Guys, I study lots of proprietary lubricant data in the course of my business. The values listed in the bar graph seem very low for all oils tested that I am familiar with. Amsoil has, since its inception been a low volatility oil by design. i.e. they need the oil to be stable and no cook off over longer drains.
Both "TGA and NOACK" tests are Thermal Gravimetric Analysis tests. NOACK is designed to measuring the volatility stability of the oil. TGA measure distillation levels at different temps and times.
The key methods are ramp up time and temp, max temps, and time at max temp.
Keep in mind D5800/NOACK is method of measuring volatility stabililty on the BENCH. ILSAC used to require that 22% max loss limit for pass/fail at 250C for one hour.
The oil is heated to 250C and held at that temp for one hour, air is then introduced into the chamber and maintained at a constant flow rate under slight vaccum. After the hour is expired the amount of oil remaining in the chamber is weighed,and compared to the original weight of the oil.
The source of the data shared here was Institute of Materials. Their own in house lab is Savant labs and the SAVLAB volume test is based on their protocal. They perform a very respected broad based bench testing protocal for blinded off the shelf motor oils and publish the results. Last time I priced the book it was $10,000 for a regions oil. Asian, US, and Europe are available. So my data was independent of any in house testing that could be construed as biased.
The reason the values change is that formulations change, rapidly!
For instance in 2000 data for ATM 10w30 was 7.06%.
2004 showed ATM 10w30 at 12.06%
Most AFFORDABLE lubricants will test in the 12-15% range by independent lab.
Like any measure of a formulation there are tradeoffs. Synthetics may be extremely resistant to volatility (burn off = adding oil) but oxidation resistance is poor. Both contribute to thickening oil and throwing add balance out of whack.
Use oil analysis and logging top ups as a accurte way for you to "in-house" test your chosen brand. Don't use one bench test measure to guide your oil purchase decision.
The 2005 Amsoil D5800 tests above show me excellent volatility stability that engine sequence testing is causing formulators to ensure.
The poorest reading above is a good one in a well maintained car. Some of the relatively good results are not consistant with UOA we see daily!
lobo11's test was obviously a bit harder on motor oil but the relative measure of the values is a good one from what I can tell on overview.
Additives like Auto-RX will not do as well alone as they do in the host oil at design treat levels. It is designed to be a cleaner not a motor oil!