Many advances in “Digital Twin” technology used by design tools have come about recently, so Navy has the responsibility to evaluate tools and processes in order to develop the next generation early stage ship design work space so we do not continue to design tomorrows ships with yesterdays tools.
Here we present “Digital Twin” solutions to application of product model technology, high performance computing, and early stage design tools playing an important part in the development of future Navy Ships. The subject of design tools is explored from the perspective of how they improve the early stage ship design process as well as their role in gaining insights and supporting oversight during the detailed ship design and construction phases.
Change is a permanent fixture to the acquisition process, and open architectures and the availability of standards for the definition of product models has the potential to improve the early stage design process. Until recently, investigation into simulation-based shipbuilding systems and their applications have been incomplete, and design demonstration of constructed ships is only partially done for use in sales or assembly simulation of welding robots.
Efforts aimed at the realisation of virtual shipyard in an integrated work space have yet to be conducted since new concept ships include thousands of different parts and, therefore, the performance and cost of new computing power to process huge amount of information can be difficult to achieve.
But there are important reasons more investigations have yet to be conducted. Generally, shipbuilding is order based and a new design is made for each new ship. Furthermore, the whole process from contract and design to building happens concurrently.
The cycle of design/manufacturing/ maintenance is repeated, where through each stage, information becomes more detailed. Many processes are mainly dependent on agent labour quality and require qualitative information making it difficult to extract detailed information in early stages. The only possible way to overcome these limitations and reduce the lead-time of the process from design to manufacturing is through the use of computers.
Since computers are already widely used and the current level of their usage is very high, simulation-based shipbuilding process can lead to greatly reduced manufacturing time than can conventional sequential process from design through process planning to manufacturing. Evaluation of production capability in the early design stage is very important. The importance is manifest in relations between production cost and the possibility of cost reduction. In other words, early evaluations of design and manufacturing activities occupy a low percentage of total cost, but they have big effects on cost reduction and product quality.
However, in general, early evaluation requires a vast amount of information on design and manufacturing, and frequently it is not an easy task. If a simulation-based shipbuilding system is used, all elements and activities both in design and manufacturing that are required for product development will be modeled in a computer-based product model, and the whole design and manufacturing process can be simulated in computer working space.
In the concept design phase, the ship design organisation should explore large sets of potential design alternatives using design space exploration and automated visualisation tools to rapidly populate a design space with performance, cost, and risk metrics. There are several areas of improvement that have been identified to fill in a complete toolset for concept design. As we continue to identify and prioritise working toolsets, actions should be taken to improve those areas.
During this phase the level of detail may be relatively low, but the design is extremely dynamic with goal to identify solutions that are feasible. New tools will quickly at a low level of detail identify windows of feasibility considering many variables including; cost, weight, arrangeable space, powering, and many others.
Ships are large and complex products and have a long development cycle. It is widely recognised throughout the engineering world that decisions made during the conceptual design phase have the largest impacts on cost, performance, and schedule. Many of the critical requirements levied on a warship require complex assessments to verify that they are met such as hull fatigue life, vulnerability/shock performance, signatures, and topside sensing/communication performance.
High level of design definition is required and is typically not available until the detailed design and construction phase. In the current design efforts, results verify if a ship design meets its requirements come after their opportunity to influence the design. Because of the limited amount of tool integration, and a manual ship design definition process, the Navy enterprise usually driven to select one design alternative early in the design process. Much of the rest of the design effort is spent detailing and reworking this single alternative to meet the requirements and cost goals.
Decisions made early on in the ship design process have large impacts on ship functionality that are not quantified until the design is mature. Often these impacts are only vaguely understood at the outset of the design cycle, and by the time that the impacts are fully understood it is too late to make significant changes. An example of this could be the vulnerability of the ship: in order to asses vulnerability, a detailed layout of compartments and distributed systems is needed, but early on in the ship design, when sizing decisions are made, detailed layouts are not available and a ship designer has little more than rules-of-thumb to base these crucial decisions upon.
To establish a simulation-based shipbuilding work space, the most important factor is successfully resolving the planning problem. Planning in the shipbuilding process is the process of devising method to minimise effects of design change and delay, supports the appropriate manufacturing method, and maximises the usage of resources in decision making.
But if planning and production are done without the optimisation of decision making during the design and manufacturing process, unexpected delays, unwanted modifications, and various other problems may occur. In other words, to solve problems after the beginning of the planning and production requires changes in plans such as the reassignment of resources. In short, in order to accurately plan the shipbuilding process, the technology for implementing dynamic simulations is required.
With High Performance Computing as an enabler, the vision is to explore all downstream implications of decisions made during the initial concept development and apply that knowledge as early on in the design process as possible. In the vulnerability example used above, for instance, an automated tool could rapidly produce a full range of feasible ship arrangements from a basic shell of a ship, and then a vulnerability assessment could be performed on each of these many design variations and the resultant range of achievable levels of vulnerability can be fed back to the designer—with all of the highspeed computation happening behind the scenes so designers are instantly aware of the vulnerability implications of the sizing and arrangement of the ship.
