Determining Reliable Load Assumptions in Wind Turbines using SIMPACK

Transcription

1 SIMPACK AG, Friedrichshafener Strasse 1, Gilching, Germany SIMPACK News News SIMPACK September Customer Application Simulation of the Dynamic Behavior of Aircraft Landing Gear Systems 06 Customer Application Electronic Stability Program (ESP) for Trucks on the Daimler Driving Simulator 10 Customer Application Simulating Tank Vehicles with Sloshing Liquid Load 13 Customer Application Development of a SIMPACK User Routine for Dynamic Light Rail Vehicle Gauging Simulations 16 Customer Application Determining Reliable Load Assumptions in Wind Turbines using SIMPACK Simulation of the Dynamic Behavior of Aircraft Landing Gear Systems Simulation and the resulting prediction of the dynamic behavior of an aircraft and its landing gear system during ground maneuvers is an essential part in the design process. A realistic estimation of unwanted oscillations, such as gear walk and shimmy, for the landing gear and the whole aircraft can be readily obtained with an MBS-model. It is then possible to adjust the model for changes in the structural design of the airframe and the landing gear so as to optimize the... See page 2 Electronic Stability Program (ESP) for Trucks on the Daimler Driving Simulator Determining Reliable Load Assumptions in Wind Turbines using SIMPACK 20 Customer Application Simulation of Drivetrains on Wind Turbines within the Framework of Certification with SIMPACK 21 News SIMPACK Academy Events in Customer Application Coupling of MBS and CFD: an Oscillating Aeroelastic Wing Model 25 Software Gear Pair Enhancements with SIMPACK Version 8904 The Electronic Stability Program (ESP) safety system is a common feature in passenger cars nowadays and has proven to be an essential component in the reduction of traffic accidents. Electronic "stability control" and "rollover control" have been available since 2001 for commercial... See page 6 The intensified efforts to provide alternative and renewable sources of energy led to a substantial worldwide growth in the wind turbine industry over the last few years. The advantages of this clean energy source and the recent successes in increasing power output rates for on- and offshore wind turbines is extremely... See page SOftware Fatigue Analyses with FAT4FEM in SIMPACK's PostProcessor 30 Software 200 New Features and Improvements introduced with SIMPACK Version 8904

2 Customer Application Reinhard Lernbeiss, TU Wien, Institute of Mechanics and Mechatronics Simulation of the Dynamic Behavior of Aircraft Landing Gear Systems Simulation and the resulting prediction of the dynamic behavior of an aircraft and its landing gear system during ground maneuvers is an essential part in the design process. A realistic estimation of unwanted oscillations, such as gear walk and shimmy, for the landing gear and whole aircraft can be readily obtained with an MBS-model. It is then possible to adjust the model for changes in the structural design of the airframe and the landing gear so as to optimize the aircraft stability. INTRODUCTION Vibrations resulting from the elastic behavior or from dynamic loads may result in material fatigue and failure. Loads acting on the landing gear at touch-down are of major interest. Addressing these issues during testing is crucial. Elastic properties of the airframe and the landing gear have an essential influence on their dynamic behavior. Emphasis was placed on developing an elastic model of the whole aircraft so as to realistically simulate a complete landing and rollout with its braking, and in particular, the influence of the landing gear. To demonstrate the capabilities of this approach an existing aircraft Emphasis was placed on developing an elastic model of the whole aircraft... (Airbus A ) was selected. To control the aircraft during flare, touch-down and roll-out, a control system was introduced, capable of achieving any desired angle of the aircraft in relation to the runway at the exact moment of touch-down. This enabled the simulation of possible crosswind conditions as well as different landing techniques applied by the pilot. It was possible to pre-select certain values of residual vertical speeds upon touch-down. Additionally, the application of the brakes was performed by an automatic braking system combined with an anti-skid system. DYNAMIC PHENOMENA Undesired oscillations can occur in the longitudinal, lateral and yaw directions. Longitudinal vibrations are normally induced by changes of vertical and longitudinal loads acting on the wheels. They can be generated by landing impact or during braking and are commonly called gearwalk. The lateral and yaw oscillations are called shimmy oscillations when generated by self excitation forces. However, such vibrations may also be induced by asymmetric conditions occurring at landings with prevailing crosswind. Even the asymmetric structure of the landing gear itself, as occurs on most main landing gear systems, may be a source of unstable conditions, see Fig. 1. Rolling motions of the wheels about their longitudinal axis also exist. forward longitudinal-oscillation gear walk Fig. 1: Dynamic phenomena rolloscillation lateral-oscillation shimmy forward yaw-oscillation SIMULATION MODELS To generate a good approximation of the mass distribution of an aircraft like the Airbus A320 and the structural properties, existing and accessible data were used together with the statistical mass approximation method published by Raymer. These data were the main sources used to establish a CAD-model of all of the structures. To comply with the goal of simulating different loading cases of passengers, cargo and fuel, a balance calculation similar to those applied before each flight of the real aircraft was used. The data obtained for the air-frame structure was pre-processed to facilitate the generation of an elastic model in SIMBEAM. The elastic structure was assembled with the use of beam elements accounting for 2 SIMPACK News September 2010

3 Reinhard Lernbeiss, TU Wien, