Engineers employ their accumulated knowledge and experience to guide iterative cycles of modification, analysis and assessment to an ‘optimal’ final design. Unfortunately, time, money and workforce are always limited and much too often design processes stop prematurely when a feasible solution has been derived. All essential design requirements may be met but the results are rarely optimal.
Over the past two decades computers have become indispensable tools in engineering; however, they sit idle for most of the time. The development of computer hardware has outpaced software development as well as our usage of this resource. Computations that 20 years ago took five minutes to complete on a US$30,000 workstation are now finished in two seconds on a US$3,000 personal computer (PC). In fact, today’s multiprocessor PCs provide more number-crunching capability than most of us ever use.
Optimal results and better utilisation of available computing capabilities can be achieved by better integrating available software into the design process. As an example, this article concentrates on a hydrodynamic hull-shape optimisation procedure that combines advanced computeraided design (CAD) tools and numerical seakeeping analysis with nonlinear programming algorithms. The resulting software package supports the development of offshore structures with optimum seakeeping characteristics. Although hull shape and its influence on wave forces and motions constitutes only one of many aspects in the design of floating platforms, good seakeeping behaviour is essential for the successful operation and safety of any offshore structure. Similar approaches can be and are used in structural design. Innovative tools always transform the work process itself. In our case, relief from tedious interactive hull modelling shifts the engineer’s focus to the definition of desirable design properties and requirements. In order to better understand the benefits and cost of hydrodynamic hull shape optimisation, we begin our discussion by illustrating the drawbacks of the traditional hull development process.
Traditional Hull Design Process
Developing a hull shape starts with creating a digital model within an interactive CAD system (see Figure 1). Modern 3D modelling systems represent hulls with B-spline and non-uniform rational B-spline (NURBS) curves and surfaces. Their shape is manipulated by control meshes, which consist of a set of points in space – so-called vertices. Almost any shape is possible. However, designing surfaces with pre-defined properties (area, volume, centroid, etc.) is, at best, tedious. As indicated by the arrows in Figure 1, the engineer continues shifting vertices until the required properties have been derived. A subsequent seakeeping analysis yields response amplitude operators (RAOs) of pressure, wave forces and motions of the platform. In the performance assessment and comparison of design alternatives, environmental conditions at the target area of operation are accounted for by short- and long-term wave statistics. Whenever a design stage yields unsatisfactory results the process is re-started by implementing the required changes in the digital hull model.
As depicted in Figure 1, engineers are involved in every step and valuable work hours are spent manipulating vertices of the hull model and with pre- and post-processing of seakeeping analysis. The process is restricted to office hours and leaves vast computational resources unused. As the process is so time-consuming, often only a few design alternatives are investigated. The best of the alternatives that satisfies all essential design requirements is chosen as the final design. Rarely will it be an optimal solution.
To View full article : Click here
|
