Dow: Designing the Next-generation Railcar with TRIZ
By Sue Reynard
The interaction of process improvement and innovation tools can be seen in one of Dow’s projects – to design the next generation of rail tank cars for transporting the most hazardous chemicals.
“When we looked at this project, we recognized that traditional brainstorming and creativity techniques could not take us to the level needed in terms of improved safety and security,” said Henry Ward, Dow’s director of Transportation Safety and Security, Global Supply Chain. The railcar design is a long-term project with multiple partners. Launched in April 2006, as of April 2007 it was at the design concept stage, near the end of the Explore phase of Design for Six Sigma (DFSS). The project is led by a core team of 10 people from the five participating companies – Dow, Union Pacific Railroad (the largest freight carrier in the United States), Union Tank Car Co. (which has built many of the tank cars in Dow’s fleet), the Federal Railroad Administration and Transport Canada. To date, more than 60 people have been engaged on various sub-teams made up of design engineers, university researchers, contractors and specialists in emergency response, rail operations, rail maintenance, research and development, and logistics.
Combining TRIZ and DFSS
The benefit of interweaving the Theory of Inventive Problem Solving (TRIZ) with DFSS became apparent immediately, according to Black Belt Jennifer Trager, a Dow senior logistics specialist and the project manager overseeing the railcar team. “When you work through a house of quality – the design matrix done at the beginning of a QFD [quality function deployment] effort – you identify a lot of design or functional ‘conflicts,’ situations in which you traditionally face design trade-offs and may encounter compromises,” she said. Resolving conflicts is a key strength of TRIZ; the tools help teams find creative ways to provide the required functionalities while avoiding the conflict. “With TRIZ, we’ve taken the term trade-off out of our vocabulary,” Trager said.
For example, the team came up with eight acceptability criteria for the design options, including performance, size, carrying capacity and weight. Ward explained that in traditional railcar design, steel is added to improve strength for crash protection. But train track restrictions limit the maximum weight of a car. Another conflict was that adding steel reduces the amount of commodity that can be transported, he said. At this point, the team latched onto a core TRIZ principle: Don’t reinvent; use solution concepts discovered already. At Dow, that means not only accessing TRIZ databases to find conceptual solutions for similar problems from many other industries, but also maintaining databases of effective technologies and solutions used within the company.
The team learned that the Dow automotive business had already been challenged to develop materials that would provide crash protection without adding weight (for fuel economy) or using space (so cars could still be roomy). “Their criteria were exactly what we were looking for,” Ward said. The team realized that the structural foams the automotive group had developed could be used in the new tank railcar to provide the necessary crash worthiness with superior performance, good economics and less weight.
In another case of “stealing” from themselves, the team glommed onto the work a safety team was doing to develop mechanisms that allow constant monitoring of a railcar’s location and the condition of its contents. “The company is currently adding such devices onto existing railcars, and we’re going to integrate them into the design of these next-generation cars,” Ward said. That way, the company will be able to monitor the location and status of every railcar in its fleet and immediately detect any signs of tampering.
One of the key requirements to be addressed in the design was thermal protection of the tank cars in case of fire, such as in a derailment. But that type of insulation typically takes up a lot of space – another conflict. Once the team was able to define the technological problem – providing thermal insulation with minimal use of space – TRIZ databases helped them track down the solution in the aerospace industry. “With an inch of the material they use,” Trager said, “we can get the same thermal protection as 4 to 8 inches of the current material we use.”
The other main contribution that DFSS and TRIZ made to the railcar design process was forcing the team to look at the design functions separately. As Ward explained, in current railcar design the tank serves three functions: It carries the load of the train; it has to contain the commodity (acting as a pressure vessel); and it has to protect the tank from external forces, such as in a derailment.
In the course of completing the functional analysis required by DFSS, the team realized that it was not necessary to require that the tank component fulfill all three functions. “We could separate the train loads from the tank and separate the protective function from the tank,” Ward said. That separation opened new design avenues, which the team is exploring. “DFSS also is helping us not just look at this railcar from a design point of view but to consider the whole life cycle,” Trager said. “So we’re thinking about how easy it is to fabricate and service, whether it will pass industry and government regulations, its maintenance requirements, and how it can be managed in an emergency response.”
The team’s goal is to have prototype railcars built in early 2008 and then start comprehensive crash testing. “We should start replacing cars in 2010 and complete the conversion of our high-hazard fleet by 2014,” Ward said. “This is an extremely aggressive timeline given that we’re looking at new technologies that have never been used in the railroad industry before. But DFSS and TRIZ are making it possible.”
A Leap of Faith
Deciding to integrate toolsets like DFSS and TRIZ into a Six Sigma deployment requires a leap of faith because the gains are often quantifiable only in hindsight. But Costa thinks Dow’s experience can serve as a benchmark for others. “Most corporate initiatives burn brightly for 18 months or two years, then die out,” he said. Dow has been at this for nine years and has seen a return of “thousands of percent over investment.”
Costa credits Dow’s success to constantly adapting their Six Sigma methods to the needs of the company. “We found that if we articulate the right strategies and get the right toolsets to execute them in a systematic way, using a very proven method, then our probability of success is much higher than companies who aren’t as disciplined,” he said.
Making its quality program Six Sigma centric – with the governance structure of Belts, Champions, Sponsors and the use of reviews – has enabled Dow to keep the toolsets and operating disciplines integrated with corporate strategy, Costa said. He urges companies to make sure they do integrate their efforts rather than pile on disjointed efforts. “Look at your company and understand where you will get the earliest payback, what elements of the income statement are most critical to manage first.” Then, he said, step back and blueprint the methodologies that will help you achieve the necessary level of performance.
“Addressing some challenges requires incremental improvements in performance,” Trzcinski said, “whereas addressing others requires a true step change – a breakthrough in performance. Integrating innovation tools with Six Sigma methodology gives Dow people a broader set of skills and tools. It enables them to apply the right tools at the right time to achieve the desired strategic result.”
Sue Reynard is a freelance writer and a frequent contributor to iSixSigma Magazine.