Predicting the performance of a sail design is important for optimising the performance of a yacht, and Velocity Prediction Programs (VPPs) are commonly used for this purpose. The aerodynamic force data for a VPP is often calculated using Computational Fluid Dynamics (CFD) models, but these can be computationally expensive. A full VPP analysis for sail design is therefore usually restricted to high-budget design projects or research activities and is not practical for many industry projects. This work presents a method to reduce the computational cost of creating lift and drag force coefficient curves for input into a VPP using both multi-fidelity kriging surrogate modelling and data from existing sail designs. This method is shown to reduce the number of CFD simulations required for a desired accuracy when compared to a single-fidelity model. A maximum reduction in the required computational effort of 57% was achieved for model-scale symmetric spinnaker sails. For the same number of simulations, the accuracy of the model predictions was improved by up to 72% for scale-symmetric spinnaker sails, and 90% for asymmetric spinnakers.

Dynamic Velocity Prediction Programs are taking an increasingly prominent role in high performance yacht design, as they allow to deal with seakeeping abilities and stability issues. Their validation is however often neglected for lack of time and data. This paper presents an experimental campaign carried out in the towing tank of the Ecole Centrale de Nantes, France, to validate the hull modeling in use in a previously presented Dynamic Velocity Prediction Program. Even though with foils, hulls are less frequently immersed, a reliable hull modeling is necessary to properly simulate the critical transient phases such as touchdowns and takeoffs. The model is a multihull float with a waterline length of 2.5 m. Measurements were made in head waves in both captive and semi-captive conditions (free to heave and pitch), with the model towed at constant yaw and speed. To get as close as possible to real sailing conditions, experiments were made at both zero and non-zero leeway angles, sweeping a wide range of speed values, with Froude numbers up to 1.2. Both linear and nonlinear wave conditions were studied in order to test the limits of the modeling approach, with wave steepness reaching up to 7% in captive conditions and 3.5% in semi-captive ones. The paper presents the design and methodology of the experiments, as well as comparisons of measured loads and motions with simulations. Loads are shown to be consistent, with a good representation of the sustained non-linearities. Pitch and heave motions depict an encouraging correlation which confirms that the modeling approach is valid.

. In this paper, we propose a new control method for the next generation of sailboat autopilots. These new systems will need to manage more actuators to control the hydrofoils, which is going to significantly increase the energy requirements. So, this method is aware of the autopilot power consumption. It uses a model predictive controller to manage the actuators. This controller uses a dynamic model of the actuator, running in real time, to anticipate the future behavior of the system. Once the predictions are made, it determines the future control sequence to apply in order to follow the reference trajectory. To do so, it minimizes a cost function which takes into account two criteria: the precision of the system and the energy. With the proposed control method, skippers can focus on one or the other criterion depending on their goals and the boat's energy balance. We apply this method to one of the autopilot's subsystems, namely the rudder control. The electric actuator intervening in this control loop and the load representing the force opposed to its motion are modelled to design the control law. The first results of that method are compared with a standard autopilot. We increase by 40% the precision level and we are able to reduce the consumption by at least 20%. This work provides the first necessary components of a future autopilot that will control the whole appendages to a three-dimensional piloting. Moreover, this type of management is a first step towards possible fossil fuel-free sailboats.

. In this paper, we propose a new control method for the next generation of sailboat autopilots. These new systems will need to manage more actuators to control the hydrofoils, which is going to significantly increase the energy requirements. So, this method is aware of the autopilot power consumption. It uses a model predictive controller to manage the actuators. This controller uses a dynamic model of the actuator, running in real time, to anticipate the future behavior of the system. Once the predictions are made, it determines the future control sequence to apply in order to follow the reference trajectory. To do so, it minimizes a cost function which takes into account two criteria: the precision of the system and the energy. With the proposed control method, skippers can focus on one or the other criterion depending on their goals and the boat's energy balance. We apply this method to one of the autopilot's subsystems, namely the rudder control. The electric actuator intervening in this control loop and the load representing the force opposed to its motion are modelled to design the control law. The first results of that method are compared with a standard autopilot. We increase by 40% the precision level and we are able to reduce the consumption by at least 20%. This work provides the first necessary components of a future autopilot that will control the whole appendages to a three-dimensional piloting. Moreover, this type of management is a first step towards possible fossil fuel-free sailboats.