Ship Design involves complex interactions between many disciplines, and reconciling the needs of one system against others becomes a delicate balancing act. The convergence of various discipline-specific ship models into a valid single design is made up of discipline specific modules i.e. hull geometry, gross arrangement, hull structural design, resistance and propulsion, power plant sizing, weight estimation, and area/volume.
With spiral design approach disciplines are addressed one at a time before moving to the next one, and multiple iterations are performed through the spiral process in order to converge into a single solution. Each loop is a serial process that must be done in order, and control of each design variable must be carefully executed. The tech modules are highly coupled so that the dynamic process of integration is stable and converges on a solution.
In a Set-Based Design approach, which has been identified as a preferred approach for the development of future Navy design efforts, discipline-specific designs are done in parallel across a broad design space. This process is designed to improve the flexibility of the design by delaying key decisions until the design space is fully understood, but the parallel nature of the approach also makes it an ideal fit for High Performance Computing HPC application.
Set-based design relies on engineering interaction and judgment for creating the set information from each discipline and integrating the results from multiple disciplines in order to find a set of possible solutions. The vision for Navy design tools is to move to a automated high-end toolset that integrates many information dense design definition tools with high fidelity physics-based analysis tools.
This toolset will be able explore many ship design alternatives to populate a feasible design space. This design space will be used to perform real time agent-based cost-benefit trades on ship configurations during the requirements definition process.
There systems could be used to explore the design space to ensure correct design is selected before signing a contract to build a ship. Direction outlines types of tools and tool developments needed. Accomplishing these ambitious goals will be a challenge, but is essential for crafting affordable, executable ship programs for future missions.
Current applications have been built and maintained by the Navy to function as principal tool used in earliest stages of ship design by combining ship design disciplines into one integrated whole-ship model that represents a balanced design, but does not produce the level of design definition required for many of the higher-level assessments required in the later stages of ship design.
When a design progresses beyond concept design to the stage where a more detailed assessment is required, the design integration provided between disciplines by current applications is lost. Existing tools typically require their own custom format of input and most time is spent preparing the input, often manually recreating design essentials already created in another tool. This recreation accounts for most of the time, cost, and error associated with assessments.
The effort to solve this time-delay and configuration control issue between high-end tools is focused on digital representation of the ship designed to be expansible to include all information necessary to perform any assessments and store the results of those assessments, functioning as operational hub, while detailed discipline-specific tools represent the spokes in a ship design cycle.
In addition to the issue of configuration between disciplines, many aspects of ship design do not have sufficient tools and models in existence, and increasingly rely on subjective engineering judgment. For instance, we are often looking at new and innovative ways to estimate ship manning requirements or sustainment costs at an early stage. But developing and improving the individual high-end tools is is not as simple as implementing direction into a computer system.
Tools need to be verified and validated; problems must be easy to set up and run; structural process generation must be easy and quick; tools must be built to run effectively and efficiently on massively parallel computers; and, results must be timely. Many of the tools we use are highly specialised, and not employed for practical use beyond ship design. Results must be visualised and packaged in a way so it is easy to understand by both the design engineers and programme executives so smart and timely decision making process is well-executed.
To establish a simulation-based shipbuilding work space the most important factor is successfully resolving the planning problem. Planning in the shipbuilding process is the process of devising processes that minimises the effects of design change and delay,
supports the appropriate manufacturing method, and maximises the usage of resources in decision making.
However, if planning and production are done without the optimisation of decision making during the design and manufacturing process, unexpected delays, unwanted modifications, and various other problems may occur. In other words, to solve problems after the beginning of the planning and production requires changes in plans such as the reassignment of resources.
In short, in order to accurately plan the shipbuilding process, technology for implementing design simulations is required. Simulation-based shipbuilding based on 3D systems can be described as a concept that enables the product and process simulation of a wide range of the whole of the product life cycle including the design, production, and maintenance under virtual computer work space.
To realise simulation-based shipbuilding, it is necessary to perform both function modeling of the design and operational resources related to ship functions based on a 3D virtual ship prototypes and process modeling of the manufacturing processes, plans, and manufacturing resources. The results of this modeling and information of virtual ship prototypes must be shared between shipyards, programme offices, engineering firms, etc.
Ultimately, our goal is to shrink the time required to generate a sufficient amount of information to make informed design decisions early in the ship design process before the requirements are set and cost of the ship is locked in. By considering an integrated computational ship model as a “virtual prototype,” several design iterations are possible in a far shorter amount of time than a single design-build-test cycle of a traditional prototype.
Decisions made early on in the ship design process have large impacts on ship functionality that isn’t quantified until the design is mature. Often these impacts are only partially understood at the outset of the design cycle, and by the time engineers figure out the impacts it is too late to make significant changes. Take for example assessement of ship vulnerability requires detailed layout of compartments and distributed systems, but early on in the ship design, when sizing decisions are made, detailed layouts are not available and a ship designer has little more than rules-of-thumb to base these crucial decisions upon.
With High Performance Computing acting as an enabler, the vision is to explore all downstream implications of decisions made during the initial concept development and apply that knowledge as early on in the design process as possible. In the vulnerability example used above, for instance, an automated tool could rapidly produce a full range of feasible ship arrangements from a basic shell of a ship, and then a vulnerability assessment could be performed on each of these many design variations and the resultant range of achievable levels of vulnerability can be fed back to the designer—with all of the highspeed computation happening behind the scenes. Thus, the designer is instantly aware of the vulnerability implications of the sizing and arrangement of the ship.
The function model has a close relation with ship design. To be a virtual prototype that is meaningful in engineering sense, its behaviour should be exactly expressed. In other words, behaviour based on physical characteristics should be expressed. In this sense, the function model is the model used for expressing the behaviour of its object system in the simulation of provided functionalities. It should convey system behaviour pattern and output responses to control input.
The function model must be able to take input from other system models, and provide output for the simulation of other systems. Also, it should be connected with the geometric and characteristics portion of the product model. The process model is an indispensable model for ship manufacturing- requires the expression of the product and process. In other words, not only the information of the product to be produced, but also the assembly sequence, the manufacturing technology or assembly process according to the production plan should be incorporated as well.
In addition, as an example of the process model, it includes certain behaviour to quantify events happening during the building process such as welding distortion. To express many production patterns of ships, multiple process models are required. The creation of building models plays an important role in the creation of adequate virtual prototypes or production of prototypes.
The virtual prototype is in the form of a product model, function model, and process model combined, and should be manipulated in virtual workspace. For the virtual prototype to achieve the real time response required for design and development, a high power computer that can recalculate the vast geometry info that composes a part of the product model is required.
Currently, computer systems that can process such a large amount of digital parameters in a design work space is just now becoming available.. However, designers, developers, operators, and manufacturers use only part of product model warehouse according to their needs, therefore, a collection of small prototypes adequate to each function rather than one complete prototype with all the necessary functions is more realistic.
In addition, for the efficient operation of simulation systems using virtual prototypes, an important issue to resolve is real-time processing in the area of high performance visualisation. While many computer systems focus on 3D solid objects, high performance visualisation focuses on surface rendering. The basic expression unit of an image is the polygon, and frequently millions of polygons are necessary for realistic frames. In this sense, partial simplification is necessary for product models to perform as geometry models for virtual prototypes.
Statistics are now available to engineers through automation and high-speed computing, to allow for better capture of uncertainty into the design process, but it allows several single aspects of a ship design to be explored comprehensively on their own before comparing them to ensure convergence and feasibility of the ship design as a whole.
In addition to linking ship structures, hydrodynamics, and susceptibility models for instance, the front end can link to force models and the back end can link with cost and affordability models to provide a full picture to decision makers so that timely decisions can be made with confidence.
Current next generation shipbuilding systems automate conventional processes and increase productivity and quality, but they are limited in introducing fundamental changes to conventional processes or accepting rapidly advancing technologies efficiently.
Recently, much attention has been paid to simulation-based manufacturing, which can model and simulate the whole process of manufacturing products including the design.
This simulation-based manufacturing environment is called virtual manufacturing, digital manufacturing, or virtual factory, and it is implemented in concepts such as virtual shipyards, and digital shipbuilding.
In order to develop practical systems based on contract provided here, a product model based 3D system is required. Furthermore, technologies for the process of block stand-up, assembly, processing, cutting, and testing should also be developed with operational verification.
In the future, by applying virtual simulation technology to the design, modeling, simulation, manufacturing, testing, and information systems under simulation-based manufacturing environments, a basis for concurrent engineering systems can be built and will be a principal factor affecting the competitiveness of the shipbuilding industry.
Here, we have discussed the emerging tools, modeling, and product info integration work space being developed to support early stage of Navy ship design. It is true Navy ship design was performed well before any of the advanced computational capabilities we seek today were available, but with current constraints to include rising cost of ships and the increasing complexity of technology, we cannot afford to not have the most powerful tools available.
Leveraging the power of “Digital Twin” Simulations to design, build, deploy and sustain the ships of the future while improving on existing process tools is likely to become even tougher to figure out.
1. Virtual Shipbuilding Solution Test Space
2. Function Simulation Process Simulation
3. 3D product Model System
4. Virtual Dock Assembly Simulation
5. General Equipment Arrangement
6. Planning & Scheduling Resource
7. Process Resource Transportation
8. Job Resource Worker, Operator
9. Operation Resource Workplace
10. Design Resource Performance