The costs of building and running prototype engines are high, but they’re extremely daunting when each cylinder of a marine engine provides more than 1000 kW (1340 hp). That’s prompted Wärtsilä Corp. to invest heavily in hardware and software so modeling, simulation and virtual testing can be used to shorten development time and ensure that all design requirements are met.
The Finnish company’s marine engines are among the world’s largest. Far from the company’s largest engine, the Wärtsilä 46 was built to run on heavy fuel oil, marine diesel oil, or light diesel so it can run in ports and at sea. It is still so hefty that building a prototype and testing it is quite costly. The four-stroke engine has a cylinder bore of 460 mm (18.1 in) Cylinder output is 1200 kW (1610 hp)/cylinder. The smallest model stands 4930 mm (194 in) and weighs 97 t (107 ton), while the largest is 5863 mm (231 in) tall and a hefty 233 t (257 ton).
“Prototyping is really expensive; fuel costs alone are really substantial,” said Juho Könnö, digital design platform manager at Wärtsilä. “We want to promote virtual testing more and more. We are also doing more virtual engineering, going into multimodal models.”
Könnö, speaking at Dassault Systèmes’ Science in the Age of Experience conference in Chicago, noted that Wärtsilä has come a long ways since it started modeling in 1973 with finite element method (FEM) calculations. Continued investments in hardware have yielded powerful clusters. In recent years, software costs have outstripped hardware expenditures.
This software is helping the company move to requirements-based validation, in which the requirements are continuously used to ensure that all parameters are being met. That’s important because all requirements impact multiple components.
“At the component level, each component has dozens of requirements. Sometimes, you can have over 100 elements that all need to be examined to be sure they meet customer requirements,” Könnö said. “There are no standalone requirements. The need for good fuel economy breaks down into 10s of requirements.”
Making simulation and validation accessible to a broad base of engineers has helped shorten development time. In the past, requests were e-mailed to simulation managers who often didn’t get to the request for a day or two. That process repeated when managers were asked for time to run a simulation, so it wasn’t uncommon for engineers to wait about a week for a 3-hour simulation.
Now, simulation software is easier to manage and more engineers have direct access to the computing cluster, so many of the delays have been eliminated. The evolution to 3D imagery also makes it simpler for engineers to drill down deeply to understand the design and eliminate points that cause problems. Könnö is pressing to further democratize software and minimize any pain associated with simulations.
“My vision is that when they’re finished, users will have a sweet taste in their mouths, and that they can’t wait to do their next simulation,” he said.