LEVERAGING SIMULATION FOR COMPETITIVE ADVANTAGE

Transcription

1 LEVERAGING SIMULATION FOR COMPETITIVE ADVANTAGE SUMMARY Dr. Rodney L. Dreisbach Senior Technical Fellow Computational Structures Technology The Boeing Company Simulation is an enabler for the development of new innovative products. As advances in engineering modeling and simulation methodologies are made, it is imperative that they provide for more tightly coupling of technology, processes, and people with the business objectives so that a company s knowledge (most of which is tacit in nature) can be shared, managed, and reused appropriately. It is through nurturing of innovation to exploit a company s intellectual capital across different domain groups and organizational boundaries that a competitive advantage is realized. There have been some incredible advances in product simulation techniques for developing new aerospace vehicles during the past seven (7) decades. However, more advanced process and knowledge management techniques, and more advanced physics-based computational techniques for performing designanalysis-optimization-synthesis activities concurrently, are needed to attain higher levels of prototyping the overall functions of a product in a virtual environment. These needs present new challenges during the 21 st century for more innovative design, analysis, and simulation techniques. Typical of these needs is the ability to perform coupled solutions of multi-physics-based problems using intelligence that is inherent to the simulation processes. The goal is to perform these solutions in a realistic manner, and in real time, so that the lifecycle of an aerospace vehicle can be virtually simulated before physical prototyping is initiated. 1: Introduction Success of a business in the global marketplace requires a breadth of complex factors to be addressed in developing a competitive advantage. These factors, which tend to be strongly interdependent, can be grouped under Business, People, Processes, and Technology (Figure 1). Simulation, in a broad sense, plays a key role within each of the four groups. From an engineering analysis and simulation perspective however, most of the capability and application developments have been focused primarily on the Technology group, with only some recent advances being made in the Processes and People groups. It is imperative that advanced simulation

2 methodologies be more tightly coupled across all four of the groups to effectively address futuristic business environments. Figure 1: Major Factors Associated with a Competitive Business Advantage 2: Knowledge and Innovation The intellectual capital of a company defines the seed bed for innovation that must be nurtured and continuously exploited for: Reduced costs and cycle time in designing, developing, analyzing, manufacturing, and supporting a product Accelerated time-to-market for new products Improved quality and safety of new products To effect these objectives, a company s knowledge strategy and business strategy must be tightly coupled where much of a company s most valuable knowledge is tacit; it is embedded in the minds of the employees. Innovation is a key in exploiting the knowledge of an extended enterprise such that a competitive advantage in the company s business strategy is realized, along with enhanced performance and productivity (Figure 2). Innovation is not something that happens by accident; it is not defined as a eureka moment! Instead, innovation that is focused on a company s business strategy must occur in a disciplined manner that allows for risk assessment to be performed during a project, and where the people are open and able to share knowledge across different domain groups and organizational boundaries. New knowledge is the necessary raw material for innovation and creativity to grow. Furthermore the creation of knowledge and innovation are generally very closely associated with the development of new products and services.

3 Figure 2: Knowledge and Innovation for Competitive Business Advantage Most innovations are the result of a series of incremental improvements that are combined when addressing new product or service opportunities. Consider, for example, the major innovation of the first powered airplane flight by the Wright brothers in This product culminated with not only designing and building a unique airplane but it also included a new 12 horsepower, fourcylinder engine with stringent weight requirements, a new 8.13 ft. propeller that was later found to be 66% efficient (compared with propellers that were considered acceptable for ships at the time if they were 50% efficient), etc. A second example of a major innovation was the development of a privatelyfunded spacecraft capable of reaching sub-orbit, as a result of the Ansari X- Prize competition to address the growth of public interest in space tourism. The prize was won by the US commercial company Scaled Composites, founded and led by Burt Rutan, which successfully achieved sub-orbital flight with its air-launched one-man SpaceShipOne craft during June 21 of This flight demonstrated the practicality of commercial space flight. It also led to other companies developing five and six-manned vehicles for commercial passenger sub-orbital flight, as well as proposals for orbital space hotels and commercial circumlunar missions! Another innovative product on the horizon is the Boeing 787 commercial transport airplane that incorporates new advanced composites for most of its primary structure. The advantages of choosing such composite materials include: increased range and payload, and reduced weight due to higher strength-to-weight ratio; improved environmental performance; enhanced passenger comfort via cabin air that has more moisture and higher pressure; larger, more integrated structure and increased design options, such as onepiece barrel sections and larger windows; reduced manufacturing flows, reduced tooling and easier assembly; and reduced maintenance by the airlines

4 due to higher fatigue and corrosion resistance. Rollout of the first 787 is planned for July 8, 2007 and its first flight is planned for the following month! Reflecting on the foregoing innovations that occurred over the past century, and the idea that many technical experts believe a similar amount of progress will take place during the next 20 years, imagine what new major innovations we will see during the upcoming years! But this will not happen unless major advances are made in the capabilities of our engineering modeling, simulation, and visualization systems. 3: Simulation and Product Development The development of new advanced products is a systems integration problem that requires detailed insight into how the myriad of product design parameters interact with each other. To achieve advanced functionality and reliability, but with decreasing costs and maintenance, multiple design sub-systems of a new product must be optimally integrated. This is most effectively performed through an innovative environment whereby simulation is the key enabler. When developing a new product, the overall simulation capabilities must allow the knowledge and experience of the technical workforce spanning many domains of expertise to be shared, managed, reused, and exploited via an appropriate computer-based infrastructure. These capabilities must span the lifecycle of the product, beginning with the performance of conceptual product studies, through detailed product design and certification, and on to product support activities. In addition to having the computational systems available for performing the necessary simulations, the design and analysis processes by which they are used and reused should be managed and controlled. Too often the associated processes are not sufficiently formalized, not making them reusable, thereby introducing wasted time, inaccuracies in the simulation results, and potentially a negative impact on the product s quality and safety. The CAx/IT infrastructure required to achieve these improvements at an industrial program level must provide a foundational architecture that enables the various tools to interoperate seamlessly and processes to be managed efficiently. The architecture must enable the data to be accessed when needed in real-time across the enterprise and made directly usable for subsequent tasks. For full benefits to be realized, a far more holistic approach to planning and implementing new technologies is required. It requires one with a balanced emphasis on the CAx/IT infrastructure and business process transformations necessary to integrate new technologies. It must enable effective real-time collaboration, as well as accommodate various cultural and organizational changes that may be needed throughout an extended enterprise. Executive management and strategists of engineering and manufacturing companies are continually bombarded with an ever-growing array of tools,

5 technologies, and solutions that suppliers promise will bring immediate gains in productivity and efficiency. They are frequently left challenged with trying to assess the true benefits of the tools relative to their overall business operations. After investing heavily in the necessary software, training, and services, they often discover that their operating costs have increased, and their business expectations were never met. This often occurs because implementation is often performed at a local departmental or functional level without consideration of the global impact to the company-level culture, processes, and business operations. Generally, the core competency of a software developer has been to focus its resources on developing and servicing its products exclusively isolated from the dynamic, heterogeneous environment in which the customer will inevitably deploy the Commercial Off The Shelf (COTS) and proprietary tools. Many suppliers of engineering simulation tools overlook the order-of-magnitude benefits in overall cycle-time and cost reductions that could be achieved when their best-in-class solutions are implemented within an integrated, processcentric environment. Because of the manner in which business must be conducted today, industrial users have been taking a global perspective on the deployment of new technologies within their proprietary processes to maximize their competitive advantage in developing new products. As the functionality and number of CAx computing tools increases within an engineering organization, isolation of the different functional groups and the end-user teams responsible for product development grows. The incompatibility of the data structures and formats of the different tools often results in task-level optimization while operating within a particular tool, at the expense of the overall process when the global flow of product information is considered. Hence, productivity at the macro-process level can be very low. Communications can become strained between the various contributors of a product development team who are often in different geographic regions of an extended enterprise. Even organizing the program into multidisciplinary Integrated Product Teams (IPTs) often does not resolve the underlying issues of data access, data quality control, data security, product data and process management, and data-relationship management across the multiple systems used by a team or across a program. Data quality, poor configuration control and management between data objects, loss of design intent, and poor visibility of product data often compromise attempts to perform concurrent engineering, distributed real-time collaboration, or rapid design iterations with controlled change propagation. What is required is an effective process-centric strategy for exploiting the new technology. Some of these concerns, as related to a design and simulation framework, are being addressed by industry, including those efforts made by the European-sponsored Value Improvement through a Virtual Aeronautical Collaborative Enterprise (VIVACE) Project.

