By Kelley Stirling
NEW NAVAL PLATFORMS—SUCH AS THIS SHIP-TO-SHORE CONNECTOR BEING TESTED AT THE DAVID TAYLOR MODEL BASIN—ARE BEING FABRICATED WITH A PROCESS CALLED “SET-BASED DESIGN.”
Set-based design has become the preferred approach for early-stage ship design at Naval Surface Warfare Center Carderock Division, according to engineers working on the Navy’s future ships and submarines.
Dr. Jason Strickland and Jeff Hough invited engineers and academics to Carderock to learn more about set- based design and how Navy engineers and designers can work together using this approach during a summit at Carderock Division’s headquarters in West Bethesda, Maryland, in August.
“We wanted to start to have a really honest conversation about what is set-based design, what’s not set-based design, how it’s different than what we’ve done, and how we start to apply it with a common language,” said Strickland, a senior naval architect from Carderock’s future ship and submarine concepts branch. The summit was, in part, sponsored by the Office of Naval Research (ONR), and included engineers and designers from all the different naval warfare centers, Naval Sea Systems Command, Naval Air Systems Command, and Space and Naval Warfare Systems Command.
Two University of Michigan professors, Dr. Matthew Collette and Dr. David Singer, spoke to the group about the evolution of design and what set-based design is. Several Navy engineers presented examples of how set- based design already has been used for the Navy and how it could work for other naval design efforts.
Point-based design is the traditional approach to designing ships and other vehicles, meaning there are decisions being made at iterative points during the process. Singer said traditional design approaches may mean redesign for design failures or late changes in requirements.
“The core principle for set-based design is delaying the design decisions until trade-offs are fully understood,” Singer said. “In set-based design, we want to establish feasibility before commitment.”
This process allows for a diversity of solutions, especially if requirements change at some point in the process. If the requirements change during the design process, the designers can easily evaluate the new requirements against the sets of designs and systems they have already established for feasibility. In general, Singer said that both design approaches are successful and that point-based design makes sense when the requirement is only a modification on an existing ship, vehicle, or system. But for more complex designs, set-based design could make more sense. Point-based design also functions best with a highly experienced workforce that has developed the engineering intuitions needed to make critical design decisions early on.
“Traditional design is done in stages,” said Collette, who spoke about the evolution of design. “Complexity is addressed by not dealing with all aspects of the design at each stage.” He added that this is often called “over-the- wall” design, where each stage gets thrown over the wall to the next stage and the stages are not communicating throughout the process.
Collette explained that, in some ways, set-based design is similar to concurrent engineering, where there is communication back and forth throughout the process. Concurrent engineering occurs when different stages of the design process are being worked at the same time. Set-based design allows the customer the opportunity to be part of the decision process as well.
“The reason we developed set-based ship design is to try to maximize your chance of success,” Collette said. “It allows people to make more decisions later in the process.”
Dr. John Burrow, deputy assistant secretary of the Navy for research, development, test and evaluation, and keynote speaker at the summit, is passionate about the benefits of set-based design.
“I lived it, I breathed it, I practiced it, I demonstrated it and I briefed it to the most senior people,” Burrow said of his role in designing an amphibious combat vehicle for the Marines. He added that set-based design allowed the designers and the customers and everyone in between to have healthy discussions about the requirements.
There may be hundreds of thousands of possible sets of designs when initial requirements are put into a design tool to help process the information. As more design and analysis data is added to the design space, the sets can be reduced in size, and infeasible solutions removed. In the end, there still may be more than one design option, whereas in point-based design, the experienced engineers may already know at what point to start their design and ultimately will end up with one design to present to the customer early in the process.
Set-based design has become a more viable option with the host of technology that is available or becoming available. Singer said in the 1950s when ships were being designed, there was a lack of information and so the design space only included a small number of possibilities. By the 1980s, the design space was on information overload. But the ability to process the information was limited.
