dnewton3
Staff member
I got my data right from the SWRI sheet, and I double checked what I typed; no mistakes. However, if their sheet had an error, I would only know how to regurgitate what they publish and not be able to discern a mistake in regard to air intake temps, etc. But, their data seemed reasonable to me if applied in real world ops, at least at face value. I would defer to A Harman and dusty and TT as they have experience in real world trucking and heavy duty diesel development engineering.
Just don't loose sight of the fact that this is ALT (accelerated life-cycle testing) and they do manipulate things purposefully to cause/induce the failure modes they seek. It's not "wrong" to do this, but it often does not translate well into reality for most folks. They do this to cause a delineation of some parameter in the test subject, be it oil, filters, hoses, belts, tooth brushes, audio speakers, semi-auto rifle bolts, etc etc. Keep this in mind about ALTs:
- The data they generate does often reveal some product performance disparity when multiple options are present; one thing will most often usurp another.
- The data they generate gives a very clear view of how things perform in a world we likely may never visit because they conditions they induce are not typically replicated by life's actual events.
The fact that these two Amsoil fluids did so well should not be ignored. This is a very abusive test, and they fared very well. But even the normal dino oil surpassed the test threshold by 6 hours. If you were to run your heavily loaded rig uphill for 7 straight days non-stop, and had very high ambient temps, and never OCI'd in that duration, this may make you want to consider using a premium PAO product. If not, well then it's kind of a moot point. What we cannot assume is that any linear relationahip exists in the data stream. Failure may come rather abruptly towards the end of each unique product performance collapse. 30 hours of "test" time is a pass. But once you accelerate into phase two, it's likely a parabolic escalation of abuse.
The DD scuffing test is very, very abusive. Anything that passes that test will do well in most any application the real world will present. Being able to surpass that test by some gross margin is admirable, but it's hardly necessary.
Or, to put it into one of my infamous analogies ...
Do you really need to be able to run a marathon, if you're only walking the mail to the end of your driveway?
Just don't loose sight of the fact that this is ALT (accelerated life-cycle testing) and they do manipulate things purposefully to cause/induce the failure modes they seek. It's not "wrong" to do this, but it often does not translate well into reality for most folks. They do this to cause a delineation of some parameter in the test subject, be it oil, filters, hoses, belts, tooth brushes, audio speakers, semi-auto rifle bolts, etc etc. Keep this in mind about ALTs:
- The data they generate does often reveal some product performance disparity when multiple options are present; one thing will most often usurp another.
- The data they generate gives a very clear view of how things perform in a world we likely may never visit because they conditions they induce are not typically replicated by life's actual events.
The fact that these two Amsoil fluids did so well should not be ignored. This is a very abusive test, and they fared very well. But even the normal dino oil surpassed the test threshold by 6 hours. If you were to run your heavily loaded rig uphill for 7 straight days non-stop, and had very high ambient temps, and never OCI'd in that duration, this may make you want to consider using a premium PAO product. If not, well then it's kind of a moot point. What we cannot assume is that any linear relationahip exists in the data stream. Failure may come rather abruptly towards the end of each unique product performance collapse. 30 hours of "test" time is a pass. But once you accelerate into phase two, it's likely a parabolic escalation of abuse.
The DD scuffing test is very, very abusive. Anything that passes that test will do well in most any application the real world will present. Being able to surpass that test by some gross margin is admirable, but it's hardly necessary.
Or, to put it into one of my infamous analogies ...
Do you really need to be able to run a marathon, if you're only walking the mail to the end of your driveway?
Last edited:
