test? - ie, did they do a separate wear test for dust 0-5u and another separate test for 0-10u, 10-20u, etc? What was the dust size break down and volume for the 0-80 micron dust test? Apparently the reduction in engine wear rate in the 30-60 min period of the test is due to the particles breaking into smaller particles, but what's strange IMO is if that's the case then you'd have more smaller partilces to do wear, and as the Cummings and this test show, even the smaller particles cause wear. Did the study say how small the particles broke down to ... was it something like 1 or 2 microns?
They performed separate tests with each type of dust. The dust samples had the same total mass, not the same number of particles. The 0-80 micron dust was "Standardized Fine Air Cleaner Test Dust", probably based on old SAE standards but similar to modern ISO Fine A2, which has a fairly linear distribution by mass.
They weren't able to measure the final particle sizes, but explained their reasoning to why they thought the wear rates indicated that the dust was all getting ground down to the same size:
"Another piece of important information is shown in Table
6. The second column shows the wear rates in milligrams/
hour caused by each of these fractions of road dust during
the second half hour after the add. Note the similarity of
the second half-hour rates of fractions larger than 0-5 mi-
crons to the first half-hour rate of 0-5 microns. They are
almost identical. Something is happening to the dust as the
test time progresses to make it all wear like the 0-5 micron
fraction did initially. There is no filter in the circuit, so
none of the dust is being filtered from the oil.
The Standardized Fine Air Cleaner Test Dust acts in the
same fashion, and a further measurement of the wear rate
occurring during the time between the first hour and the third
hour shows that it continues to decrease with time.
We have hypothesized that the dust is changing in parti-
cle size by grinding in the oil pump and between the wear-
ing surfaces of the engine. Some change in particle size
distribution of and reduction in the amount of recirculating
contaminant undoubtedly does occur from settling out of
particles, particularly the larger ones. However, the data
show that the 0-5 micron fraction also changes quickly, and
it is hard to visualize that a significant amount of these small
particles settles out of lubricating oil in half an hour.
Attempts at measuring the particle size distribution of
contaminants in crankcase oil by us, and by others have been
largely inconclusive, because of the many variables involved,
and because of the difficulties of sampling and measuring.
However, we do know from these studies that, in general,
contaminant grinding does occur in an engine. Also recent
work by the SAE Filter Test Methods Committee has shown
that the contaminant mixing pumps on the SAE standard oil
filter test stand, as well as the oil flow pump itself, do cause
contaminant grinding to a marked degree.
This information might tend to support a theory that the
original particle size of a test dust is of little relevance be-
cause it is quickly ground up by the engine into small parti-
cles less than 5 microns in size. If we were concerned only
with sump clean-up over a long period of time, this premise
might be acceptable, but since we are concerned with en-
gine wear, and since the grinding of contaminant in the en-
gine obviously causes engine wear, we must be concerned
with the particle size of test dust. "