At General Electric?s technology centre in Bangalore, researchers want to look inside an engine and examine its metal parts to be able to predict its health. But they don’t need to take it apart to do that. For the American company that has an installed fleet of over 6,000 steam turbines used for power generation globally, the solutions to that problem could mean a huge business benefit, potentially helping to extend engine service intervals by predicting accurately and non-destructively.
It?s also the single biggest research project at GE?s John F Welch Technology Centre (JFWTC) in Bangalore, the company?s largest research unit outside the US, where a team is developing new prognostics technologies based on the data from engines running across the world.
?Today, we can predict stuff. What we are thinking is from these signatures, from these sensors, can you predict more than what you know today?? explains Mano Manoharan, General Manager of Global Research at JFWTC. ?What we are looking at is, can you non-destructively evaluate the health of a machine, a complete machine.? GE?s research push into ?machine health? is a global, multidisciplinary project and Bangalore is leading pieces of it, like monitoring the health of engine blades using data from engines that are under contractual service agreements. ?The traditional way we used to measure whether a material was good or not you would cut it up, look at the section and see. Can you non-destructively evaluate the health of a part, lets say the blade inside without taking the whole machine apart, and see what is the potential remaining life of the blade. This is not known today,? says Manoharan. He added that they were looking at gas turbines and steam turbines now and perhaps wind turbines in the future. ?We are looking at transport related but those are tougher problems.?
The attempt is akin to the structural health monitoring capability being developed for the aviation industry using sensors which allows for real time analysis of an aircraft structure to increase safety and reduce maintenance costs, except that the complexity increases in a turbine because of its rotating parts. More so in the case of an aircraft engine that operates in varying conditions.
?That?s why they want to prove the technology on the land-based turbines,? says A C Raghuram, former head of the Failure Analysis Group at the Bangalore-based National Aerospace Laboratories, explaining that techniques such as acoustic signal monitoring are now reasonably well understood for studying cracks on aircraft structures. ?Mainly, failure is because of crack propagation.?
Even a one year extension on a turbine with a life of about 30 years, and whose service intervals could run into weeks, represents a substantial benefit, says T A Abhinandanan, Professor at the Department of Materials Engineering at the Indian Institute of Science (IISc). ?There are models people are developing. Sometimes it’s no fancier than a co-relation.? Eventually, according to Manoharan, the ongoing lab-scale research should be able to say when a part in a real machine underwent a certain cycle based on the signature and when it needs maintenance.
?That is why it is a big deal for us. We have a huge fleet of gas and steam turbines. So if we can add value, the customer can use it more efficiently. Or we can provide the right upgrade to them, then they could become more efficient…that’s what these things will be able to do,? says Manoharan. He added that JFWTC plans to begin discussions for collaboration with institutions such as IISc and the National Metallurgical Laboratory in Jamshedpur to better understand materials. ?We are beginning to think about that route. We are beginning those discussions.? Apart from `machine health’, some of the research at JFWTC includes the development of software for optimization solutions such as improving the fuel economy of a locomotive.