Lean Propels Turbine Engine Production
Sustaining--and growing--lean philosophy at Pratt & Whitney
By Jim Destefani
Editor's note: Founded in 1925, Pratt & Whitney (P&W; E. Hartford, CT) today manufactures nearly half the jet engines used in the world's commercial aircraft fleet as well as military engines that power the US Air Force's front-line fighter aircraft and numerous air forces around the world.
A division of United Technologies Corp. (UTC; Hartford, CT), P&W produces engines and engine components in older facilities, using skilled labor coupled with high-tech and often proprietary processes. Facilities include five "mega-campuses," each of which encompasses more than 1 million ft2 (93,000 m2) of floor space, as well as 40 overhaul, repair, and service centers around the globe.
P&W launched its lean manufacturing initiative in the early 1990s. Since then, the company has run thousands of kaizen events in its various plants, and has delivered year-on-year cost reductions in overall operations in each of the last four years.
Kip Wyman is P&W's Chief Manufacturing Engineer. Part of his role is sustaining and growing the company's lean manufacturing efforts. He recently talked with Manufacturing Engineering about those efforts.
ME: How long has there been a lean manufacturing initiative at P&W?
Wyman:We've been actively pursuing lean manufacturing since the early 1990s. We started in our North Haven, CT, facility, at our Turbine Airfoil [turbine blades and vanes] Division--the highest-volume plant at P&W. The plant manager there realized that, with the volume that plant had, he had a real opportunity to increase capacity and lower costs. He took on the challenge of implementing P&W's first lean cell.
ME: Can you tell our readers about this plant's operations?
Wyman: This facility is very vertically integrated. We bring in the turbine blade castings, then do all the required machining and secondary processing. That includes coating, drilling, nondestructive testing, final inspection, moment weight analysis...all the things that go into making a turbine airfoil. Then we deliver completed parts to our assembly floor.
In North Haven, at one time we had 20, twelve-axis, dual-spindle creep feed grinders for airfoil production. Those machines were fully automated, palletized, with full toolchangers. They were just monsters. It typically took one operator to run a machine and two skilled tradesman to keep it running.
The plant manager realized he couldn't continue this massive, full-automation strategy that everyone jumped into in the mid-1980s, so he challenged the work force to come up with a lean grinding cell. Working with our own machine repair organization, we designed and manufactured the first two lean grinders ourselves. They were very lean machines, very bare-bones, and with them we put our first cell together. It was a learning cell, and we had some success, which triggered another cell.
That cell was enough to convince us to abandon the mass automation process for airfoil grinding. So we decided to recapitalize all of our airfoil grinding operations into lean grinding cells.
ME: Why did you build those first machines in-house? Was there nothing in the market that fit your requirements?
Wyman: This was an employee-driven initiative, and our maintenance organization wanted the first opportunity to try to provide a machine to fit our specialized grinding requirements.
There were machines in the market that could've been modified to fit, but we decided to design and build the first machines ourselves from the ground up.
These were very simple machines, not elaborate by any means. But when we recapitalized the entire grinding operation, we selected a tried-and-true supplier (Blohm) who modified standard machines for our application. We wound up purchasing 38 of them.
ME: So, in terms of sheer numbers of machines, you can't say that switch was lean. What did the plant gain from the new process?
Wyman: We did increase the number of machines, but we went from twelve-axis machines to three-axis and five-axis machines. This initiative was successful enough that it allowed us to implement lean across the North Haven plant product line. That led to a 28% reduction in floor space--enough to actually consolidate the plant with another plant in E. Hartford, CT. So we actually shut down the North Haven facility in 2001.
We also recapitalized a good portion of our coating operations. Most of the coatings we produce are thermal barrier coatings, but we have some wear coatings as well.
ME: Wouldn't it be harder to lean out that type of operation?
Wyman: One of our biggest challenges was, first, we were dealing with old facilities and we weren't going to build new ones. We decided to invest our capital in equipment, not buildings. So, moving stuff around and reorganizing was tough, and requirements like utilities, electrical demand, and chilling capacity presented quite a challenge.