The Department of Defense’s High Performance Computing Modernization program office sponsored, and Carderock developed, the Computational Research and Engineering Acquisition Tools and Environments (CREATE)-Ships tools to help process design options. Within CREATE-Ships, there are several design tools, such as Rapid Ship Design Environment (RSDE), Integrated Hydrodynamics Design Environment (IHDE) and Navy Enhanced Sierra Mechanics.
Dr. Alex Gray, a naval architect and set-based design expert with Carderock’s future ship and submarine design tools branch, said as set-based design becomes more widely accepted, these tools are the key to making it a viable design methodology.
“When we have hundreds to thousands of points, how do you narrow that set?” Gray asked. “And how do you cross-analyze?”
The tools available allow Navy ship designers to add additional analysis beyond just basic ship stability. Right now, the RSDE tool helps ship designers generate concept points that are architecturally feasible for a naval vessel. Is the ship floating upright, is there enough displacement, is it structurally sound? The information gained can then be
input into other tools that are not in the RSDE environment, such as the IHDE tool, which can then provide a resistance and seakeeping analysis. Carderock developed the Leading Edge Architecture for Prototyping Systems (LEAPS) data environment as the means to supply common engineering data between different engineering tools.
As the tools are being developed within the Navy, the idea is to integrate them all so that these computations on different levels become automatic, generating concepts within minutes.
Amy Markowich is the director of Naval Air Systems Command’s battlespace simulation department and is also the Department of the Navy’s modeling and simulation executive. She wants to help the Navy’s warfare centers develop and share the tools necessary to make set-based design usable across the Navy.
“We want a continuum, a computational prototyping environment or a common modeling environment, where we can assess designs early, make choices, start building the prototype, but then have all our capabilities tied together to be able to evaluate a prototype. That means tying our labs, our models, and simulations together in a common environment,” Markowich said during the Carderock summit. “How can we collectively take this to the next level?”
In most cases of ship or submarine design, engineers and architects have used a point-based approach, because this is what they learned in school.
“We as engineers were not trained to do set-based design in school,” said Jeff Hough, the Navy Warfare Center Distinguished Engineer for Ship and Platform Design and Integration. “Ultimately, if the Navy is using set-based design, we have to get the universities to start teaching set-based design as an approach.”
Singer said students receive lectures on set-based design in graduate school, and it is mentioned in undergraduate classes, but the majority of design that is taught is point-based design. In some cases, doctoral students have focused their dissertations on set-based design. Strickland is one of four Ph.D.s in Carderock’s future ship and submarine concepts branch who studied set-based design under Singer at the University of Michigan.
Strickland said Singer is recognized as the person who developed set-based design as it applies to ship design, noting that Toyota gets the credit for creating the design approach originally. Singer trained the ship-to-shore connector (SSC) design team on set-based design and worked out the process for using set-based design on the amphibious landing craft. The SSC is being built as a replacement for the Navy’s air cushion landing craft class of vehicles.
Part of the problem is that the schools do not have the tools necessary to use set-based design, such as LEAPS, RSDE, and IHDE.
“They don’t have these tools at the universities right now,” Gray said. “But I think it’s possible, if they had a group of engineering students that were interested in doing a naval vessel, they could teach a set-based process for that.”
From an “on-the-job-training” aspect, Hough said one of the benefits of set-based design that the Navy did not really anticipate was that it provides a means to do design with inexperienced engineers and architects. In point- based design, there generally needs to be someone with experience in ship design to offer up the first design that fits the requirements. On the other hand, as inexperienced designers work in the set-based process, they are looking at a multitude of options within the design space that allows them to gain knowledge on something they had never done before.
“To be able to develop and train, in stride, inexperienced people so they can support design is a huge benefit,” Hough said, adding that most engineers and architects will design one or two ships in their career, so everyone coming in to the ship-design workforce is inexperienced, including some that have been there for a decade.
For Burrow, set-based design is a must for the Navy. “It’s not about the textbook. It’s not about how they teach you to do it,” he said. “It’s about how you make sense of large, complex systems and make informed decisions and make them right. It’s not about giving me a requirement and swearing to it. It’s about working with me to develop the requirement.”
