Hello Rolly, I see your point, but I firmly believe that "random" and "accelerated" have distinct meanings in the context of reliability-centered maintenance (RCM). Random failures refer to unexpected events that can occur regardless of maintenance efforts, whereas accelerated failures suggest that a component is deteriorating faster than its intended design life. Thank you, Mike.
When discussing failure, it is important to consider it from two perspectives. The first is the physical degradation of equipment, which is influenced by design, operating conditions, and maintenance practices. While we can take steps to slow down the rate of degradation, the wear mechanism is inherent and cannot be completely eliminated. This results in unexpected failures, which may be caused by neglect, misuse, or interference with the equipment.
On the other hand, the cause of degradation is deterministic, meaning we understand what leads to it. By managing the rate of deterioration, we can extend the lifespan of the equipment, shifting the distribution of failure times to occur later in its life. This does not change the predictability of failure times, as the underlying wear mechanism remains the same.
In addition to physical degradation, there is the statistical distribution of failures to consider. Failures may occur randomly at any point after commissioning, influenced by the progression of degradation. While we can influence the likelihood of certain events, such as bird strikes on aircraft, the randomness of such occurrences remains.
Overall, it is important to recognize that while the physical nature of failure is fixed, we can impact the statistical probability of failures occurring. Understanding the interplay between physical and statistical factors can help in predicting and managing failures effectively.
In the foreseeable future, it seems unlikely that the random nature of bearing failures can be modified. There was a study conducted by a bearing company, possibly 15 years ago, that showcased the variability in failure times among a group of about 20 bearings tested under the same conditions. The range in failure times was significant, with a possible 10x or more difference between the first and last failures. This data highlights the importance of implementing condition monitoring to predict failures, as there were no initial failures within the group. The study demonstrated that it is not possible to predict when a bearing will fail based on a set number of hours or years of operation. Replacing bearings before the first failure would result in wasting approximately 95% of their total lifespan and require multiple replacements over time. Looking ahead, there may be opportunities for improvement in bearing manufacturing, potentially through tighter tolerances to cluster failures under identical operating conditions.
It is important to note that while the degradation mechanism cannot be altered by external factors, we do have the ability to impact the rate at which degradation occurs. For instance, in cases where failures occur randomly due to factors such as bird strikes on aircraft, the underlying mechanism remains unchanged. However, we can implement various measures to reduce the frequency of these occurrences, although the intrinsic randomness of events cannot be eliminated. In situations where physical degradation is a result of wear and tear, such as with clutch plates, tires, general corrosion, or exchanger fouling, steps can be taken to extend the lifespan or efficiency, but the fundamental nature of the mechanism remains the same.
One exception to this is when there is a consistent pattern of failure at the equipment level, which is often due to a combination of multiple failure modes leading to equipment breakdown. In such cases, proactive maintenance and design adjustments can be made to address specific failure modes and potentially shift the pattern towards gradual wear and tear.
Looking towards the future, there may be opportunities for improvements in bearing manufacturing processes. Tighter tolerances could potentially lead to a more clustered pattern of bearing failures, as issues often arise from installation, lubrication, alignment, or stray currents rather than inherent manufacturing defects. It is worth noting that many bearings do not reach their intended 'design life' in real-world applications, suggesting that efforts to improve performance should focus more on user practices rather than manufacturing processes alone. Remember, when it comes to complex tasks, like brain surgery, it's best to leave it to the experts!
Joe, The mechanism of degradation is immutable to external forces, but we can impact the degradation rate. For instance, in cases where failure occurs randomly due to physical factors such as bird strikes on aircraft, the underlying mechanism remains unchanged. While we can implement various measures to reduce the frequency of occurrences, the inherent unpredictability of timing persists. Similarly, in situations where physical degradation is due to wear and tear, like with clutch plates, tires, corrosion, or exchanger fouling, we can implement strategies to extend the lifespan or scale factor. However, the fundamental degradation pattern cannot be altered. Only when equipment failures exhibit a constant hazard pattern, resulting from a combination of different failure modes, can proactive measures be taken to mitigate some failure modes and transition towards wear and tear as the predominant pattern.
Regarding the manufacture of bearings, there may be potential for improvement if tighter tolerances are applied. Failures in ball and roller bearings are typically attributed to installation errors, inadequate lubrication, misalignment, or stray currents, rather than flaws in the manufacturing process. It is uncommon for these bearings to reach their intended 'design life' in real-world applications. Therefore, the focus should be on enhancing practices at the user level rather than solely relying on manufacturers to address these issues.
- 23-07-2024
- Shawn Thompson
Quote: Originally shared by Vee: The six failure patterns are derived from a comprehensive analysis of failures within the aviation industry, specifically referencing data from 12 years of failure records at United Airlines. While these patterns may not be universally applicable across all sectors, they have significant relevance in the context of aviation maintenance strategies. This research stems from the well-known bathtub curve concept, which was further developed through a detailed examination of aviation failures, resulting in the identification of six distinct failure patterns. This seminal work by Nowlan and Heap laid the groundwork for what is now known as Reliability Centered Maintenance (RCM).
It is intriguing to note that a significant portion of failures, approximately 68%, fall within the last failure pattern characterized by both infant mortality and random failures. However, the specific breakdown between infant and random failures was not documented. Furthermore, the root causes behind these failures remain unclear – whether it be inadequate lubrication, human error, modifications, or other factors. While Nowlan and Heap successfully categorized the potential behaviors of various aircraft parts in failure scenarios, the exact reasons behind the prevalence of random and infant failures remain elusive.
It is evident that with the right tools, technology, and knowledge, these patterns can potentially be influenced and altered to improve overall reliability and maintenance practices within the aviation industry. Warm regards, Rolly Angeles.
Rolly, it appears that there may be some confusion in your understanding of the facts. It is claimed that approximately 68% of failures fall under the last failure pattern, involving both infant mortality and random failure. However, it should be noted that there are no wearouts in pattern F. Additionally, the statement that all parts of a Boeing 747 or aircraft were studied is inaccurate, as the 747 did not exist at the time. The percentage breakdown between infant and random failures was not specified. It is important to verify the source of such claims, as it is the first time I am hearing that researchers were 'amazed' by this discovery. From my recollection, the infant mortality component was around 2%, but I stand to be corrected. Nowlan and Heap did make efforts to understand the physical nature of failure, unlike the TPM source cited, which seems to make unsupported statements. The basics, such as lubrication and potential modifications, were considered in their analysis. Nowlan and Heap extensively categorized the behavior of aircraft parts upon failure, but the specific causes of random and infant failures were not definitively determined. It is essential to thoroughly review their work before drawing conclusions. The majority of failures exhibited a constant hazard pattern, with only a small percentage attributed to infant mortality. Therefore, it is unlikely that wearout, a physical degradation process, played a significant role. The airline industry maintains meticulous records, prioritizes failure prevention, adheres to strict inspection protocols, and is closely regulated, making it improbable that basic TPM approaches would be disregarded. While it is suggested that failure patterns can be altered, it is ultimately up to individual interpretation. It is crucial to ensure that information is accurately understood before forming opinions.
Quote: It is important to note that pattern F does not experience wear outs. In fact, about 68% of failures are attributed to the last failure pattern, which combines both infant mortality and random failure. According to John Moubray's RCM study on a civil aircraft, different failure patterns were identified, with pattern F accounting for the highest percentage of failures at 68%. This pattern initially shows a high incidence of infant mortality failure that gradually transitions to a constant or slowly increasing probability of failure.
Please provide the source of your statement, as it is the first time I have heard of anyone being 'amazed' by this information. While it is believed that the infant mortality component in pattern F is around 2%, it remains unclear what portion of the failures are attributed to infant and random failures separately since pattern F is a combination of the two without any wear out mechanisms.
It is worth noting that the airline industry maintains accurate data, strives to eliminate failures in design, follows strict inspection and compliance regulations, and is heavily regulated. Disregarding basic Total Productive Maintenance (TPM) practices in such an industry seems implausible. While the industry's maintenance practices may have been weak in the 1950s, improvements have been made over the years, as evidenced by the decrease in failures and crashes per million takeoffs. The adoption of RCM by the airline industry was a response to the weaknesses in their maintenance structure experienced in the 1950s.
I share my experiences and insights on this forum to learn and grow. My observations are based on personal experiences and knowledge. It is important to acknowledge that there are exceptions and discrepancies between theoretical principles and real-world applications. I have encountered instances where parts fail randomly, but through thorough analysis and fractography, their lifespan has been significantly extended. While I see flaws in RCM, particularly in its decision diagram which may not suit all industries, modifications can be made to tailor it to specific needs.
Although TPM has its own shortcomings, the focus remains on continuous improvement. There is no one-size-fits-all solution that can be universally applied to every plant. Warm regards, Rolly Angeles, Teacher.
I sincerely apologize for unintentionally causing you distress, Rolly. Please accept my apologies for any discomfort my actions may have caused.
In agreement with Rolly's insight, I ponder: Can a part's pattern evolve from random to age-related post modification? If so, should our aim be to transform patterns of random or infant mortality failure into manageable wear out or age-related patterns? This was our original hypothesis. I foresee lively debate and wish to further complicate the discussion. Random failure patterns can stem from various causes, some discussed and some not. Here are a few reasons: 1. The degradation rate varies widely. 2. Degradation is influenced by multiple factors (such as starts, stops, and run hours). 3. The initial failure trigger is often a random event (installation issues, lack of lubrication, contamination, shock loads, etc.). 4. The X-axis may be misleading - cracks in a pressure vessel attributed to time instead of cycles, ball valve wear linked to process flow rather than stops and starts. 5. Components may have multiple failure mechanisms. The data available may not provide enough detail to determine the factors at play and their significance.
Suppose I work as an engineer at a Photo Film company, operating equipment that applies light-sensitive emulsion to paper. The line runs in the dark, and each bearing failure on the 1200 rollers costs me 10 hours of downtime and $30,000 in scrap. With an MTBF of 10 years for the bearings and 2400 in total, I'd face five failures per week, jeopardizing the business. The solution? Transforming random failures into wear-out patterns by investing in quality bearings, precise installation, and preventive measures. A shift to age-related patterns can ensure smoother operations.
Transitioning to a gas plant job, I encounter valve failures during shutdowns, extending downtime and incurring the boss's ire. Though data suggests random failures, I discern wear as the root cause, occurring during start-ups and shut-downs. By refocusing the analysis from time to starts, the wear becomes evident, enabling proactive measures and resolving the issue. Through these examples, we see the transformative potential of altering failure patterns. A stimulating journey indeed. Food for thought always. Regards, Steve www.pmoptimisation.com.au P.S. Ever wondered why aircraft companies log take-offs, not landings? Answer: Embarrassing data entry - explaining 100 take-offs but only 96 landings to the boss can be tricky.
In a famous quote by Steve Turner, he humorously highlights the challenges of data entry - imagine having to explain to your boss why you have recorded 100 take offs but only 96 landings. Is it a case of the Bermuda triangle, or the 911 effect?
- 25-07-2024
- Jasmine Howard
Thank you for your helpful responses, Eugene.
- 25-07-2024
- Victor Thompson
It is likely that the number of landings surpasses the number of takeoffs due to the higher quantity of components. My apologies for the dark humor, but I couldn't help but make the observation.
Are you wondering if it's possible to change a pattern from random failure to wear out failure? This could be a game-changer in the field of reliability maintenance. Join the discussion on this intriguing topic and explore the untold story of failure patterns. By altering patterns, we may discover new solutions and opportunities in our profession. Let's delve deeper into the world of reliability-centered maintenance and explore the possibilities of improving reliability by changing patterns. From identifying metallurgical aspects to addressing human factors, there are various factors to consider when tackling random failures. By implementing strategic modifications, such as secondary filtration or impeller screens, we can shift from random failures to wear out patterns. While achieving 100% wear out may be challenging, making small changes can lead to significant improvements in reliability over time. Let's continue to push the boundaries of reliability maintenance, one step at a time. Warm regards, Rolly Angeles - Educator
- 25-07-2024
- Vanessa Carter
Hello Rolly, I must commend your dedication and persistence in promoting your "random wear out" principle, despite facing challenges in gaining wider acceptance. History has shown that persistence in innovation is crucial for progress. I wholeheartedly agree with your emphasis on continuous improvement, which is essential not only in our professional lives but also in our personal lives. It is evident from your previous posts that you value spending quality time with your family, just like myself.
However, I have reservations about the effectiveness of turning back the clock, as your principle seems to suggest. The emergence of proactive (predictive) maintenance has revolutionized the industry by enabling us to foresee potential failures through advanced instrumentation. Relying solely on wear out timings for maintenance programs is limited, as failure patterns can vary among identical components. It is a fundamental truth that predicting the exact moment of failure is beyond our capabilities due to the unpredictable nature of machinery.
While I appreciate your efforts to promote your principle, I believe it may have a fundamental flaw that could hinder its widespread acceptance. Nevertheless, I wish you the best in your endeavors. Take care, Mike.
Quote: It appears that there is a fundamental issue that may be difficult to overcome. Hang in there, Mike... When facing a reliability issue with data indicating random failure patterns, it is crucial to explore alternative solutions. One common approach is to investigate the mechanisms of failures to determine if they are truly random. In my previous post, I outlined five potential reasons for failures appearing random. I believe there is potential to shift from random failures to age-related patterns, with necessary modifications. Real-life examples, such as a company implementing safe life strategies for bearings and a misinterpreted data scenario with a shut-down ball valve, support this viewpoint. Like Rolly, I am skeptical of attributing random data solely to random occurrences and believe modifications can alter this perception. It is possible that Vee and Mike share similar beliefs, but from my perspective, not all failures with random data are truly random (as randomness applies to various distributions) and can be influenced by adjustments. Regards, Steve at www.reliabilityassurance.com.
I completely agree with Steve on addressing reliability issues by utilizing tools for further investigation and problem-solving. However, when it comes to maintenance strategies like RCM exercises or PM optimization, I find it hard to accept the notion of completely changing established algorithms. It seems like Rolly is proposing a shift towards a more proactive and planned maintenance approach, which may contradict the trend we've been following for the past 30 years. Apologies if I've misunderstood the concept. Regards, Mike.
Mike, it seems we are discussing different aspects here. In many cases, like when implementing PMO or RCM, we typically assume that equipment failures occur randomly. This is often due to uncertainty about when a particular component was installed and how many hours it has been in operation. In such instances, when failures occur, we may opt not to perform preventive maintenance or instead focus on condition-based maintenance strategies. However, if the failure involves moving parts, we usually conduct a more in-depth investigation. Sometimes, we discover unique patterns in these components, leading us to identify issues like infant mortality and other contributing factors. While reliability issues may not be rare, they are more common than one might think. Research by Nowlan and Heap revealed that 11% of failures were related to wear out parameters. In reality, this percentage is likely higher, especially with improved data and efforts to address reliability issues like infant mortality. It's true that human involvement can result in random failures, but with a growing emphasis on quality across industries, such failures can be minimized. As evidenced by our experience with the photo film maker, addressing reliability concerns can significantly reduce instances of infant mortality. Cheers, Steve
Hello Steve, going back to Rolly's point about properly maintaining our equipment through precision engineering, lubrication, and correct installation. It is suggested that by doing this, we can predict when components will wear out, as they near the end of their design life. This leads to the question: should we implement a preventive maintenance schedule based on the design life of specific components, such as changing bearing X every 5 years? In this scenario, should we eliminate predictive maintenance, as failures are now more likely to be age-related? Looking forward to your thoughts. - Mike.
It is important to understand the different causes of failures in order to address them effectively. Failures can occur randomly or begin at random times, known as memory-less distributions. This means that each event is independent of the previous ones, making the start time of failure unpredictable. While the term "random" may be misleading, it is essential to focus on the prevention of non-age-related failures.
For example, physical causes like race damage in ball bearings can lead to failures regardless of proper installation and maintenance. On the other hand, failures caused by random events, such as a block of wood damaging an impeller, can be corrected through specific actions like using a grated filter. However, it is crucial to understand that these corrective measures may not apply to memory-less distributions.
In cases where failures are non-age-related or memory-less, traditional methods may not be effective. It is crucial to differentiate between age-related and non-age-related failures and address them accordingly. By implementing best practices, the average age of failure can be increased, shifting the curve to the right. Ultimately, it may not be possible to convert a non-age-related failure into an age-related one, emphasizing the importance of understanding the unique characteristics of each type of failure.
Is it possible to predict the lifespan of bearings accurately based on design life alone? While the initial instinct may be to answer "no," the reality is that, in theory, it is actually a "yes" - albeit with some factors to consider. When it comes to addressing age-related random failures in bearings, the key lies in implementing hard time maintenance practices alongside proper engineering, lubrication, and installation techniques. While some may argue that bearings are unpredictable due to external influences, such as human touch causing premature failure, others believe that the lack of metal-to-metal contact means bearings themselves do not wear out. Instead, it is the contamination of lubricants that leads to wear and damage. In industries like gas plants, where monitoring hundreds of fan belts can be challenging, condition-based maintenance may not always be practical. Additionally, the lack of information regarding installation dates and run hours further complicates the decision-making process around bearing replacement. Therefore, the use of condition-based maintenance becomes crucial for accurately assessing the health of bearings and determining the optimal time for replacement based on design life and performance history.
Imagine the impact of reducing random causes and potential starting initiators in various scenarios, such as ensuring everyone uses proper gloves when installing bearings. As Steve mentioned, failures can stem from random factors or unpredictable start times. It's crucial to address these potential issues to prevent future complications.
In an intriguing twist to this discussion, there is a manufacturing company rumored to have found that running equipment until it fails is more cost-effective than implementing predictive and preventive maintenance. Instead of dedicating resources to maintenance, they have quick-response "crash" teams on standby to rush to repair sites in case of a breakdown. One can only assume that safety and environmental consequences are handled differently in such cases. It's safe to assume their inventory storage must be extensive.
- 25-07-2024
- Quentin Foster
Steve mentioned that discussing fan belts is more relevant in situations where condition monitoring may not be as practical, such as in a gas plant with a large quantity of belts. Belts, tires, clutch discs, and brake pads are prone to age-related failures, often with weibull shape factors exceeding 3.
Steve wonders what would happen if most of the random causes of failure were eliminated. While efforts are made to minimize bird strikes near airfields and install coarse gratings to protect sea water pump intakes, complex assemblies often have multiple causes of failure. Each failure may have a different cause or combination of causes, leading to unpredictable start times. Monitoring the condition of components is key to identifying potential failures early on. Even something as simple as a ball bearing is a complex assembly with various potential failure initiators. In contrast, journal bearings typically wear down due to age-related factors. JoeP is cautious about the future development of bearings that fail in an age-related manner, questioning whether it is truly an improvement. Vibration analysis remains a vital tool in predicting failures, regardless of the pattern of failure. The focus should be on the scale factor or Mean Time Between Failures (MTBF) to improve reliability and performance, rather than solely on the pattern of failure. Age-related patterns may offer predictability in failure times, allowing for more proactive preventive maintenance. However, the primary goal should be to increase the scale factor or MTBF to benefit the customer's business financially.
In the context of failure analysis, Vee emphasized the importance of the 'scale factor' or the age at which approximately 2/3rd of items have reached their end of life. This concept aligns closely with Mean Time Between Failures (MTBF), suggesting a strong similarity between the two terms. Can we infer that the scale factor is essentially synonymous with MTBF, or is there a subtle distinction to consider?
I agree with Vee in most cases. I also agree with Rolly to some extent that changing patterns can add more value than just focusing on scaling up. In situations where Condition-Based Maintenance (CBM) is not feasible, it is important to address the issue rather than accepting it as the norm. Choosing, installing, and maintaining bearings for a designated lifespan is crucial. Simply increasing the scale factor may not be the most effective solution. It ultimately comes down to determining the most cost-effective and practical approach. Total Productive Maintenance (TPM), Reliability-Centered Maintenance (RCM), CBM, and "horses for courses" (HFC) are all strategies to consider. Is it ever not beneficial to alter the failure pattern, or have I overlooked something? - Steve
In a scenario where a random failure is age-related, the importance of predictive maintenance (PdM) should not be underestimated. PdM, along with condition-based maintenance (CBM), goes beyond simply monitoring failure patterns. It is a crucial step towards achieving proactive maintenance practices. By analyzing data from vibration and oil analysis, the true cause of problems can be identified. For example, consider a ball bearing's lifespan, which is determined by the number of revolutions before deterioration sets in. In a reactive maintenance approach, the bearing is only replaced once it fails, leading to downtime and unexpected costs. By implementing preventive maintenance, scheduled replacements are made assuming an age-related failure. However, this approach can lead to premature replacements or unnecessary replacements of bearings in good condition.
On the other hand, predictive maintenance involves monitoring the condition of the bearing using non-destructive techniques like oil and vibration analysis. This enables accurate detection of potential failures and allows for scheduled replacements, preventing surprises for maintenance teams. Moving towards a proactive maintenance mode, where PdM data is utilized, can further extend the component's life cycle. By analyzing data and addressing issues like contamination, the bearing's true fatigue life can be achieved. This not only minimizes downtime and costs but also maximizes savings in the maintenance function. In conclusion, the key lies in maximizing the component's life cycle to achieve efficiency and cost-effectiveness in maintenance practices.
In reality, the pattern itself is not as important as the 'scale factor' in determining the reliability of an item. The scale factor, or the age at which nearly 2/3rd of items fail, is crucial for improving performance and ultimately saving the customer money. While age-related patterns can help predict when failures may occur, the focus should be on increasing the scale factor or MTBF to enhance reliability. Paying attention to patterns is important too, as it can help maximize the lifespan of a part and save costs for the company. It is essential to strike a balance between the two aspects. Warm Regards, Rolly Angeles, Teacher.
In response to Rolly's quote, it is crucial to consider not only the importance of patterns in determining the lifespan of a part, but also how effectively utilizing the entire lifespan can ultimately save the company money. It is essential to understand that the average lifespan of all items can be improved through continuous improvement efforts, leading to an increase in Mean Time Between Failures (MTBF) and ultimately improving uptime and profitability. The failure pattern primarily influences the selection of maintenance tasks, rather than determining the average life of the part. It is important to tailor maintenance tasks based on whether the pattern is age-related or not, as this can help in deciding the appropriate approach. Age-related patterns guide age-based maintenance tasks, while non age-related patterns may call for Predictive Maintenance (PdM) or other actions depending on the consequences of failure. Ultimately, it is essential to recognize that age-based maintenance tasks alone are not sufficient, as demonstrated by N&H's findings that only 11% of failures in their industry were accounted for by age-related factors. It is a misconception to believe that age-related failures are superior to other patterns, a concept that seems to be shared by Steve, Mike, and Joe.
I haven't been keeping up with this discussion, but something caught my eye and got me thinking. The important takeaway is that failure patterns don't necessarily dictate the overall lifespan of a product, which ultimately impacts profitability. Do you have any additional factors to consider, Vee? It's crucial to remember that the average lifespan is only as reliable as the timeframe analyzed, the demographic being studied, and how closely it aligns with the operational environment. Perhaps I've missed the main point here, do you have any insight to share?
Hi Vee, I'm also finding it difficult to grasp Rolly's concept. -Mike
Daryl asks: What are the caveats that should be considered in evaluating the average life? The accuracy of the average life depends on factors such as the time period studied, the size of the population analyzed, and the similarities in the operating environments. In previous discussions, I have highlighted the importance of considering independent and identical conditions, as well as operating contexts and the use of approximations.
Hey everyone, I'm going to take a quick break but I'll be back soon. Feel free to join in by sharing your photos, Vee, Steve, Mike66, Joe, and everyone else. Thanks for joining!
Thank you, Vee, for your patient response and for directing me to the relevant posts. I had not yet gone through the details and was seeking clarification on a specific statement. I appreciate your help in clearing things up.
In the realm of business, it is crucial to understand that the failure pattern of a product does not necessarily define its average life span. This distinction is key as it directly impacts the profitability of a venture. If you are interested in delving deeper into this concept, I recommend exploring resources that discuss the relationship between product lifespan and financial success. Feel free to reach out if you require further clarification. Best regards, Steve.