Also, some of our processes involved very large batches and very long cycle times. Some heat treating processes take more than two days. Many of our coating processes take hours, not minutes. They process multiple parts at a time. So we had to basically design around some of those highly specialized processes, many of which are very specific to P&W. We own the patents, we own the technology, and one of the differentiators between Pratt & Whitney and our competitors is our ability to run our turbines hotter and faster than anyone else. So those processes were a big challenge for us.
Another challenge was they're very large. They take up enclosed rooms on the manufacturing floor and they require environmental controls. We operate acid lines, coating lines, strip lines--things that are not necessarily the friendliest environmentally. So they require special rooms, special controls, special environmental barriers around them. When you're trying to go to single-part flow or put in a cellular concept, that's a very big challenge.
ME: Were there challenges to lean implementation on the human side? Was there resistance?
Wyman: You'll always have nonbelievers until you get some momentum. But the initial cell was a grass-roots effort, developed by a cross-functional team of skilled trades, manufacturing engineers, and operators, and it really was the operators that made the cell successful.
When we started migrating the lean initiative across the enterprise, the workforce responded cautiously. As soon as we got some results, people started seeing the quality increases and reduction in lead time, and more and more people came aboard.
So I wouldn't say the workforce was a barrier. They were a little cautious until they saw some movement, and they didn't know what the end result would be. But that was true with a lot of us until we actually started going in the lean direction. But now, it's part of the culture at P&W.
ME: How did the hourly workforce react to the closing of the North Haven plant?
Wyman: Those types of things never go down well. But we saved many of those jobs; many employees were able to make the commute to East Hartford, about 35 miles away.
We've been very careful not to tie any job reductions to any of our lean manufacturing strategies. We avoid that at all costs, and we find jobs for anyone who is displaced elsewhere in our operations.
ME: How many kaizen events have you run since that initial work on the grinding cell?
Wyman:In Connecticut operations, which includes three manufacturing sites, we average three kaizen events per month. Usually each event is four days long.
ME: Can you tell our readers about some of the lean tools that you've used over the years?
Wyman: We actually use two main guides. Our parent company, United Technologies, has an overall continuous improvement plan called Achieving Competitive Excellence [ACE; see the article "Jetting to Success" in the May 2003 issue of Manufacturing Engineering]. It's been in place since the mid-1990s, and it encompasses both a manufacturing strategy and our quality strategy.
ACE is always evolving and growing, and it's our base toolbox. We learn as we go and adjust our standard work, we learn from other organizations, we bring in consultants on a regular basis. But ACE guides both our manufacturing plants and our business practices.
Two years ago, we introduced Modularization as the next step in our ACE journey. Lean is now fairly mature in all our component manufacturing operations, so modularization is the next step in terms of moving lean concepts upstream to our assembly operations. It's really how we intend to continue growing lean manufacturing at P&W.
Traditionally, we did all engine assembly at our Middletown, CT, facility. What we will be doing is assembling modules at the component plants, then delivering fully certified, subassembled modules to Middletown for final assembly.
We started modularization two years ago, but the kind of work we do requires a significant amount of government oversight, so change of this scale is difficult. It takes time, and there's a lot of testing and validation required. We're actually working on our first engine, which is a military product, the F-119 engine for the F/A-22 Raptor. We're in the process of developing specifications and buying our new assembly line. The modules, and which plants will have responsibility for their subassembly, certification, and delivery, have been identified. We've worked closely with our military partner and the government on this product, and we have high hopes for not only quality improvements but also lead time reduction.
ME: Modularization sounds a lot like initiatives at some aerospace and automotive OEMs. But they're pushing the detail work down to subcontractors, while P&W will continue to do the work in-house. How will you benefit from modularization?
Wyman: In aerospace, Boeing and Airbus are both trying to bring in more subassembled components. Airbus brings in complete wing and fuselage assemblies and just bolts them together.
