51
FUELS & LUBES INTERNATIONAL
Quarter Two 2015
gives a greater chance of repeatability
and reproducibility, so Toyota se-
lected Southwest Research (SwRI) to
design and document the construc-
tion of the Golden Stand. TEI then
supplied the stands according to the
SwRI design, and eight stands now
exist: four at SwRI, three at Intertek
and one at Lubrizol.
“This has been a well-received
concept,” Kowalski said, because the
stands were easy to install and have
displayed minimal variability.
Ford is sponsoring both of the
tests completely new to GF-6. One of
these, the ChainWear test, addresses
the concern that the valve timing
chain is sensitive to new oil formula-
tions. This test was supposed to have
been introduced for GF-4 and GF-5,
and hopefully will be ready for GF-6.
It is now being developed at SwRI
and Intertek.
The second new test for GF-6
monitors low-speed pre-ignition
(LSPI), a growing concern with
small-displacement engines. Ford
is using the same engines for the
chain wear and LSPI tests, a 2012
2.0-L L4. Initial testing discriminated
among three oils of low, medium and
high LSPI occurrence. The cylinder
pressure measures LSPI events, and
earlier problems with the pressure
transducer have been resolved. New
transducers and a new ECU are
duplicating prior testing with low
and high LSPI oils, and the new ECU
calibration allows a certain number
of LSPI events to occur without
damaging the engine.
This number of new tests for one
category is unprecedented, Kowal-
ski said. Also unprecedented is the
simultaneous development of the
PC-11 category for heavy-duty diesel
engine oils. In previous iterations,
the releases of the gasoline and diesel
categories were staggered. Some of
the same people are working on both
specifications, so human and mate-
rial resources have been stretched
thin.
Because of these new challenges,
it is not surprising that the category
has so far been about four and a half
years in the making. With only one
new test, GF-5 took four years to de-
velop, Kowalski said. But, she added,
“We are seeing the light at the end
of the tunnel.” All parties involved
have signed the Memorandum of
Agreement and have agreed on the
Needs Statement. All funding for the
precision matrix has been collected
as well. Matrix test labs and reference
oils have also been designated.
There are two pressure points
for completing the testing: First, four
out of the five tests for the current
GF-5 category will cease to be avail-
able in 2016 and 2017. And, “OEMs
[original equipment manufacturers]
want to take advantage of the fuel
economy benefits of oils that comply
to these higher specifications,” said
Kowalski, “and they can’t do that
until the oils are available in the
marketplace.”
Insight into a
single test
Kaustav Sinha, project manager for
Chevron Oronite leading the GF-6
and Dexos1™ developments globally,
gave an update on the Chrysler
Oxidation and Deposit Test for GF-6,
on behalf of the test development
team comprising members from Fiat
Chrysler Automotive, Shell, Chevron
Oronite, SwRI and Haltermann.
Sinha’s talk emphasised just how
much a test needs to go through
before it can even enter the industry
precision matrix. The test uses a
2014 PentaStar 3.6-L V6 engine (the
Sequence IIIG runs on a 1996 GM
Powertrain 3800 V6 engine, which is
no longer in production). It runs for
90 hours, with a six-ounce addition
of oil every 20 hours. SwRI, IAR,
Lubrizol, Afton and Ashland have in-
stalled the test and contributed to the
prove-out matrix data generation.
The test focuses on oxidation
performance as well as kinematic
viscosity increase (%)—which is an
indicator of a lack of oxidation stabil-
ity— and weighted piston deposits
(WPD). It also has to be backward
compatible with the Sequence IIIG
test, so much of the test development
involves comparing the reference
oils for the old and new tests to make
sure that the GF-6-compliant oil is
compatible for vehicles that had been
using the GF-5 oils.
The developers had to collect
data on repeatability, reproduc-
ibility and discrimination for traits
including percent viscosity increase
(PVIS), WPD, phosphorus retention
and compatibility with 0W-16 oils.
In each case, the focus was to make
sure that the reference oils could
discriminate performance for each
aspect as well as they could for the
IIIG. When showing this process for
PVIS, Sinha showed how one of the
reference oils for the Chrysler test
consistently had very low results—
that is, performing very well. “If I am
developing a test,” Sinha said, “I don’t
actually learn that much from this
oil because it will always pass. How
is it going to respond to the changes
I’m going to make?” During engine
test development, borderline oil, i.e.
sensitive to changes in test severity,
plays an important role in providing
meaningful insights into the test ap-
petite before settling on the final test
procedure. A passing and/or failing
oil with statistically demonstrated
separation on the test parameters
satisfies the discrimination aspect of
the test, he said.
Compared to the Sequence IIIG,
some of the procedures of this test
have become more stringent. The oil
consumption for the IIIG test was
3.2-4.8 L, and for the Chrysler test
it is 2.2-3.1 L. Oil addition for the
IIIG was 18 ounces every 20 hours,
whereas for the Chrysler test, only six
ounces are added. Sinha commented,
“With the Chrysler engine, basically,
you can run the entire test on a full
charge, which is due to the reduced
volatility effect that the test has.”
This reduction in volatility, Sinha
explained, is one consequence of
updating a test from the 1990s to
newer hardware. Volatility refers
to howmuch oil leaves the engine
during its operation by “volatilizing”
into the exhaust. The less oil that
volatilizes, the closer the oil stays to
its initial viscosity. Sinha said that the
test developers found an SAE paper
from 2007 (SAE 2007-01-1961 –A.
Boffa and S. Hirano, “Formulation
Impacts on Sequence IIIG Viscosity
Increase”) that showed that initial
viscosity increase in the IIIG test is
due to oil thickening from volatility,
and this allowed them to estimate the
volatility effect using PVIS. “There
has to be a correction factor for
volatility,” Sinha said.
Kaustav Sinha
