The simpler the question, the more difficult it is to provide an accurate answer, and sometimes, answering “is this bearing faulty?” is not easy.
Firstly, “is the bearing faulty?” is a bearing condition recognition problem, with the object of recognition being “faulty”. The most complex part is defining what “faulty” means. Does it refer to a bearing that cannot rotate?
However, sometimes, in certain circumstances, a faulty bearing might still be able to rotate and some people continue to use it. In this case, is the bearing considered “faulty”? In this vein, some would ask: this bearing is not in great condition, but how long can it still be used? This then introduces another term, “residual life”. So, does a so-called “not so good” bearing with “residual life” count as a “faulty” bearing?
The overall issue remains, what kind of bearing is considered “faulty”? What is the precise answer?
It’s not hard to see that to answer if a bearing is faulty, one must first clearly define “faulty”.
For a normally operating bearing, its operational state is in accordance with the design intent. This state of “accordance with the design intent” is the normal state. Hence, it’s not difficult to define a state that is “not in accordance with the design intent” as an “abnormal” state.
In the above definitions and discussions, one point needs to be clarified: what is “design intent” and what represents it? For a bearing, design intent can basically be described as stable operation and flexible rotation.
Therefore, stable operation means no hindrance and smoothness during rotation. The best parameters to describe this state are vibration parameters. In terms of flexible rotation, low friction is an important manifestation, and friction manifests as heat, so temperature can characterize some manifestations of friction.
This is why in related standards for condition monitoring, for bearings, there are always vibration and temperature signals. This means that vibration and temperature can be used to characterize the design intent.
For a normally designed bearing, the vibration and temperature levels can be determined during operation. Once the vibration and temperature of the bearing exceed the levels in the normal state, it can be said that it is in an “abnormal” state, also referred to as an exceptional state.
The question arises, does “abnormal” mean “faulty”? To answer this question, we need to return to the definition of “fault” itself. Readers can certainly refer to the definitions in relevant international standards, but the overall definition of “fault” includes two aspects:
1. The performance of the equipment is reduced, but it can still operate;
2. The equipment is completely inoperable.
The first condition is functional degradation, and the second is functional loss. Both fall within the range of malfunctions defined by standard specifications.
Viewed from such a definition, it can be considered “broken”. Internationally adopted standards for vibration and temperature provide the benchmark for determining if something is “broken”.
Though strictly defined, in practical engineering, minor “malfunctioning” scenarios always emerge. That is, the bearing has minor issues, but can still operate for a relatively extended period. Many practical applications, due to cost considerations, maintain the status quo. At this point, the bearing is “not good,” but is not categorized as “broken”.
Here, the definition of broken is determined by onsite personnel based on actual conditions. Under these circumstances, it’s challenging to apply a universal standard to judge whether the bearing is broken.
However, establishing a benchmark for when a bearing is “broken” is a balancing act between usage costs and future risk.
For instance, if a bearing has operated for 5000 hours and shows signs of vibrational abnormalities, each extension of usage time increases the risk of the bearing becoming completely incapacitated. Balancing the extension of time and the chance of failure to achieve an optimal value is the ideal time for maintenance.
Please note, the optimal maintenance time mentioned above is part of the optimization process of predictive maintenance, which could be elaborated further when possible.
Did you notice that a seemingly simple question, when answered accurately, involves complex concepts and logic, and requires various analytical methods? Onsite workers hope for simple prompts to handle issues, but the judgment logic behind these simple prompts becomes increasingly difficult.