The disciplines of machinery lubrication and vibration analysis are both critical elements of a reliability-focused culture. Both enable us to stay on the left side of the revered potential failure-failure (P-F) curve. Unfortunately, even in the modern era, these programs often exist in a disconnected state, forgoing the value achievable with their union in a connected preventive and predictive maintenance ecosystem. This article intends to touch on the relationship between lubrication and vibration, explain how we can use the relationship to our advantage, along with providing some points of consideration to bring these disciplines together in a cohesive design.
Machinery lubrication encompasses the practices relevant to maintaining adequate lubricant conditions within machine components moving in proximity to each other. If a machine rotates, slides, or pivots, chances are that it is either lubricated or has a designed wear surface installed (e.g., bushing). The lubricant may provide wear reduction, cooling, contaminant removal, and protection from contaminant ingression, depending on the formulation and application. One of the most important parameters of a lubricant is viscosity—a measure of resistance to motion or flow. The viscosity is related to lubricant film thickness, which ultimately determines the amount of separation between moving machine components, and thus the friction between them. Maintaining the optimum fluid film is essential to minimizing wear and vibration.
Any experienced reliability professional can attest to the correlation between the lubricated state of a machine and its detectable health indications: an improperly lubricated machine will generally operate with greater temperature, noise, and vibration. This anecdotal evidence is supported by studies found in professional literature. One such study, by Ripin and Yusof (2018), tests the impact of fluid film thickness in rolling element bearings on the overall vibration levels. According to the study, whether starting with an unlubricated bearing or a lubricated bearing, vibration levels decrease over time initially as bearings “wear-in”. As a new bearing accumulates operating hours, surface asperities inside the bearing are worn-down (i.e., polished) permitting smoother contact between the bearing’s raceway and its rolling elements. However, this decrease in vibration is short-lived, much more so in the unlubricated bearing. Overall vibration levels are shown in the study to rise if the fluid film is not established and maintained in a nominal range after the “wearing-in” has been completed. As expected, the lubricated bearing also showed significantly less surface roughness (a proxy for wear) over time during the testing.
Given that the state of a machine’s lubricant condition is a leading indicator of machine life over time, it is important to understand how vibration analysis can provide insight as part of a continuous monitoring program. The vibration signature of operating equipment can be used to detect cases in which the fluid film provided by the lubricant becomes disrupted or sub-optimized. Lubricant degradation will present indications in the time-waveform or spectrum due to elevated friction or the presence of impacting. Special techniques such as high-frequency enveloping (HFE) are particularly useful in detecting such signatures. When detected, the vibration analyst can then communicate to the maintenance team that a machine requires lubricant replenishment, change, or perhaps investigation of improper lubricant delivery. Better still, the frequency of lubricant delivery can be adjusted based on the machine’s needs. This is important because the lubrication needs of a machine change over time due to wear and environmental factors, thus rendering purely time-based strategies ineffective.
While vibration analysis can help optimize lubrication programs and provide early warning of machine health issues, achieving high reliability of lubricated assets requires a more robust approach. Machine preventive (including lubrication) and predictive maintenance asset management plans should be developed using formalized methods such as Reliability Centered Maintenance (RCM), Failure Modes and Effects Analysis (FMEA), or Preventive Maintenance Optimization (PMO). Outcomes of the aforementioned methods will necessitate the inclusion of additional domains of expertise if actionable monitoring or maintenance tasks are desired. Regarding lubrication programs specifically, it is strongly recommended to adhere to the guidance in the International Council of Machinery Lubrication (ICML) Standard 55.1, which covers program design, training, equipment, etc. Regarding vibration analysis programs, sites should include the following: appropriate monitoring posture (e.g., online versus periodic based on equipment criticality); experienced and certified analysts (ISO/ANST Category 3 or 4); a platform allowing timely notification, diagnostics, communication, and monitoring of program and asset health; scalability to the enterprise level. Most important of all is to tie these strategies and programs together, sharing knowledge and supporting continuous improvement of asset health.
Machines do not last forever, but their lives can often be better aligned with business objectives by administering effective maintenance and monitoring programs. In this vein, two common pathways for life extension include lubrication and vibration monitoring. Improper lubrication leads to vibration through fluid film degradation and the resulting surface contact. If not detected and diagnosed, such conditions can prevent timely intervention into further development along the P-F curve—i.e., moving towards the dreaded far right-hand side of the curve. The best approach to maintaining a high level of machine health is one that intertwines both lubrication and vibration monitoring best practices, rather than permitting them to operate as siloed programs.
ReferencesRipin, Z. M., & Yusof, N. (2018). The Effect of Lubrication on the Vibration of Roller Bearings. MATEC Web of Conferences, 217. doi:https://doi.org/10.1051/matecconf/201821701004