Ship design for high-performance ships like mega-yachts should be based on high-performance tools including sophisticated analyses. Increasingly we rely on »virtual experience« from dedicated and well chosen simulations to guide design decisions. Scope and depth of these simulations have developed very dynamically over the past decade.
Structural analyses
Finite-Element analyses (FEA) for global strength within the elastic material domain have been standard for[ds_preview] a long time. These simulations were the starting point for more sophisticated analyses, e.g. fatigue strength assessment, ultimate strength assessment, etc. Finite-element analyses require load specifications which involve external hydrostatic and hydrodynamic loads. GL.ShipLoad, Cabos et al. (2006), supports efficient load generation for global FEA of ship structures, allowing overall rapid and reliable analyses. Hydrostatic and hydrodynamic computations are integrated into the program. GL.ShipLoad supports the generation of loads from first principles (realistic inertia and wave loads for user supplied wave parameters). The result is a small number of balanced load cases that are sufficient for the dimensioning of the hull structure.
FEA is today also the standard choice for ship vibration analyses. The computations require longitudinal mass and stiffness distribution as input. The mass distribution considers the ship, the cargo and the hydrodynamic »added« mass. The added mass reflects the effect of the surrounding water and depends on the frequency. One can either use estimates based on experience or employ sophisticated hydrodynamic simulations. For local vibrations analyses, added mass needs to be considered if the structures border on tanks or the outer hull plating. Because of the high natural frequencies of local structures, FEA models must be quite detailed including also the bending stiffness of structural elements.
For very high frequencies (structure-borne noise), the standard FEA approach to vibration analyses is impossible due to excessive computational requirements. For a frequency of 1,000 Hz, a FEA vibration model would lead to several million degrees of freedom. However, for acoustics, information is required only averaged over a frequency band. Therfore, an alternative, far more efficient approach based on statistical energy analysis can be employed, in our case the Noise Finite Element Method (GL NoiseFEM) of Germanischer Lloyd, Cabos and Jokat (1998), Cabos et al. (2001). GL NoiseFEM predicts the propagation of noise by analyzing the exchange of energy between weakly coupled subsystems. Validation with full-scale measurements shows that the accuracy of GL NoiseFEM is sufficient for typical structure-borne sound predictions for the frequency range between 80 Hz and 4000 Hz, Wilken et al. (2004).
Computational Fluid Dynamics (CFD)
Seakeeping is closely related to passenger comfort and as such always an important topic for mega yachts. For many seakeeping issues, linear analyses (assuming small wave height or small wave steepness) are appropriate and frequently applied due to their efficiency. The advantage of this approach is that it is very fast and allows thus the investigation of many parameters (frequency, wave direction, ship speed, etc.). Non-linear CFD simulations are usually necessary for the treatment of extreme motions. These simulations require massive computer resources and allow only the simulation of relative short periods (seconds to minutes). Combining intelligently linear frequency-domain methods with nonlinear time-domain simulations allows exploiting the respective strengths of each approach, El Moctar (2005). The approach starts with a linear analysis to identify the most critical parameter combination for a ship response. Then a non-linear CFD analyses determines motions, loads, free surface (green water on deck) and speed loss in waves.
CFD is the most appropriate tool to support practical rudder design, GL (2005). The propeller is typically modelled in a simplified way using external forces distributed over the cells which cover the location where the propeller would be in reality. The sum of all axial body forces is the thrust. The body forces are assumed to vary in radial direction of the propeller only. This procedure is much faster than geometrical modelling of the propeller (by two orders of magnitude) at a negligible penalty in accuracy (about 1 %). The procedure has been extensively validated for rudder flows both with and with-out propeller modelling. The same approach for propeller and rudder interaction can be applied for podded drives, Junglewitz and El Moctar (2004), or waterjets. Here, the detailed insight in the flow helps also in trouble-shooting with cavitation occurring on appendages and waterjet inlets.
Fluid-structure interaction is more recent field where simulations have brought significant progress. Menzel et al. (2008) present a case study for vortex-induced vibrations in a mega yacht. Such hydrodynamically induced vibrations cannot be prevented with current state-of-the-art design approaches and the traditional trial-and-error approach is tedious and costly. A combination of measurements and simulations allowed solving this complicated vibration problem in relatively short time. The presented approach is time and cost efficient, therefore clearly superior to the traditional trial-and-error approach.
HVAC (heat, ventilation, air condition) simulations involve the simultaneous solution of fluid mechanics equations and thermodynamic balances, often involving concentrations of different gases. Super and mega yacht applications include for example smoke propagation and flow conditions (buoyancy and turbulence) on helicopter decks affecting safe helicopter operation. Although wind tunnel tests still are popular and widely used, CFD offers the advantage of overcoming scale effects which can be significant if thermodynamic processes are involved, El Moctar and Bertram (2002). It also allows formal optimisation of ship or funnel designs for smoke dispersion on decks, Harries and Vesting (2010). Ventilation systems are still largely designed based on simple estimates, resulting often in over-dimensioned and energy inefficient systems, sometimes in insufficient performance. CFD can guide more intelligent layout, based on reliable detailed insight.
At present, zone models and CFD tools are considered for fire simulations in ships. Zone models are suitable for examining more complex, time-dependent scenarios involving multiple compartments and levels, but numerical stability can be a problem for scenarios involving multi-level ship domains, HVAC systems and for post-flashover conditions. CFD models can yield detailed information about temperatures, heat fluxes, and species concentrations. However, the time penalty of this approach currently makes CFD unfeasible for long periods of real time or for large computational domains. Nevertheless, applications have graduated from preliminary validation studies to more complex applications for typical ship rooms (accommodation, atrium, engine room), Bertram et al. (2004). Alternative designs (in terms of equivalent fire safety) allow today larger design freedom for the naval architects.
Final remarks
Simulations are a powerful tool for ship designers. They allow a reliable sneak preview into assorted aspects of vessel performance at an early design stage; they also open the door for unconventional designs. The much larger design freedom is very much appreciated by creative designers and discerning owners alike.
Acknowledgements
Many colleagues at Germanischer Lloyd have supported this paper with their special expertise, supplying text and/or figures, namely (in alphabetical order) Andreas Brehm, Christian Cabos, Bettar El Moctar, Stefan Harries, Jürgen Jokat, Axel Köhlmoos, Holger Mumm, Tobias Zorn.
Author: Prof. Dr.-Ing. Volker Bertram
FutureShip – A GL Company
Volker Bertram