6 4: A Strategic Aerospace Initiative The aeroelastic structural design process is iterative because of complex aerodynamics interacting with complex aerospace vehicle structural arrangements (Figure 3). To obviate exhaustive static and dynamic physical laboratory and flights tests for optimally sizing the various structural components for all flight regimes, extensive analytical and computational methods are used during the design, development, and certification of flight vehicles. Figure 3: The Iterative Aeroelastic Design Process for Aerospace Vehicles A strategic initiative at the Boeing Commercial Airplane Group (BCAG), known as the Product Simulation Integration (PSI) Project for Structures, has been underway to reduce costs and cycle time in the design, analysis, and support of commercial transport airplanes. The Products are the airplanes designed and built, and the services provided to customers for their airplane operations. Simulation includes the analytical and test processes performed to predict in-service behavior of the airplane structure in support of design

7 requirements and objectives. Integration is the close binding of design, analysis, manufacturing, and support processes with the associated product information as it supports reduced costs and cycle time. Fundamental to the success of the PSI project in meeting its goals is establishing standard processes and tools, associating lifecycle information with the product definition data for easy, reliable, and consistent retrieval, and adopting industry standards for sharing these data throughout the product s lifecycle. Standard Processes and Tools reduce variability in the way airplane products are designed, analyzed, and supported, thus lowering training, computing, product support, and sustaining costs. Standard tools and computing systems also provide a common look and feel, along with easy access to multiple computing operating systems and environments where required. By synchronizing and managing the analysis and simulation processes, along with the tools and data, the need for interfaces and the transfer of large numbers of files between applications can be avoided. Additional benefits include reduced product design cycle times that allow for more productivity in addition to increased creativity and innovation. In addition to ensuring the processes and tools are lean relative to their reliability, cycle times, cost, and accuracy, their enrichment through on-going incorporation of new technologies and best practices is essential. These objectives are often accomplished through the capture and reuse of knowledge associated with the product and the design processes by using advanced software techniques commonly referred to as Knowledge-Based Engineering (KBE). Areas of such CAD/CAE applications include the definition of relationships between the myriad of product design and analysis parameters, capturing design/analysis intelligence via rule-based applications for reuse, and developing standard digital templates for reuse of well-defined sub-processes. Associating Lifecycle Information with the Product Definition Data, the records substantiating the design decisions, strength, durability, damage tolerance analyses, and service history of the airplane parts and assemblies are made available for derivative airplane design and analysis while sustaining current configurations. To be successful, these data must be available for the life of the airplane products. The data also need to include analysis and test data that may not necessarily be physically linked, but at a minimum they should be logically linked. Adopting Industry Standards for Sharing Product Information helps to address the fact that evolving computing software and hardware systems have made the task of information retrieval increasingly difficult with time. The best opportunity to preserve the data generated today and to minimize regeneration tomorrow is through the adoption of standards for information

