Because of the complexity of modern propulsion systems it is useful to simulate manoeuvers on naval vessels before sea trials. This kind of preparation makes a significant contribution to the adaptation of ship automation
Regarding naval vessels, there are different concepts for ship propulsion. Only considering the recently built frigates, an interesting development can[ds_preview] be seen. In combination with a »cross-connect« gear, the classical use of a diesel engine and a gas turbine allows any desired interconnection of drive units. This leads to more options to respond to the operational profile, particularly with regard to fuel efficiency.
A modern CODLAG (COmbined Diesel-eLectric And Gas Turbine) propulsion system concept offers a number of other advantages and potential benefits. Not only in terms of acoustics, the system brings its assets fully to bear in the electrical driving mode: Since the motors can be coupled directly and exclusively to the propeller shaft, the gear and transmission do not play a role in this driving mode. The deployment of the torque has the characteristics of an electric motor. This has a positive effect on both acceleration and dynamics. The flexible positioning of the electrical power generators is another major advantage regarding mass distribution and noise.
Not least because of the combination with the gas turbine, the use of a CODLAG system results in many practical challenges for the design engineers. One example is the problem of overshooting shaft speed during the stop maneuver of the vessel. Especially when operating with electric drive motors, this process is sensitive and it is important to predict it safely.
By using modern simulation and development tools, the engineer is able to find answers and solutions to special problems and complex issues already at the design and construction phase or during sea trials. The basis provided by driving and maneuver simulation using the system simulation tool Matlab Simulink offers a solution to ship motion equations in the time domain. In order to do this, the hydrodynamic coefficients calculated with the commercial CFD software Ansys CFX are just as essential as the integration of ship components. With respect to the equations of motion, the system is first reduced to three equations starting from a Taylor Series expansion. The movement of the ship in the plane of the water surface with the translations x, y and the rotation around z as well as the corresponding forces X, Y and the moment N are considered.
For the simulation of any maneuver, especially stop maneuvers, the entire speed range of u > 0 is relevant. Thus, in the solution of the Euler equations, not only the difference to the design speed u = u-u0 is considered, but also the respective speed u. Therefore, this approach requires the consideration of several ship speeds including u = 0 for the determination of the coefficients.
This rather high effort can be systematized. In addition, the numerous towing tank, circle tank or planar motion tests can be calculated with the help of modern RANSE (Reynolds-Averaged Navier-Stokes Equations) flow simulations. Applied systematically, they provide very good results in a reasonable time. An important stage of development is the definition of a feasible approach for the thrust and reverse thrust of the controllable pitch propeller. Here, the relatively free inflow to the propeller plane of a typical twin screw naval vessel proves to be beneficial because the wake fraction is comparatively small.
First numerical open water tests on the propeller over the entire pitch range are carried out. The curves of thrust and moment according to the pitch and the advance coefficient are conditioned and parameterized using polynomial approaches. In the ship simulation, the thrust or the thrust effect of the propellers are modelled with an equivalent axial volume force. Thus, the propeller-hull interactions are approximately determined. The resistance curves as a function of different magnitudes of propeller thrust or reverse thrust practically include the thrust deduction fraction. Due to this approach, new derivatives and coefficients emerge. The index »T« is chosen to show their dependence on the thrust.
For example, if the lateral force on the ship according to the lateral velocity is specified by the coefficient Yv, the new coefficient XT represents the longitudinal force according to the thrust.
The longitudinal force due to the rudder angle d, the thrust T and the ship speed u, for instance, is represented by XdTu and characterizes the interaction between propeller and rudder. Because the propeller jet also generates forces and moments at the hull and the rudder at ship speeds u = > 0, coefficients like YdTor Nd3T also emerge.
The number of coefficients depends on the physical conditions around the ship and the maneuvering requirements. It has to be decided up to which order of polynomial the curves of the numerical forces and moments are approximated. Oversimplification leads to a reduction in the number of coefficients and a quick solution, but can also lead to relatively large variations in the results. Another sensitive issue is the choice of the parameter range for speed, yaw rate or rudder angle. The full parameter range that is expected during the maneuvers should be captured in order to avoid uncertain extrapolations in the polynomial approximations.
The implementation of driving and maneuvering simulation is performed with the system tool Matlab Simulink. On the left, there is the control panel or the bridge. On the right, the different program blocks are described. The driving philosophies for ship operation with a high degree of automation, as well as the bow rudder or numerical wind forces and moments, can almost be implemented 1:1. The output options are manifold. In addition to the display elements in Simulink, typical circle or zigzag maneuvers in a 2D grid or trail plots on sea charts can be realized. Apart from the ship hydrodynamics, there are almost no limits to modelling mechanical and mechatronic components. To capture the previously described problem of shaft speed overshoot when stopping the vessel, the relevant physical parameters like moments of inertia according to the driving modes are implemented in the system of equations.
Since the adjustment and control times of all components are stored in the simulation, the crash stop maneuver can be calculated realistically. At maximum lever position, the propeller pitch linearly runs up to the maximum according to the adjustment characteristics. The actual revolution speed follows linearly up to the point where the limit of moment is reached. Since no pitch adjustment is done in the simulation yet, the actual speed does not reach the target speed. With the start of the crash stop maneuver, the maximum negative pitch is chosen. Depending on the speed, a range where the propeller is driven by the flow and acts as a turbine is exceeded. The simulation reflects this phenomenon of overshooting the actual speed. Here, the expected percentage is not negligible. The task was to develop an approach which guarantees a reduction in overshoot and simple and safe implementation in the automation.
An increase in the speed of pitch adjustment is virtually impossible. The introduction of additional inflection points in the pitch gradients and the definition of an optimum range and a suitable pitch respectively, with the help of an investigation of different variants, turned out to be successful.
Once again the benefits of the complex engineering come to the fore. The complexity of modern propulsion systems and related problems often cannot be mastered by the system supplier alone. Shipbuilders and shipyard are challenged more than ever to successfully merge the systems. The results of the simulation were verified on an actual ship. This allows the systematic implementation of corresponding changes in ship automation to ensure successful sea trials.
Authors: Markus Druckenbrod
Senior Manager Propulsion Systems,
Mechanical Engineering,
thyssenkrupp Marine Systems
markus.druckenbrod@thyssenkrupp.com
Christian Thieme, Project Engineer, Computational Fluid Dynamics,
Research & Development
thyssenkrupp Marine Systems
christian.thieme@thyssenkrupp.com
Markus Druckenbrod, Christian Thieme
