7
STRUCTURES AND STRUCTURAL DYNAMICS
Computational structural mechanics and dynamics (CSMD) covers a broad spectrum of activities, including computational methods for predicting the response, failure, and life of structures and structural components and automated methods of structural synthesis and optimization. Also, CSMD is an important component in the multidisciplinary analysis and design of many engineering systems.
RECENT ADVANCES
Major advances in CSMD continue to take place on a broad front. The new advances are manifested by the development of sophisticated computational models that simulate mechanical, thermal, and electromagnetic responses of structures; efficient discretization techniques (viz., improved finite element, boundary element, and hybrid/analytical numerical methods); stochastic-based modeling; and computational strategies and numerical algorithms for nonlinear static, dynamic, and fracture problems. Large software systems now exist for detailed modeling and analysis of structures. These systems are currently being applied to some rather complicated problems, and this is affecting the design cycle of engineering systems and their components.
Life prediction methods are required to perform structural integrity assessments of engineering systems. For metallic structures, empirical methods for life prediction are being replaced by fracture-mechanics-based computational methods. For composite structures, more work is needed to understand the mechanisms of failure initiation and propagation before computational methods can be widely used by industry. Because of the difficulty in understanding and modeling the failure phenomena, computational methods for strength and life predictions are lagging behind those of response predictions.
FUTURE DEVELOPMENTS
CSMD is likely to play a significant role in the future development of structures technology, as well as in the multi-disciplinary design of future engineering systems. For this to happen, major advances and computational tools are needed in a number of key CSMD areas. To this end, the research community must address a number of primary and secondary pacing items, many of which are discussed in this appendix. In identifying the pacing items, three factors are taken into account: characteristics of future engineering systems and their implications on CSMD, future computing environments, and recent and future developments in other fields of computational technology (notably fluid dynamics and computational mathematics) that can be adapted to CSMD.
Primary and Secondary Pacing Items
The primary pacing items in CSMD include detailed modeling of complex structures, prediction and analysis of failure of structural components made of new materials, effective computational strategies for large systems, computational methods for articulated dynamic systems, and quality assessment and control of numerical simulations. (The last pacing item is covered in Appendix 7.) The secondary pacing items include integration of analysis programs into computer-aided design/computer-aided engineering (CAD/CAE) systems and predata-postdata processing and the effective use of visualization technology. (The secondary pacing items are not discussed in this appendix.)
Detailed Modeling of Complex Structures
One of the most important steps for the accurate prediction of the response of a complex aerospace structure is the proper selection and sequencing of mathematical and discrete models with varying degrees of complexity. Hence, there is a need for development of automatic model generation facilities as well as smart interfaces to the analysis and design systems. The smart interfaces will be artificial intelligence (AI) based expert systems that run on workstations and can help the engineer in the initial selection of the model, its adaptive refinement, selection of the solution procedure, constraint representation, and interpretation of the results.
Work is currently being done by AI researchers on the development of intelligent computational tools that combine numerical techniques with traditional AI methods, particularly symbolic methods and computer vision. These tools are designed for automatic preparation, execution, and monitoring of numerical simulations and automatic interpretation of their results. These new tools have high application potential in many facets of CSMD, including preparation of a discrete model from high-level specification in symbolic form of structural characteristics and qualitative description of structural response characteristics using computer vision techniques to recognize them (see Appendix 4).
Predictions of Failure of Structural Components
Practical numerical techniques are needed for predicting the failure initiation and propagation in structural components made of new high-performance materials within terms of measurable and controllable parameters. Examples of these materials are high-temperature materials for hypersonic vehicles; piezoelectric composites; and electronic, optical, and smart material systems for space and other applications. For some of the materials, accurate constitutive descriptions, failure criteria, and damage theories are needed, along with more realistic characterization of interface phenomena (such as contact and friction). The constitutive descriptions may require investigations at the microstructure level or even the atomic level, as well as carefully designed and conducted experiments. (Numerical modeling of new materials is covered in Appendix 6.) Numerical simulation of failure is still a challenge and is feasible only under restricting assumptions. Considerable work is needed in this area.
Solution of Large-Scale Structural Problems
A number of large-scale problems exist for which solutions are not feasible, even on present-day large computers. Examples of these problems are dynamics of large flexible structures incorporating the effects of joint nonlinearities and hysteretic damping, interaction problems of large structures with harsh operational environments such as space structures for extraterrestrial bases, and large-scale multidisciplinary design problems. Solution of these problems requires the development of effective multilevel strategies and hierarchical modeling techniques. Promising multilevel strategies include
hybrid modeling/analysis techniques and partitioning methods. In hybrid modeling/analysis techniques, different analytical, numerical, and experimental techniques are combined to predict the response of the structure. Partitioning methods are based on the intuitively obvious and well-established practice of breaking a larger problem into smaller subproblems and generating the solution with information provided by the individual subproblems. Hierarchical modeling is a conceptual change from traditional finite element modeling. It is a strategy to handle structural response characteristics covering a wide variety of length scales (e.g., from global stress/displacement distribution in a composite component to stress/strain distribution in the fibers and the matrix). It can be performed by an initial selection of multiple mathematical models in different regions of the engineering system to take advantage of the simplifications available (through the known local nature of the response). The mathematical models used can range from simple one-dimensional continuum models to three-dimensional micromechanical models followed by adaptive refinement of the models for those regions in which the sensitivity of the response to the modeling details neglected exceeds a prescribed limit.
Computational Methods for Articulated Dynamical Systems
Broad classes of machines that function in space and on the earth's surface involve coupled effects of deformation, angular velocity, and acceleration relative to an inertial reference frame. Deformation may be thought of as relative to an accelerating frame that is associated with each moving component of the machine. Such applications include deployable space structures, orientation control of space systems, space and surface-bound robot and manipulator, ground vehicles such as cars and trucks, aircraft landing gear, and manufacturing equipment. Such machines are designed to control motion and transmit force in the presence of large relative motion between components.
Despite the breadth of these applications and their importance in many fields of engineering and applied science, dynamic simulation methods to predict motion and structural deformation of the components of such systems have lagged significantly behind the much better developed field of finite element structural analysis. While considerable progress has recently been made in computational methods for articulated structural dynamics (often called multibody dynamics or flexible multibody dynamics), much remains to be done in the
development of algorithms and computer software for automated formulation and solution of the differential algebraic equations of motion that describe the dynamics of articulated structures. Such simulation tools must be capable of accurately predicting coupling among bodies that make up the system, coupling between gross motion (angular velocity and acceleration) of individual bodies and their associated deformation fields, and prediction of constraint forces that act between components. Only then can effective performance and failure analysis be carried out prior to fabrication and hardware testing.
Another major objective of articulated structural dynamics is to provide accurate and effective simulation of dynamics and control of hydraulic subsystems that are used to impart desired motion to the system. Structural-control interaction is becoming increasingly important for lightweight extremely flexible articulated structures whose motion is controlled by feedback systems that can cause unwanted interaction and even instability. Finally, with the advent of space telerobotic systems, interaction among the human operator, the control system, and the flexible articulated structure must be accounted for. This will require real-time simulation of the articulated structure and its associated controls, in order to carry out experiments with the operator in the loop.