Sensors in Condition Monitoring: Making Optimal Decisions

Sensors and solutions for condition monitoring can be as varied in their architecture, capabilities and price range as vehicles are for transporting people. Just because a vehicle has a wheel doesn’t make it a car, and there is a big difference between a car that costs $100,000 and one that goes for $20,000. The differences are for good reasons. The most obvious variables are performance, form factor, functionality and which “bells and whistles” are included.

This kind of variety isn’t much different in the world of vibration sensors. Some are wireless, some are small, some are fast and some are slow. Some include algorithms, while others do not. An accelerometer is the most basic level of sensor and doesn’t include any of this — not even the amplifier that boosts its signal — so it is readable by industrial controllers.

When looking at vibration sensors, a prospective customer will want to understand what information will come from that sensor, how often they will get that information, what mathematic analysis has already been done or can be done with it and how long that sensor will last. For example, a wireless sensor may be cost-effective, but is simpler and slower than a wired alternative because the battery life becomes such a limiting factor. An astute prospective consumer will also find that some vibration sensing systems, as opposed to “vibration sensors,” are significantly more expensive because there are layers of mathematical analysis that can be performed on vibration information. The more complex the math and analysis, the more expensive one can expect the vibration solution to be. A quick search online reveals that people have spent their entire careers studying vibration. Tapping into their experience will not come cheap.

Vibration sensors are a common solution for customers to start with, but one customer in the food processing industry started their experience with condition monitoring using temperature and humidity sensors instead of vibration. After a few steam leaks raised the humidity in their control room to a level causing several VFDs to fail, they decided they needed to be more proactive. Their decision wasn’t based on the first event, but after three times over the course of 15 years. At $10k per event just in hardware, not including the labor, and up to 10 hours of downtime per event, they’d had enough. Perhaps the technology wasn’t really in place the first couple of times, but now their systems were in a place where implementing a solution was simple and inexpensive.

Simple and Inexpensive Solutions

Just as the customer started pondering possible solutions, one of our team members shared the Balluff condition monitoring sensor solutions with the customer. While vibration and temperature were the initial focus, humidity is what really caught their attention. The cost of $1,100 for a couple of sensors to monitor a 50-by-300-foot room was an inexpensive solution in the face of failures that had cost 10 times, if not 100 times, as much

Furthermore, the implementation was not difficult. The most challenging part was running cables across the room from the sensors to the IO-Link masters the customer already had in their system. They simply had to plug these sensors in and then program their existing controls to throw an alarm when the humidity value went too high. The effort was so minimal that they didn’t even have to take the system down as they could update the programming while the system was up and running.  After setting the alarm value at 60% humidity to start, the PLCs’ HMIs were set to show alarms on them while the PLCs would send out text messages to alert the supervisors. Even mounting the sensors was simple with their magnets easily sticking to the overhead crossbeams.

At that price and ease of implementation, they are already way ahead financially over even just one of the steam leak events with only factoring in the cost of hardware, not to mention the downtime. The customer now has confidence that they can rapidly respond to a condition and prevent an event that has cost them tens of thousands of dollars in the past.

The customer has already moved on to implementing the same sensors — without the humidity option — on several of their 50 Hp grinder motors. Their grinders are similar to large blenders, so the vibration is critical as the blades have caused problems before. Benchmarking the normal vibration level gives them a value that they can compare against. If the value goes higher, their PLC knows to throw an alarm to the HMI and texts in the same way as they use the same IO-Link master and controls architecture. They even already had the IO-Link master in place with open ports, so their cost was limited to the sensors themselves.

Vibration Analysis

Troubleshooting a large motor with vibration data can tell a knowledgeable person whether the failure is in the bearings, the windings or even what it is coupled to. Although, this data must include vibration in all three axes, X, Y and Z. It also must include frequency analysis. This means the sensors must be fast enough to catch the highest frequencies likely to be seen in those failures. The frequency and direction of the vibration data are key data in diagnosing the various failures.

For this customer’s grinders, however, they are focusing on just one failure mode which is not direction nor frequency dependent. An inexpensive vibration solution will catch this change, so an expensive solution is not needed. The first time the customer catches a vibration problem, they will be glad they have their simple and inexpensive condition monitoring solution in place,

The key to making the proper decisions on condition monitoring is to understand what kind of information is required, how often that information is needed and what kind of budget is in play. Then, after consulting with an expert, the most cost-effective and efficient solution can be selected.


Article featured in Processing Magazine

Printer-friendly version