8 exchange. Then, in principal, it is possible to unplug the old analysis or information management tool and plug in a new one without extensive data conversion and disruption to the engineers and customers. 5: Challenges for Innovation In developing future aerospace vehicles during the 21 st century, challenges abound for more innovative technologies and products than ever before. These needs are being driven by increased demands for efficiency, safety and multifunctional operational requirements placed on future aerospace systems. There are many opportunities that currently exist for advancing numerous areas of computational mechanics to virtually simulate, in a realistic manner, the lifecycle of an aerospace vehicle before physical prototyping is initiated. Current design/analysis tools are mostly stand-alone, for example, with most tools operating in a local environment and with little integration. An integrated, comprehensive computing architecture for a multi-physics design/analysis/optimization system that addresses the lifecycle of an aerospace vehicle does not exist. Free exchange of accurate product definition information is difficult since proprietary data representations are typically used. Standards are needed for data modeling and for sharing information and knowledge. Also, product data redundancy is prevalent with many different data models being created via translations to specific technology application codes. It has been reported that in some environments, analysts can spend 75% of their time cleaning up geometry (deleting features in CAD models to allow them to be meshed for analysis) after being imported from CAD to CAE applications. So far, a great deal of focus has been on optimizing the mathematical models of the product and not the product itself (e.g., strength optimization of the structural gages associated with finite element models vs. the geometry, topology, or topography associated with the product). Moreover, increased demands on the operational requirements of products have provoked interactions between multiple technology domains. Although the focus has been on federated data environments, an integrated data environment is preferred where fully-coupled solution techniques spanning multiple physical domains are needed (e.g., the simulation of combustion on structural response). For the most part, in developing a new product, the fidelities of the design constraints spanning multiple technical disciplines are different. Smart techniques are needed for product definition information representation, and for mapping and integration in support of the continuous design evolution process that begins with product conceptual studies. Furthermore, simulation of product lifecycle systems using a product information management system that relies on a common logical, single-source of data is essentially non-

9 existent, and where costing tools and methods in support of developing new products are currently inadequate. The demand for more realistic simulations has been growing rapidly. This inherently requires new modeling, simulation, and solution capabilities that are mathematically closely coupled between the different disparate but interrelated physical elements, and which need to be based on higher levels of time-space accuracy in simulating real physical phenomena. Typical aerospace design scenarios require the interaction across multiple technology domains such as structural strength, stiffness, durability and damage tolerance, thermal, fluid dynamics, systems and controls, crash, etc. Today, solutions to multi-physics problems are typically overly-compromised via expansive assumptions that must be made by the analyst (e.g., decoupling of analysis fields such as combustion simulation from structural response simulation). The current throughput of computational mechanics solutions continues to be marginally acceptable for single-disciplined engineering problems. In addition there is a need for concurrent engineering solutions of multi-physics-based problems that will most likely rely on currently advancing technologies such as intelligent knowledge agents, innovative computing frameworks and management systems, along with clustered hardware utilizing multi-core chip technology and advanced human-computer interfaces. Other advances in computing hardware will soon take us beyond the current supercomputing speed record of 280 trillion (10 12 ) flops by the IBM BlueGene to levels of a quadrillion (10 15 ) calculations per second. Today s video game players are already experiencing real-time realistic modeling and visualization facilities! Similar capabilities of engineering modeling, simulation, and visualization systems for experiencing real-time multi-physical response scenarios are yet merely a vision. Such an environment would truly allow our technical workforce to think creatively, and with increased accuracy, reliability, and productivity. Realistic simulation of multi-physical problems at the speed of human thought should be our vision! 6: Conclusions Incredible advances have been made in multiple areas of computational mechanics technologies and in process implementations within industry for developing new aerospace vehicles during the past seven (7) decades. However, more advanced computational engineering techniques for performing design-analysis-optimization-synthesis activities concurrently, in satisfying the multi-functional operational specifications of an aerospace vehicle, are needed to attain higher levels of product functional prototyping in a virtual environment. Major advances are required in numerous areas of computational mechanics to virtually simulate, in a realistic manner, the lifecycle of an aerospace vehicle before physical prototyping is initiated.