Randall Young wrote:
>
> Michael / Jonmac :
>
> I agree thoroughly with the comment that you shouldn't use lower viscosity
> oil than an engine is designed for.
>
> However, full synthetic has been available for some time in 20W-50, which
> should be plenty thick enough for our old engines. I've in fact been
> running my TR3 (rather hard at times) on it for several years with no
> apparent ill effects.
>
> BTW, the SAE weights are in fact directly related to viscosity measured at a
> particular temperature, not to "lubricating capacity" or "shear and film
> strengths". Lubrizol provides a nice chart at
> http://www.lubrizol.com/referencelibrary/readyreference/6-OilClasses/physica
> ls.htm .
>
> Note also that "W" (as in 20W-50) does not refer to "weight" but in fact
> "Winter", because the viscosity for the W grades is determined at a low
> temperature (different for each weight). A particular oil formula is rated
> by the lowest, then highest grade that it meets the standard for. So, for
> instance, the '20W' part of 20W-50 means that it is no thicker than 4500
> centipoise (cP) at
> -10 C, the '50' part means it is no thinner than 3.7 cP at +150C.
I checked the Lubrizol site, and it verifies what Randall says. I stand
corrected, to a degree. However, the standards used for viscosity
measurement, I will note, use two tests with two different
measurements--one which measures in centiStokes (this is a test which
measures by gravity) and one in centipoise which measures the resistance
to movement of a stirrer in the viscous media and is independent of
specific gravity).
CentiStoke measurements of viscosity must be corrected for specific
gravity to obtain absolute viscosity. So the centipoise readings Randall
cites above are more accurate than those supplied to me from another
site elsewhere. But, a centipoise is a measure of shear strength, so I
only have to eat half a crow. <smile>
And, from the Castrol site, the difference in specific gravity of their
synthetic vs. mineral oils is about .86 vs. .89, so in terms of absolute
viscosity, the mineral oil viscosity will be about 3-1/2% higher than
advertised, if the oil viscosity is measured by the gravity cup method
in centiStokes.
> AFAIK, there is no such thing as "30W", simply because the SAE has not
> defined it.
That is likely a throwback from the days when only single-weight oils
were available, and might be a colloquialism meaning "30-weight."
Having said all that, I still stand with the original observations about
viscosity in general. Two of the three examples I cited were in using a
full-synthetic oil available in the 1970s as only a 5W-30. An oil of
this grade, regardless of its synthetic nature, is not up to the demands
of very hard use, unless the engine is designed for such an oil, and
viscosity affects flow through and out of the bearings. That
observation, from a strictly physical point of view, is hardly to be
challenged. I can be faulted, in part, for assertions that
high-temperature viscosity is lower for synthetics--admittedly, my
opinion is colored by personal observations from the use of the earliest
versions of Mobil 1, then only available as a 5W-30, and the general
belief that, when it comes to keeping engines alive, more viscosity at
high temperatures is better.
This is part of the reason why so many international racers used castor
bean oil in the earlier years of racing. It had adequate lubricity
(certainly not as good as modern oils, though), and also had the unique
property that its viscosity went up as its temperature increased. The
higher the viscosity at a given temperature, the greater the oil film
thickness. The greater the oil film thickness, the less likelihood of
engine parts kissing each other at high speed, thus rendering them one
part, rather than many associated parts. That's the point of having
sufficient viscosity to the oil--pressure stays up, and the oil film
thickness is sufficient to maintain running clearances without contact
of rotating parts.
Randall is correct. An adequately viscous oil, whether a synthetic or
mineral oil, will do well in normal service. Latching onto the latest
synthetic, if of inferior grade, will still do an engine damage.
Apart from this, there is the business about additives attacking seals.
If a commonly available seal fits a Triumph, and it's been formulated
for newer synthetics, great, no problem. If it hasn't been, it may
erode.
A last point... several people have sent me examples from the Web on
viscosity measurements which indicate that to meet a certain API spec,
an oil has to behave in particular ways. Some of those measurements,
provided by the oil manufacturers, are, inexplicably illogical or
possibly simply wrong or not truthful. The manufacturers provide the
proof that the oils meet the specifications. The API doesn't test them,
ordinarily. API only issues the spec, the testing requirements and
testing methods. It's up to the manufacturer to assert compliance.
Keep in mind that John Macartney's original post on this subject was to
pass on the wisdom of the oil manufacturers themselves that contemporary
full synthetics are not recommended for use in older engines. Maybe they
know something they aren't fully advertising to the general public,
either about oil additives, oil properties, or both.
So, the simple test should be, if you pull the dipstick on a hot engine
and the oil runs off it like water, and if the engine is measurably
noisier than it was before you filled it with the latest synthetic, the
oil's probably too thin at temperature, regardless of what the
manufacturer asserts, regardless of its grade or composition.
Cheers.
|