There have been several designs that have been developed within the Navy using set-based design, the SSC being one of the first where Singer provided the framework for the Navy to use set-based design for that vehicle, as well as future vehicles. The SSC is set to be delivered to the fleet in 2017.
To help foster the idea that set-based design is the best option, Carderock held a demonstration project on early ship design in 2012. Two design teams were established: one using set-based design and the other using point-based design. Both teams were given the same requirements, as well as the same design tools, with the exception of the set-based design team also having the RSDE design tool. On the set-based design team, the engineers were mostly inexperienced and the point-based design team had one team member who was experienced in ship design.
The point-based design team came to a solution quickly, whereas the set-based team needed a little more time to build the sets before coming to a set of options they could provide the customer.
Most engineers recognize that there are more costs up front using the set-based design approach, but that the costs balance out, and in some cases will be less overall at the completion of a new ship. The costs usually come in the form of more engineers working on the project and taking more time to develop the sets of possibilities based on the requirements from the customer. But the savings come from less time being spent on finding the right design when a requirement changes; for example, if the customer now wants the ship’s minimum speed to be 30 knots instead of 28 knots, and needs more affordable designs with lower risk. The set-based design team will be able to go to their set of designs and find that ship without having to redesign.
This is what happened in the design demonstration project. The teams were given a requirement change late in the design process.
“The point-based team had to pretty much start over. They had experience, there was a learning curve, so they were able to do it quicker the second time, but they came up with another point design,” Hough said. “The set-based design team was able to plug it into the design space and come up with a design much quicker.”
For the Future
Gray said that the Navy is evaluating using the set- based design process to look at technology insertion studies. As an example, if set-based design is being used to develop a combat system, the design space will include parameters such as area, volume, power, and cooling requirements. When designing for a particular combat system, minimum and maximum ranges of the requirements are assigned.
“Taking the set-based approach, we are going to look at that minimum and maximum and everywhere in between,” Gray said. This will generate sets that don’t necessarily have real-life combat systems in them, but the design sets may still be useful in developing an understanding of how the ship behaves when toggling between different power and cooling requirements. “When someone comes to you from the outside, maybe someone from [Naval Surface Warfare Center] Dahlgren Division, and says, ‘we think this future weapon is going to have these properties,’ we can just pull from the set we’ve already developed and say, ‘I have a ship design that more or less has those exact physical properties.’”
Technology insertion studies like this also may provide a resource for determining if it makes sense to invest money, research, and development for a particular technology in the long run, Gray said.
A Community of Practice
“A community of practice for designers and practitioners is critical,” Singer said of implementing set-based design across the warfare centers. He added that academia does not have the benefit of communities of practice.
Strickland and Hough hope the August summit, which was directed to naval engineers and architects who may actually be responsible for executing set-based design in their engineering organizations, was the first stepping stone for creating a community of practice that will help to establish a common language on how to use set-based design.
“This is a self-organizing community of practice,” Strickland said, speaking of the group attending the summit. “I would like to see it become more formal and a more regular thing, maybe a larger quarterly meeting or monthly for project-specific things. But I think it needs to grow in that fashion.”
“We are trying to inform people so that when you hear the term set-based design, you actually know what it is and you can apply it when it’s appropriate, and you can become champions of it as I have,” Burrow said.
Strickland says the next summit will be later this year and will include university affiliated research centers and federally funded research and development centers, both of which do a lot of the defense work. He said he expects the next summit will include “a lot less education, more application discussion.”
To further the community of practice, Hough and Strickland established a Navy set-based design community of practice page on the Defense Department’s milSuite site at: https://www.milsuite.mil/book/groups/navy-set-based-design-community-of-practices/activity. This will eventually migrate to NAVSEA’s Fusion and Wiki sites.
About the author:
Kelley Stirling is a public affairs writer with Naval Surface Warfare Center Carderock Division.