. In this paper, we propose a new control method for the next generation of sailboat autopilots. These new systems will need to manage more actuators to control the hydrofoils, which is going to significantly increase the energy requirements. So, this method is aware of the autopilot power consumption. It uses a model predictive controller to manage the actuators. This controller uses a dynamic model of the actuator, running in real time, to anticipate the future behavior of the system. Once the predictions are made, it determines the future control sequence to apply in order to follow the reference trajectory. To do so, it minimizes a cost function which takes into account two criteria: the precision of the system and the energy. With the proposed control method, skippers can focus on one or the other criterion depending on their goals and the boat's energy balance. We apply this method to one of the autopilot's subsystems, namely the rudder control. The electric actuator intervening in this control loop and the load representing the force opposed to its motion are modelled to design the control law. The first results of that method are compared with a standard autopilot. We increase by 40% the precision level and we are able to reduce the consumption by at least 20%. This work provides the first necessary components of a future autopilot that will control the whole appendages to a three-dimensional piloting. Moreover, this type of management is a first step towards possible fossil fuel-free sailboats.

The stability and the dynamic behavior are integral parts of designing hydrofoil supported sailing vessels, such as the America's Cup (AC) 50 class. The foil design and the control systems have an important influence on the performance and stability of the vessel. Both foil and control system design also drive the maneuverability of the vessel and determine maneuvering procedures. The AC50 class requirements lead to complex foil control systems and the maneuvering procedures become sophisticated and multifaceted. Sailing and maintaining AC50 class yachts is a complex, expensive and time-consuming task. A dynamic velocity prediction program (DVPP) for the AC50 is therefore developed to assess the dynamic stability of different foil configurations and to simulate and optimize maneuvers. The goal 143 is to evaluate certain design ideas and maneuvering procedures with this simulator so that sailing time on the water can be saved. The paper describes the principal concepts of developing an AC50 model in the DVPP FSEquilibrium. The force components acting on the yacht are defined based on physical principles, computational fluid dynamics (CFD) simulations and experimental investigations. The control systems for adjusting the aero- and hydrodynamic surfaces are modeled. Controllers are utilized to simulate the human behavior of performing sailing tasks. Maneuvers are then defined as sequences of crew actions and crew behaviors. In the paper examples of utilizing the DVPP in preparation for the 35th America's Cup in Bermuda are described. The DVPP is for example used to investigate the effect of different boat set-ups on stability and handling during maneuvers. With the sailing team, maneuver procedures are developed and tested. Procedures such as dagger board and rudder elevator movement and crew position are investigated and evaluated to minimize the distance lost during tacking and gybing. The DVPP is also employed for trajectory optimization during maneuvers.

. The paper presents a Velocity Prediction Program for hydrofoiling catamarans with solid wing sails. Starting from a description of the mechanical model, suitable models are identified for the forces that act on the boat components. The study is deliberately limited to means for restricted methods where computational resources and budgets are limited. Enhanced lifting line approaches are described for the wingsail and appendages. Windage is calculated from force coefficients and dynamic pressure while hull resistance is determined by means of a potential flow solver. The description of the implementation is followed by the presentation of the results. An approach for processing of measurement data is shown in preparation to compare measured and predicted performance. Good agreement was found for upwind conditions whereas varying differences are found for most of the remaining wind angles. Finally, some of the challenges encountered during the study are discussed.

In November, 2015, the Sailing Yacht Research Foundation (SYRF) published the tank test data from their "Wide Light Yacht Project" for the hydrodynamics of a modern, high performance, semi-planing yacht. This comprehensive data set, comprising canoe body with and without appendages in upright, heeled and yawed conditions, provides an important validation base for Computational Fluid Dynamics (CFD) codes. The SYRF report includes a number of comparative results from commercial CFD codes. In this paper, we present results computed using an advanced Boundary Element Method (BEM), FloSim, these calculations being compared against the test matrix of "Wide Light" measured data and also STAR-CCM+ RANS calculated results. Computer model preparation is discussed and FloSim and RANS case execution times and computer requirements are compared. FloSim is an unsteady, time-stepping BEM with a non-linear vortex wake rollup procedure and a coupled integral boundary layer method for viscous effects. The free surface wave development is treated with a non-linear procedure that includes a breaking-wave model to deal with excess wave energy issues. The program was developed specifically to bridge the gap between basic potential flow panel methods and RANS codes, with the objective of providing accurate, practical solutions on a laptop computer within a reasonable turnaround time and cost. In essence, the results presented in this paper demonstrate the development objectives have been achieved. The comparisons of FloSim's essentially "low-order" results against test data and STARCCM+ RANS calculations, provide a measure of tradeoff between calculation accuracy versus cost and turnaround time for a case.

: This study examines the hydrodynamic performance of a high performance skiff hull using three different physical testing techniques which may be used in early stage design for assessment of the upright resistance of sailing vessels. The hull chosen as a benchmark form is a high-speed hard-chine sailing dinghy, typical of modern trends in skiff design, and is broadly similar to some high performance yacht hulls. The 4.55 m hull was tested at full scale in a moderate size towing tank, at 1:2.5 scale in the same tank, and at full-scale by towing on open water. The work presented here builds on the study of Day & Cameron (2017), with the model tests repeated and re-analyzed in the present study and additional results presented. The challenges of full-scale open-water testing are discussed and several potential improvements in practice are identified for future work. Results show that the open water testing broadly matches well with model-scale tank testing, with the mean discrepancy in the measured resistance between the two around 4% over the speed range tested after correction for the presence of the rudder. Agreement is initially less good for the full-scale hull in the tank for higher speeds, both for resistance and trim. ITTC guidelines suggest that blockage may be an issue for the full-scale boat in this size of tank; comparison of the results suggests that blockage, and/or finite depth effects for the full-scale hull in the tank present a substantial problem at the higher speeds. A correction approach for the wave resistance of the full scale results using a calculation based on a linear thin ship theory is effective in this case, and results show that the full scale and model scale tests agree satisfactorily for the majority of the speed range after this correction. In addition to upright resistance in calm water, results are presented for the impact of small waves, the addition of the rudder, and variations in displacement and trim on resistance for a skiff hull. Finally, the results are compared with predictions from the well-known Delft series regression equations, Savitsky's methods, and a thin ship calculation. The thin ship approach gives good agreement for the case in which the hull is trimmed bow-down and the transom is not immersed, while the Savitsky pre-planing approach gives good agreement for the level trim case. The Delft series and Savitsky planing hull approaches do not give good agreement with physical measurements.

. The design of sails has always been done experimentally, and only recently simulations are starting to be used in the design process. This paper presents a solver that couples CFD and FEM in order to compute the deformed sail shape (flying shape) and the thrust it can provide. The complexity of the problem is both in the flow, which is fully turbulent and detached, and in the structure, which is deformable and free to move in all directions. Moreover, the coupling of the solvers has to be performed in a way that minimizes loss of information and accuracy. The CFD simulations have been run and validated with the commercial software FINE/Open. The results were reasonably satisfying when compared to the experimental data obtained by Viola et al. (2014). Consequently, the FEM solver has been successfully validated for some cases of which the analytical solution is known, due to lack of reference data for this specific case. Finally, the interpolation techniques have been implemented in Matlab and the fluid structure interaction solver has been run. From the results it can be argued that the design and flying shape of the sail are quite different and provide different thrusts. This results in the need for this type of analysis in the sail design and manufacturing process: considering the sail as rigid is not enough to have a correct representation of the spinnaker behavior. It is necessary to couple the CFD solver with a finite element solver able to describe the cloth deformation. This is true for all types of sails, but for downwind sails it is particularly important, due to their high deformability with respect to upwind sails.

. The design of sails has always been done experimentally, and only recently simulations are starting to be used in the design process. This paper presents a solver that couples CFD and FEM in order to compute the deformed sail shape (flying shape) and the thrust it can provide. The complexity of the problem is both in the flow, which is fully turbulent and detached, and in the structure, which is deformable and free to move in all directions. Moreover, the coupling of the solvers has to be performed in a way that minimizes loss of information and accuracy. The CFD simulations have been run and validated with the commercial software FINE/Open. The results were reasonably satisfying when compared to the experimental data obtained by Viola et al. (2014). Consequently, the FEM solver has been successfully validated for some cases of which the analytical solution is known, due to lack of reference data for this specific case. Finally, the interpolation techniques have been implemented in Matlab and the fluid structure interaction solver has been run. From the results it can be argued that the design and flying shape of the sail are quite different and provide different thrusts. This results in the need for this type of analysis in the sail design and manufacturing process: considering the sail as rigid is not enough to have a correct representation of the spinnaker behavior. It is necessary to couple the CFD solver with a finite element solver able to describe the cloth deformation. This is true for all types of sails, but for downwind sails it is particularly important, due to their high deformability with respect to upwind sails.

: This study examines the hydrodynamic performance of a high performance skiff hull using three different physical testing techniques which may be used in early stage design for assessment of the upright resistance of sailing vessels. The hull chosen as a benchmark form is a high-speed hard-chine sailing dinghy, typical of modern trends in skiff design, and is broadly similar to some high performance yacht hulls. The 4.55 m hull was tested at full scale in a moderate size towing tank, at 1:2.5 scale in the same tank, and at full-scale by towing on open water. The work presented here builds on the study of Day & Cameron (2017), with the model tests repeated and re-analyzed in the present study and additional results presented. The challenges of full-scale open-water testing are discussed and several potential improvements in practice are identified for future work. Results show that the open water testing broadly matches well with model-scale tank testing, with the mean discrepancy in the measured resistance between the two around 4% over the speed range tested after correction for the presence of the rudder. Agreement is initially less good for the full-scale hull in the tank for higher speeds, both for resistance and trim. ITTC guidelines suggest that blockage may be an issue for the full-scale boat in this size of tank; comparison of the results suggests that blockage, and/or finite depth effects for the full-scale hull in the tank present a substantial problem at the higher speeds. A correction approach for the wave resistance of the full scale results using a calculation based on a linear thin ship theory is effective in this case, and results show that the full scale and model scale tests agree satisfactorily for the majority of the speed range after this correction. In addition to upright resistance in calm water, results are presented for the impact of small waves, the addition of the rudder, and variations in displacement and trim on resistance for a skiff hull. Finally, the results are compared with predictions from the well-known Delft series regression equations, Savitsky's methods, and a thin ship calculation. The thin ship approach gives good agreement for the case in which the hull is trimmed bow-down and the transom is not immersed, while the Savitsky pre-planing approach gives good agreement for the level trim case. The Delft series and Savitsky planing hull approaches do not give good agreement with physical measurements.

. The paper presents a Velocity Prediction Program for hydrofoiling catamarans with solid wing sails. Starting from a description of the mechanical model, suitable models are identified for the forces that act on the boat components. The study is deliberately limited to means for restricted methods where computational resources and budgets are limited. Enhanced lifting line approaches are described for the wingsail and appendages. Windage is calculated from force coefficients and dynamic pressure while hull resistance is determined by means of a potential flow solver. The description of the implementation is followed by the presentation of the results. An approach for processing of measurement data is shown in preparation to compare measured and predicted performance. Good agreement was found for upwind conditions whereas varying differences are found for most of the remaining wind angles. Finally, some of the challenges encountered during the study are discussed.

In November, 2015, the Sailing Yacht Research Foundation (SYRF) published the tank test data from their "Wide Light Yacht Project" for the hydrodynamics of a modern, high performance, semi-planing yacht. This comprehensive data set, comprising canoe body with and without appendages in upright, heeled and yawed conditions, provides an important validation base for Computational Fluid Dynamics (CFD) codes. The SYRF report includes a number of comparative results from commercial CFD codes. In this paper, we present results computed using an advanced Boundary Element Method (BEM), FloSim, these calculations being compared against the test matrix of "Wide Light" measured data and also STAR-CCM+ RANS calculated results. Computer model preparation is discussed and FloSim and RANS case execution times and computer requirements are compared. FloSim is an unsteady, time-stepping BEM with a non-linear vortex wake rollup procedure and a coupled integral boundary layer method for viscous effects. The free surface wave development is treated with a non-linear procedure that includes a breaking-wave model to deal with excess wave energy issues. The program was developed specifically to bridge the gap between basic potential flow panel methods and RANS codes, with the objective of providing accurate, practical solutions on a laptop computer within a reasonable turnaround time and cost. In essence, the results presented in this paper demonstrate the development objectives have been achieved. The comparisons of FloSim's essentially "low-order" results against test data and STARCCM+ RANS calculations, provide a measure of tradeoff between calculation accuracy versus cost and turnaround time for a case.

