Any naval architect will tell you that computational fluid dynamics (CFD) software is the greatest thing since the towing tank. And rightly so. The simulation software eliminates the need for elaborate model tests and is conducive to a faster and smoother design process. Experts use it to model anything from underwater appendages to hull forms.
But what if a new tool were to offer naval architects even quicker solutions to their design problems? C-Job’s naval architects have dipped a toe in the water with a new CFD-driven design optimization method to help solve fluid dynamics problems, and according to them it’s a game-changer.
Down the rabbit hole of CFD simulations
The company’s use of the new in-house tool, a part of its accelerated concept design framework, started with a Data Scientist named Roy de Winter. While working as a PhD candidate at C-Job, he developed a sophisticated optimization algorithm called SAMO-COBRA to help its naval architects make quicker and better design decisions. Bob van Veen, Naval Architect and R&D Engineer at C-Job, improved the algorithm’s user-friendliness, embedded it within the company’s CFD software, and a new optimization tool was born.
For there is nothing simple about fluid dynamics or about CFD simulation software — starting with an inability to work ‘backwards’. A naval architect can’t ask the software to provide a solution that meets multiple objectives, constraints, and decision parameters. “Instead, the naval architect must manually run designs through the software until the best option is found. Each run involves waiting for the results, interpreting the data, and modifying the design for the next cycle. It’s a lot of work,” explains van Veen.
Breaking the “test, tweak, repeat” cycle
It’s no wonder, then, that naval architects can often complete only two or three iterations within the scope of their project. “By combining our CFD software with the new algorithm we can carry out as many as fifty iterations without exceeding our budgeted engineering hours,” enthuses van Veen.
Time is saved by the optimization algorithm taking over the task of adjusting a design, based on performance feedback from the CFD simulations. In a nutshell, it removes the need for manual design changes and finds the best solution itself. This makes the tool ideal for complex — and expensive —optimization problems with long simulation durations.
A match made in computational heaven
Having more time enables naval architects to explore a wide design brief early in the process, minimizing risks and costs in later design stages. The new tool also helps designers make decisions that consider both capital and operational expenses, ultimately reducing the client’s total cost of ownership.
But the benefits of this match made in computational heaven extend beyond efficiency. Being able to iterate through numerous designs at high speed gives fluids experts more flexibility and enables them to dive into new details and unknown territories. Meeting the clients’ unique requirements and discovering areas for improvement are now within the realm of possibility.
The future of ship design lies in data science
C-Job is one of a very few companies pioneering optimization algorithms in this way. Van Veen said that others are experimenting with algorithms, but there are no other naval architecture companies working with such a sophisticated optimization routine.
C-Job’s naval architects are experienced in data science, and in making the best use of the art of number-crunching in their design projects. Another example of their data-driven approach is their use of profile analyses. This involves extracting and analyzing relevant, real-world data—ranging from cloud cover and wave heights to speed and heading—right from the start of the design. This detailed understanding of how a ship operates in real-world conditions results in an accurate picture of the client’s exact needs and how best to meet them.
“The data-driven approach to ship design is where the future of naval architecture lies. Tools such as these give naval architects the opportunity to choose the best designs instead of having to generate them,” adds van Veen.
And that gives them more time to do what really matters: engineering better ships that perform more efficiently.
