Computational fluid dynamics-based multiobjective optimization of a surface combatant using a global optimization method

The main objective of this article is to describe the development of two advanced multiobjective optimization methods based on derivative-free techniques and complex computational fluid dynamics (CFD) analysis. Alternatives for the geometry and mesh manipulation techniques are also described. Emphasis is on advanced strategies for the use of computer resource-intensive CFD solvers in the optimization process: indeed, two up-to-date free surface-fitting Reynolds-averaged Navier-Stokes equation solvers are used as analysis tools for the evaluation of the objective function and functional constraints. The two optimization methods are realized and demonstrated on a real design problem: the optimization of the entire hull form of a surface combatant, the David Taylor Model Basin—Model 5415. Realistic functional and geometrical constraints for preventing unfeasible results and to get a final meaningful design are enforced and discussed. Finally, a recently proposed verification and validation methodology is applied to assess uncertainties and errors in simulation-based optimization, based on the differences between the numerically predicted improvement of the objective function and the actual improvement measured in a dedicated experimental campaign. The optimized model demonstrates improved characteristics beyond the numerical and experimental uncertainty, confirming the validity of the simulation-based design frameworks.     

Multi-fidelity optimization of a high-speed foil-assisted semi-planing catamaran for low wake

The wakes of high-speed passenger-only ferries that operated through Rich Passage, on the Seattle-Bremerton ferry route, caused beach erosion and damage to habitat. A task was initiated to design a low-wake high-speed vessel using multi-fidelity CFD based design optimization by using low-fidelity potential flow solvers for initial global design optimization and by using URANS solvers for high-fidelity tuning of the optimized design. This simulation based design process involved a close collaboration between ship designers, and hydrodynamics and CFD specialists, whose collective expertise guided the evolution of the design based on both hydrodynamic and structural aspects. The initial hull shape optimization using potential flow code was carried out by blending three different initial concepts provided by the designers. Subsequently, URANS was used to evaluate the potential flow optimized hull and to further optimize the hull configuration parameters, namely, the centre-of-gravity, demihull spacing, foil location, foil angle and slenderness ratio at different displacement conditions. The URANS based configuration optimization also took into account the far field wakes’ energy spectrum with an objective of reducing the energetic, low frequency far field wakes which are associated with beach flattening on the mixed sand and gravel beaches. Calculation of the far field wake using URANS would require an unfeasibly large domain size; therefore, a Havelock code with a source distribution matching the URANS calculated near field wave elevation was used to propagate the wakes into the far field. The end result of the optimization was a design with significantly reduced far field wake, which is currently being built for experimental testing.     

Single- and multiobjective design optimization of a fast multihull ship: numerical and experimental results

Numerical optimization of the initial design of a fast catamaran (high-speed sealift research model B, HSSL-B) has been carried out through a simulation-based design (SBD) framework, based on an advanced free-surface unsteady Reynolds-averaged Navier–Stokes (URANS) solver and a potential flow solver, and global optimization (GO) algorithms. The potential flow computational fluid dynamics (CFD) SBD was used to guide the more expensive URANS CFD SBD. The fluid-dynamic analysis of the flow past the catamaran proved that the use of the URANS solver was fundamental in dealing with the multihull interference problem. In the case investigated, the separation distance was small and the viscous flow quite distorted by the proximity of the hulls, so that only viscous solvers could correctly capture the flow details. Sinkage and trim effects, due to the high speed range and again to the small separation distance investigated, are also relevant. The initial HSSL-B geometry and three optimization problems, including single- and multiobjective optimization problems, proposed by designers from Bath Iron Works, were successfully optimized/solved, and finally an experimental campaign was carried out to validate the optimal design. A new verification and validation methodology for assessing uncertainties and errors in simulation-based optimization was used based on the trends, i.e., the differences between the numerically predicted improvement of the objective function and the actual improvement measured in a dedicated experimental campaign, including consideration of numerical and experimental uncertainties. Finally, the success of the optimization processes was confirmed by the experimental measurements, and trends for total resistance, sinkage, and trim between the original and optimal designs were numerically and experimentally verified and validated.     