The stability and the dynamic behavior are integral parts of designing hydrofoil supported sailing vessels, such as the America's Cup (AC) 50 class. The foil design and the control systems have an important influence on the performance and stability of the vessel. Both foil and control system design also drive the maneuverability of the vessel and determine maneuvering procedures. The AC50 class requirements lead to complex foil control systems and the maneuvering procedures become sophisticated and multifaceted. Sailing and maintaining AC50 class yachts is a complex, expensive and time-consuming task. A dynamic velocity prediction program (DVPP) for the AC50 is therefore developed to assess the dynamic stability of different foil configurations and to simulate and optimize maneuvers. The goal 143 is to evaluate certain design ideas and maneuvering procedures with this simulator so that sailing time on the water can be saved. The paper describes the principal concepts of developing an AC50 model in the DVPP FSEquilibrium. The force components acting on the yacht are defined based on physical principles, computational fluid dynamics (CFD) simulations and experimental investigations. The control systems for adjusting the aero- and hydrodynamic surfaces are modeled. Controllers are utilized to simulate the human behavior of performing sailing tasks. Maneuvers are then defined as sequences of crew actions and crew behaviors. In the paper examples of utilizing the DVPP in preparation for the 35th America's Cup in Bermuda are described. The DVPP is for example used to investigate the effect of different boat set-ups on stability and handling during maneuvers. With the sailing team, maneuver procedures are developed and tested. Procedures such as dagger board and rudder elevator movement and crew position are investigated and evaluated to minimize the distance lost during tacking and gybing. The DVPP is also employed for trajectory optimization during maneuvers.

. The design of sails has always been done experimentally, and only recently simulations are starting to be used in the design process. This paper presents a solver that couples CFD and FEM in order to compute the deformed sail shape (flying shape) and the thrust it can provide. The complexity of the problem is both in the flow, which is fully turbulent and detached, and in the structure, which is deformable and free to move in all directions. Moreover, the coupling of the solvers has to be performed in a way that minimizes loss of information and accuracy. The CFD simulations have been run and validated with the commercial software FINE/Open. The results were reasonably satisfying when compared to the experimental data obtained by Viola et al. (2014). Consequently, the FEM solver has been successfully validated for some cases of which the analytical solution is known, due to lack of reference data for this specific case. Finally, the interpolation techniques have been implemented in Matlab and the fluid structure interaction solver has been run. From the results it can be argued that the design and flying shape of the sail are quite different and provide different thrusts. This results in the need for this type of analysis in the sail design and manufacturing process: considering the sail as rigid is not enough to have a correct representation of the spinnaker behavior. It is necessary to couple the CFD solver with a finite element solver able to describe the cloth deformation. This is true for all types of sails, but for downwind sails it is particularly important, due to their high deformability with respect to upwind sails.

The stability and the dynamic behavior are integral parts of designing hydrofoil supported sailing vessels, such as the America's Cup (AC) 50 class. The foil design and the control systems have an important influence on the performance and stability of the vessel. Both foil and control system design also drive the maneuverability of the vessel and determine maneuvering procedures. The AC50 class requirements lead to complex foil control systems and the maneuvering procedures become sophisticated and multifaceted. Sailing and maintaining AC50 class yachts is a complex, expensive and time-consuming task. A dynamic velocity prediction program (DVPP) for the AC50 is therefore developed to assess the dynamic stability of different foil configurations and to simulate and optimize maneuvers. The goal 143 is to evaluate certain design ideas and maneuvering procedures with this simulator so that sailing time on the water can be saved. The paper describes the principal concepts of developing an AC50 model in the DVPP FSEquilibrium. The force components acting on the yacht are defined based on physical principles, computational fluid dynamics (CFD) simulations and experimental investigations. The control systems for adjusting the aero- and hydrodynamic surfaces are modeled. Controllers are utilized to simulate the human behavior of performing sailing tasks. Maneuvers are then defined as sequences of crew actions and crew behaviors. In the paper examples of utilizing the DVPP in preparation for the 35th America's Cup in Bermuda are described. The DVPP is for example used to investigate the effect of different boat set-ups on stability and handling during maneuvers. With the sailing team, maneuver procedures are developed and tested. Procedures such as dagger board and rudder elevator movement and crew position are investigated and evaluated to minimize the distance lost during tacking and gybing. The DVPP is also employed for trajectory optimization during maneuvers.