We're going that direction, but on a smaller scale. There are actually five modules that make up a military aircraft engine, and we're going to bring them in fully sub-assembled and fully validated, and bolt them together in Middletown.
The challenge for us is, most of the activities that differentiate us from our competitors are things that we really don't want outside our walls. A good portion of the activity is tied to classified projects that you really can't share with anyone without going through a very long and complicated process, and most outside companies are not equipped to handle these sensitive contracts.
So even though we're keeping the work in-house, we expect modularization will benefit us in many ways. In Middletown, we run multiple products down the same assembly line. Our long-term goal is really to have our assembly floor no longer be assembly. We want to bolt together certified modules right at the test cell and go directly into test. Our goal is to deliver these modules and, within days, not weeks, have an engine bolted together, tested, dressed out, and shipped to a customer.
All the module centers--the component manufacturing facilities--now have design responsibility for the components and now for the subassemblies. They're designing a lot of efficiency into the modules to the point where now all we have to do at the test cell is bolt the flanges together and we're ready for test. So it's really pushing design for assembly, and responsibility for quality, down to the people that are doing the work. The same person who was designing something like, for example, a fan case, now also has the responsibility for designing how that part will be assembled to the intermediate case.
We plan to implement modularization on our first legacy engine this year. Our newest product line, which is just being introduced for the new Airbus A380 super-jumbo jet, was actually designed from the ground up in the modular format. It's a joint project with General Electric, and it will be built in Middletown and received in modules. [For more on the A380, and the manufacturing philosophy of Airbus Industrie, see the article "Automation Speeds A380 Wing Assembly" in the March 2005 issue of Manufacturing Engineering.]
Another thing we do that's unique at P&W is, we have proficiency levels that we introduced into our operations organization. There are two tracks: ACE quality, and ACE lean. Basically, every employee within P&W Operations has a track and a proficiency level assigned to them, and they can move up to a higher proficiency level by gaining experience and training. Levels go up to 4, which means they've not only practiced but they can teach. We try to use this to encourage people to continue to gain knowledge and experience.
Doing that required value stream mapping. Whether it's a business process or a manufacturing process, we start with value stream mapping as a base. We identify all the processes, the handoffs, the influences, the lead times. That's a good example of one of our standard lean tools.
ME: So value stream mapping provides a baseline for looking at any process. Can you take readers through what you might do next?
Wyman: We assemble a cross-functional team in the area we're going to look at. In manufacturing, it could be anything from a single-machine cell to a very complex one. Most of our processes are already value stream mapped, so we get the undocumented value stream map for the process and validate its accuracy.
Then we start at the beginning to identify each one of the handoffs, and identify every opportunity to reduce the number of handoffs. We identify every opportunity to eliminate waste, including inventory--anywhere material is queuing up, we see if we can design it out. We identify mistake-proofing opportunities by looking at the past quality history of the part and identifying the operations that caused defects. Then we do a failure mode and effects analysis (FMEA) process to try to identify mistake-proofing opportunities.
Eventually, we identify a new lead time. That would trigger the next steps in terms of layout of the new cell, workforce training, and validation.
ME: What do you want to tell our readers about lean in general or lean at P&W?
Wyman: For four years now, we've been able to deliver year on year cost reductions, which obviously is a big part of UTC's strategy to stay competitive. We work in a very technically challenging environment with a lot of unique processes, so it's always a challenge to balance technology and cost reduction.
We continue to look for opportunities to expand our lean initiative. We rotate our lean manufacturing experts through other United Technologies companies to share our best talent and share lean manufacturing concepts.
UTC has a manufacturing technology deployment council, which consists of me and my peers at other UTC divisions, where this is a standing agenda item. We go over what each group is doing in terms of ACE activities, and that's one way we develop standard work and share best practices.
Every component manufacturing site has a lean manufacturing expert, an ACE expert; that's a staff-level manager position. In the corporate organization, I have lean experts on my staff. We have a general manager of modularization. That's how high that initiative is on the agenda.
This article was first published in the May 2005 edition of Manufacturing Engineering magazine.