Steve, you bring up a valid point, and I admit my mistake. I got too caught up in enthusiasm without much substance. However, in the scope of our discussion, we have dedicated significant time to debating the benefits of transitioning from a non-age related focus to an age-related one. I believe we may be focusing on the wrong aspect; our primary goal should always be to enhance reliability constantly. This can be achieved regardless of the failure pattern by simply paying attention to details, enhancing cleanliness, work quality, and compliance, essentially following through with our initial responsibilities. Effective practices do not necessarily require complex statistics or reliability theory, although having that knowledge can provide insight into why events unfold as they do.
Our aim should be to reach 'intrinsic' reliability from the current 'achieved' level, which often represents only 25%-30% of the intrinsic value. For instance, many organizations have increased pump Mean Time Between Failures (MTBF) from 10-11 months to 36-48 months by executing fundamental practices correctly, without involving Reliability-Centered Maintenance (RCM) or other acronyms. For individuals familiar with Weibull terminology, adjusting the scale factor, rather than the shape factor, can significantly impact performance. For those unfamiliar with Weibull, simply view the scale factor as MTBF and a shape factor of 1 as non-age related, while higher shape factors indicate rapid wearout and strong correlation with age.
Hello Vee, Mike, and All,
In a previous post, Vee mentioned that patterns help guide maintenance task selection. Age-related patterns lead to age-based tasks, while non age-related patterns may call for PdM, redesign tasks, or inaction if failure consequences are minimal. Wear out patterns simply indicate when age-based maintenance tasks should be applied. N&H demonstrated that relying solely on age-based tasks was illogical as they only accounted for 11% of industry failures. Despite this, there seems to be a misconception among some, including Steve, Mike, and Joe, that age-related failures are superior. Nowlan and Heap's study caused a shift in maintenance practices by highlighting the six failure patterns applicable to all parts, revealing that scheduled maintenance may not significantly impact item reliability unless a dominant failure mode is present.
Over the years, I have educated over 2000 individuals in my country on maintenance concepts, yet only a small percentage are aware of the six failure patterns. This lack of understanding often leads industries to reactive maintenance approaches, resulting in excessive maintenance work. Many organizations fail to grasp these patterns, leading to unnecessary reactive maintenance instead of investing in predictive maintenance solutions like thermography.
The late John Moubray often emphasized that many maintenance practices are intrusive. Steve, you have conducted PM Optimization, which involves evaluating maintenance tasks to address failure modes. Following PM Optimization or RCM, maintenance activities may decrease, reflecting more efficient practices. It is possible to modify the failure pattern from infant to wear out or random to wear out, thereby potentially reducing costs for the company.
Implementing scheduled replacements or overhauls during wear out modes can extend the lifespan of parts and minimize unnecessary replacements. Simple modifications, like adding screens to prevent foreign object damage, can significantly reduce maintenance costs. Improving MTBF and understanding the scale factor further enhances cost savings in maintenance operations.
Observing the disparity between the airline and land industries, it is evident that a lack of understanding and application of patterns plays a role in the maintenance efficiency of different sectors. Both the scale factor and understanding patterns are crucial in optimizing maintenance practices. I invite you all to share photos in the members' section.
Warm Regards,
Rolly Angeles, Teacher
When completing PM Optimization or RCM, there is typically a decrease in maintenance activities. This reduction, often around half, is quite remarkable. In certain instances, the formal program may show an increase, but this is usually due to previous informal work. Therefore, despite what the system may indicate, the actual workload has not increased. Regards, Steve at www.pmoptimisation.com.
Hello everyone, it may be beneficial to shift your perspective on condition monitoring. We implement condition monitoring because we are uncertain about when equipment will fail. If we were able to predict the exact moment of failure without condition monitoring, we would not need this process. By utilizing methods that predict failures, we can reduce maintenance costs. Failures often appear random because we lack a comprehensive understanding of the underlying causes. However, these patterns can be analyzed and broken down into identifiable factors, some of which may be related to the age of the equipment. By addressing and eliminating random causes, we can pinpoint age-related failure patterns that have a predictable lifespan. This is why I agree with Rolly's approach. We adopt this method as necessary. Regards, Steve from reliabilityassurance.com.