: This study examines the hydrodynamic performance of a high performance skiff hull using three different physical testing techniques which may be used in early stage design for assessment of the upright resistance of sailing vessels. The hull chosen as a benchmark form is a high-speed hard-chine sailing dinghy, typical of modern trends in skiff design, and is broadly similar to some high performance yacht hulls. The 4.55 m hull was tested at full scale in a moderate size towing tank, at 1:2.5 scale in the same tank, and at full-scale by towing on open water. The work presented here builds on the study of Day & Cameron (2017), with the model tests repeated and re-analyzed in the present study and additional results presented. The challenges of full-scale open-water testing are discussed and several potential improvements in practice are identified for future work. Results show that the open water testing broadly matches well with model-scale tank testing, with the mean discrepancy in the measured resistance between the two around 4% over the speed range tested after correction for the presence of the rudder. Agreement is initially less good for the full-scale hull in the tank for higher speeds, both for resistance and trim. ITTC guidelines suggest that blockage may be an issue for the full-scale boat in this size of tank; comparison of the results suggests that blockage, and/or finite depth effects for the full-scale hull in the tank present a substantial problem at the higher speeds. A correction approach for the wave resistance of the full scale results using a calculation based on a linear thin ship theory is effective in this case, and results show that the full scale and model scale tests agree satisfactorily for the majority of the speed range after this correction. In addition to upright resistance in calm water, results are presented for the impact of small waves, the addition of the rudder, and variations in displacement and trim on resistance for a skiff hull. Finally, the results are compared with predictions from the well-known Delft series regression equations, Savitsky's methods, and a thin ship calculation. The thin ship approach gives good agreement for the case in which the hull is trimmed bow-down and the transom is not immersed, while the Savitsky pre-planing approach gives good agreement for the level trim case. The Delft series and Savitsky planing hull approaches do not give good agreement with physical measurements.

. The paper presents a Velocity Prediction Program for hydrofoiling catamarans with solid wing sails. Starting from a description of the mechanical model, suitable models are identified for the forces that act on the boat components. The study is deliberately limited to means for restricted methods where computational resources and budgets are limited. Enhanced lifting line approaches are described for the wingsail and appendages. Windage is calculated from force coefficients and dynamic pressure while hull resistance is determined by means of a potential flow solver. The description of the implementation is followed by the presentation of the results. An approach for processing of measurement data is shown in preparation to compare measured and predicted performance. Good agreement was found for upwind conditions whereas varying differences are found for most of the remaining wind angles. Finally, some of the challenges encountered during the study are discussed.

In November, 2015, the Sailing Yacht Research Foundation (SYRF) published the tank test data from their "Wide Light Yacht Project" for the hydrodynamics of a modern, high performance, semi-planing yacht. This comprehensive data set, comprising canoe body with and without appendages in upright, heeled and yawed conditions, provides an important validation base for Computational Fluid Dynamics (CFD) codes. The SYRF report includes a number of comparative results from commercial CFD codes. In this paper, we present results computed using an advanced Boundary Element Method (BEM), FloSim, these calculations being compared against the test matrix of "Wide Light" measured data and also STAR-CCM+ RANS calculated results. Computer model preparation is discussed and FloSim and RANS case execution times and computer requirements are compared. FloSim is an unsteady, time-stepping BEM with a non-linear vortex wake rollup procedure and a coupled integral boundary layer method for viscous effects. The free surface wave development is treated with a non-linear procedure that includes a breaking-wave model to deal with excess wave energy issues. The program was developed specifically to bridge the gap between basic potential flow panel methods and RANS codes, with the objective of providing accurate, practical solutions on a laptop computer within a reasonable turnaround time and cost. In essence, the results presented in this paper demonstrate the development objectives have been achieved. The comparisons of FloSim's essentially "low-order" results against test data and STARCCM+ RANS calculations, provide a measure of tradeoff between calculation accuracy versus cost and turnaround time for a case.