Skip to main content

I was going to post this as a replay to a post by oceandeep but I'd really like to hear from others in regard to this subject. http://maintenanceforums.com/e...3451/m/51420123373an

This past Friday we received the results from a reputable lab chosen by our lube vendor to address a viscosity variance in an ISO 320 fluid that negatively impacted our over-all lube health KPI as well as the trend for a rather large and specific lube analysis route. The KPI is the % of total samples taken marked as acceptable, alarm, or alert. The variance caused a 20% swing from acceptable to alarm; all driven by a change in viscosity only. That's a huge and long term damaging effect to this KPI that will take some effort to repair.

We did the work, spent the money, and have taken steps considered by many to be exemplary to deliver, as the saying goes, the right lube in the right amount at the right time as diligently as possible
Some of you reading this may be familiar with our Lubrication Center as seen here: http://maintenanceforums.com/e...20728073#22520728073

Back to the point of this discussion.

The reputable lab was given multiple samples that were drawn from a 65 gal bulk storage container and a point of use container. The results were ALL over the chart. The lowest of 4 blind samples was 304 while the highest was 319 all @ 40c. According at ASTM D445 this much variance in repeatability and reproducibility is unacceptable. Now I have a another layer of potential culpability adding to my anguish as I strive for "acceptable" conditions, not perfection, simply acceptable conditions. How can we as end users of critical lab services trust the data we receive when it appears inconsistency is the norm. A big statement but in the end a true statement.

Replies sorted oldest to newest

quote:
Originally posted by noisemakr:
The reputable lab was given multiple samples that were drawn from a 65 gal bulk storage container and a point of use container. The results were ALL over the chart. The lowest of 4 blind samples was 304 while the highest was 319 all @ 40c. According at ASTM D445 this much variance in repeatability and reproducibility is unacceptable. Now I have a another layer of potential culpability adding to my anguish as I strive for "acceptable" conditions, not perfection, simply acceptable conditions. How can we as end users of critical lab services trust the data we receive when it appears inconsistency is the norm. A big statement but in the end a true statement.


I see two containers mentioned, a 65 gal bulk storage container and a point of use container. Can you match the samples to what came from where? You also seem knowledgeable enough to know that some lubricant additives will stratify or settle out of the oil. This situation is somewhat aggravated if the oil is stored in a relatively cool environment. Additives that are particularly prone to stratification include those that are organo-metallic like detergents, antiwear additives, defoamants, EP additives, and some rust inhibitors. Could this be a factor?
RM
Good Point John from PA however: the point here remains the same. The samples were drawn on the same day from the same containers, sent the next day, and I'd have to assume, tested in arrival order at that lab. All things being equal, why the variance? ASTM D445 is well defined and specific in methodology. What was previously described should not happen. Sadly this is not the only reputable lab to return varied results under exact conditions which is the source of my frustration and the foundation for my reply to oceandeep.
RM
Last edited by Registered Member
No reply was offered.
The worst part about lab inconsistencies is the stress it puts on the lubrication vendor and end users relationship. Especially if both use a different lab. Now throw in a third party lab that both agree upon and when the results come back skewed, where can you go from there? After all an ISO 220 is an ISO 220 and should pass the same test no matter what lab is testing but in the scenario just laid out, who’s lab has the final decision when it comes to warranty issues? It can get very complicated very easily. Personally, I would not use the same lab and my supplier simply for the checks and balance and data integrity.
RM
what exactly is your problem? ISO 320 means 320 mm2/s at 40 deg C plus or minus 10%. All the samples thus represent a ISO VG 320 oil. When the bulk tank that the samples are taken from is not constantly agitated, you may well experience some variance in viscosity - the way the samples were taken can have a large influence on the values observed also.

As far as the lab is concerned: they just analyze what is supplied to them - and if that varies, so will the results. The method of viscosity determination itself has a reproduce ability and repeatability that also may add some variance to the results obtained.

In this particular case all the samples obviously were within the limits as stated in the ISO VG specification.

Just another remark: viscosity often is regarded as a critical value for a lubricant. Of course it is, but one should not overstate it. The fact that there may already be 20% variation within the ISO VG 320 grade is a clear indication that there is some leeway. Another way of looking at it is the actual application. Assuming the oil is used as a gear oil, the filmthickness that can be obtained usually will vary with the square root of the viscosity. That means that for all practical purposes a somewhat more viscous or less viscous lubricant actually matters much less then people are inclined to think....
RM

Add Reply

×
×
×
×
Link copied to your clipboard.
×