Corus offers sound maintenance advice
26 Oct 2006
Scunthorpe, UK — When it comes to monitoring the condition of critical, slow-moving (0.25-80rpm), high capital value plant and machinery, companies really need to consider the benefits of acoustic emissions monitoring – particularly where there are fluctuating load conditions and where a breakdown would have a significant impact on production.
Many companies rely on vibration monitoring systems to check the condition of critical plant and machinery, but this method is not as accurate as acoustic emissions monitoring, especially when it comes to slow speed, fluctuating load applications.
It’s horses for courses really. Vibration monitoring is fine for high speed motors, fans and pumps. The equipment is cost effective, easy to use and the data is easier to interpret.
However, where components or machines rotate at less than 80rpm and operate under fluctuating load conditions, or only move through a part revolution, it is more difficult to collect meaningful data from methods such as vibration monitoring.
Acoustic emission monitoring equipment has a high sensitivity to machine faults but is also immune to audible noise and low frequency background vibration.
The problem is that many engineers are not fully aware of how acoustic emission monitoring systems can help them reduce plant maintenance costs and improve machine availability. Also, many companies simply do not possess the necessary skills in-house to interpret the data from acoustics emissions monitoring so they continue to use vibration monitoring or other devices.
CNES offers non-intrusive condition monitoring techniques to companies in industries outside of steelmaking, including aerospace and defence, automotive, food and beverage, oil and gas, civil engineering and chemicals.
Depending on the application, CNES uses thermal imaging techniques; vibration analysis; balancing and laser alignment; contamination monitoring of hydraulic systems; rotor bar testing; remote visual inspection; acoustic emissions monitoring; and debris analysis of lubricants and hydraulic fluids.
Monitoring acoustic emissions is certainly not a new method of monitoring high capital plant and machinery. The technique has been around since the early 1990s, but more manufacturing companies should be using it.
Acoustic emissions are the high frequency stress waves generated by the rapid release of strain energy that occurs within a material during crack growth, plastic deformation or phase transformation. Acoustic emission monitoring systems use surface-mounted transducers to detect these stress waves, which lie within the 25kHz to 1MHz frequency range.
CNES recently helped Corus’ Scunthorpe plant to save at least five weeks’ production time. The plant uses three Basic Oxygen Steelmaking (BOS) plant vessels, which convert molten iron from the blast furnaces into steel.
Each vessel weighs 1,300 tonnes, supported on two, 1.75m diameter spherical roller bearings. Each vessel is able to tilt to various angles up to 150° from the vertical during the process for charging and discharging purposes. The arduous duty generates high local contact loads, which over time, generates high residual stresses in the tilt bearings.
Delivery lead time for bearings of this magnitude can be as long as 12-18 months, with plant downtime for a planned bearing change requiring three weeks. The bearing cost of £60k-£80k per unit becomes almost insignificant in comparison to the overall production outage costs.”
For such critical equipment, it was essential that Corus knew the exact condition of the tilt bearings and the trends that could predict their life expectancy. The method Corus developed to do this, ‘Aquilla AE Pro’, is based on acoustic emission principles.
The software was developed by Corus Research Development & Technology, a division of 900 engineers and technologists with expertise in data analysis, condition monitoring and failure analysis, based in laboratories in the UK and The Netherlands.
Working with bearings manufacturer Schaeffler (UK), the existing bearings were removed and the new bearings fitted. This took four days of the three-week planned outage, before the panels and screens were refitted, the supporting frame removed and the plant recommissioned.