Reliabilityweb Discussion on Grease Sampling and Analysis

quote:
I seem to remember that other options are available for sampling. While browsing at the booth I remember something like a tube that could pull a small vacuum that pulled the grease. I'm just going by memory and could be mistaken. You would think that the video would show that option


Don, you must have only seen one of the videos. There are indeed three different methods of obtaining grease samples that we've developed. We named them each after the most common sampling application, but they are not necessarily limited to those applications. One, which is the bulk of the discussion on this thread, is the Electric Motor Bearing Kit, and consists simply of the threaded fitting, designed to be threaded into the 1/8" NPT drain of common motor bearings. It also includes a cap and protective container to return the fitting for analysis. The second is the Pillow Block Kit, which includes a plastic spatula, a syringe, and the sampling fitting, and is meant for locations where the grease is not contained, or purges from the equipment by design. The third, which you mention in your posting, is the MOV/Gearbox Kit. This utilizes a t-handle extension that allows the remote actuation of the sampling fitting. This allows the fitting to be inserted into a housing, say through a threaded hole normally plugged, and can travel through the reservoir to the point of contact with the gear or bearing to be sampled, where it is actuated to grab the representative sample. This and each of the methods described can be viewed on videos posted on the greasethief.com website. Grease sampling videos (Sorry Terry, I'd put that here too, but don't know how to post our videos on your website.)
We are continuing to evaluate various sampling challenges to determine the best way to get grease samples for various types of machine components. As has been discussed throughout this posting, there is no one simple way to get a sample, just like in oil analysis. To get the most from this diagnostic tool, we must evaluate each case and choose the optimal method to get a representative sample, and recognize those cases when sampling is not practical. Hopefully here we present some new tools that will enable good samples and meaningful analysis methods for when we can get a good sample.

Just to briefly address Ron and pete's comments on need for grease analysis. While I agree that other diagnostic techniques should be a part of any monitoring program, the one area that they are generally not effective at diagnosing is the root-cause of lubricant failures in greased bearings. We have done many analyses for customers with vibration, infrared, ultrasound and MCE testing, that were greatly complemented by grease analysis. Mixing of incompatible greases is a significant problem in many industries and facilities. In some cases, the mixing is not happening because of the facility mechanics, but that other greases are being introduced during rebuilds or by installation contractors. If these greases are incompatible, initial thinning of the grease can occur, followed by a release of the oil from the matrix, and then a hardening of what is left behind in the bearing, causing failure. If this condition is detected early through routine testing, a slow purging process can be used effectively in many cases to save the bearing from the onset of damage that would later be detected by these other technologies. This has been done effectively at national research facilities and pharmaceutical operations. Additionally, other proactive measures can be taken when particulate contamination is discovered in the grease. In some cases, this has been traced to a contaminated grease supply, which was the equivalent of a maching compound on the bearings. By flushing this out proactively, and testing to find the offending contaminated supply, systemic wear at the facility was addressed by removing the offending lubricant. This has also been the case in the identification of greases in grease guns with insufficient anti-oxidant packages. If through long-term or inappropriate storage, the grease has been subjected to conditions that deplete the anti-oxidants, the life of the grease can be shortened in service, and premature bearing wear can take place. If the purpose of the monitoring program is only to identify bearings just prior to failure to prevent downtime, the addition of grease analysis might not add much. But if the goal is to prevent recurring failures and maximize reliability and component life, grease analysis adds another source of information that can aid in that process.

I think this is a great discussion, and I thank other members for inviting me to comment.
Some of the skepticism expressed here is healthy, because grease analysis is not a simple endeavor. As indicated by some of the contributors, grease sampling is the real key, along with having testing options that can provide meaningful results with the very small quantities of grease that are typically representative of the lubricated component.
There is a lot to cover here, but let me start with electric motors. There are many different bearing configurations in greased electric motors, and indeed even a differency typically between the inboard and outboard bearings. Much has been published on this topic based on gut feel and intuition. We wanted to get some more reliable data on bearing greasing, so we made some test stands that would allow us to monitor regreasing of a particular common electric motor bearing setup. In this case (a 60hp outboard bearing with single-shield facing the windings, and both the zerk and drain on the open side of the bearing), we observed that there was surprisingly good mixing of the the grease in the housing when new grease was added to the bearing in accordance with regreasing guidelines. We traced the mixing process by hand packing the bearing with one type of grease, and adding through the zerk a grease with different chemistry. We were able to trace chemistry markers to show that good proportional mixing took place in this application. Things to note: 1-this configuration is not representative of all possible motor bearings, but it could be considered a worst case (large open reservoir relative to grease in bearing, and grease is not forced through the bearing to the drain); 2-this was done with fresh greases, and older and possibly dried greases may not perform the same. However, for a program with frequent relubrication and replenishment of greases to maintain acceptable consistency and lube quality, this is a promising finding.
Waiting for grease to exit the drain is another matter. I agree with some of the comments here that many electric motor bearings do not routinely purge grease at the drain hole, and for some, maybe never until the housing is overfilled. However, without a method to catch the grease at the drain, in the event that grease does exit at the hole, there is no way to capture the effluent for analysis, even if it is to figure out why grease has started to purge. Secondly, it is generally accepted that failure to remove the drain plug is a contributor to overlubrication and pushing grease to the windings. The use of a properly designed grease sampling fitting in the drain eliminates the need to ensure mechanics are routinely removing, waiting a few hours, and then replacing the drain plug. The sampling fitting can be left in place to provide purge relief if needed, and is a sentry to capture any grease that happens to exit the bearing.
The video that is referenced here is one of a series, that we will soon make available on our website. I'm going to have to make several postings here to address this issue (as Terry encourages placing content here rather than links to further information), so I'll be back to add more. Bottom line is that there are many machine configurations, and there is no one-size-fits-all for grease analysis. Just as there is no such solution for oil analysis. Remember the early days of oil analysis, when any drain you could find would do, as long as you could fill a sample bottle. There we learned that bad samples meant bad data, and in some cases, we've had to retrofit machinery with sampling devices specifically designed for the reservoir,to get good representative samples. For grease, we are going to have the same challenge, and may even need to recognize, as in the crane example just given, that the machine geometry prevents an online representative sample from being presented.