CFD validation studies for a high-speed foil-assisted semi-planing catamaran

This paper details the CFD validation studies carried out as a prerequisite for multi-fidelity CFD-based design optimization of high-speed passenger-only ferries aimed at reducing far-field wake energy that causes beach erosion. A potential flow program (WARP) and a URANS program (CFDSHIP) were validated using full-scale measurements of resistance, sinkage, trim, and far-field wake train obtained over a wide range of speeds for two high-speed semi-planing foil-assisted catamarans: Spirit (LOA-22 m) and 1060 (LOA-17 m). This study posed a unique combination of challenges for CFD modeling: the foil appended geometry required complicated surface overset grids, the effect of the waterjet and wind resistance had to be modeled, and a method had to be devised to extrapolate the calculated near-field elevation to get the far-field wake train using Havelock sources. A more concentrated effort was applied to the URANS verification and validation which forms the focus of this paper. The results show that URANS is able to accurately predict the resistance and motions for both vessels when coupled with models that account for the propulsors and air resistance. The overall accuracy of URANS for the performance analysis of the foil-assisted, semi-planing catamarans was adequate to warrant its use as a tool for subsequent design and optimization of a new vessel with significantly reduced wakes.     

Past, present and future state of the biogeochemical Si cycle in the Baltic Sea

The Baltic Sea is one of many aquatic ecosystems that show long-term declines in dissolved silicate (DSi) concentrations due to anthropogenic alteration of the biogeochemical Si cycle. Reductions in DSi in aquatic ecosystems have been coupled to hydrological regulation reducing inputs, but also with eutrophication, although the relative significance of both processes remains unknown for the observed reductions in DSi concentrations. Here we combine present and historical data on water column DSi concentrations, together with estimates of present river DSi loads to the Baltic, the load prior to damming together with estimates of the long-term accumulation of BSi in sediments. In addition, a model has been used to evaluate the past, present and future state of the biogeochemical Si cycle in the Baltic Sea. The present day DSi load to the Baltic Sea is 855 ktons y^− 1. Hydrological regulation and eutrophication of inland waters can account for a reduction of 420 ktons y^− 1 less riverine DSi entering the Baltic Sea today. Using published data on basin-wide accumulation rates we estimate that 1074 ktons y^− 1 of biogenic silica (BSi) is accumulating in the sediments, which is 36% higher than earlier estimates from the literature (791 ktons y^− 1). The difference is largely due to the high reported sedimentation rates in the Bothnian Sea and the Bothnian Bay. Using river DSi loads and estimated BSi accumulation, our model was not able to estimate water column DSi concentrations as burial estimates exceeded DSi inputs. The model was then used to estimate the BSi burial from measured DSi concentrations and DSi load. The model estimate for the total burial of BSi in all three basins was 620 ktons y^− 1, 74% less than estimated from sedimentation rates and sediment BSi concentrations. The model predicted 20% less BSi accumulation in the Baltic Proper and 10% less in the Bothnian Bay than estimated, but with significantly less BSi accumulation in the Bothnian Sea by a factor of 3. The model suggests there is an overestimation of basin-wide sedimentation rates in the Bothnian Bay and the Bothnian Sea. In the Baltic Proper, modelling shows that historical DSi concentrations were 2.6 times higher at the turn of the last century (ca. 1900) than at present. Although the DSi decrease has leveled out and at present there are only restricted areas of the Baltic Sea with limiting DSi concentrations, further declines in DSi concentrations will lead to widespread DSi limitation of diatoms with severe implications for the food web.     

Models of Wheel Contact Dynamics: An Analytical Study on the In-Plane Transient Responses of a Brush Model

This paper deals with the dynamics of the contact between the pneumatic tyre and the ground, in particular with reference to braking manoeuvres of aircraft. The transient response of the braking force is analytically evaluated in terms of the transfer functions that link the variation of the braking force to the variation of physical inputs affecting the interaction between the tread and the ground, in the longitudinal plane. The proposed models have been formulated in order to account for the effects of different reference conditions of steady braking on gains, zeros and poles of such transfer functions.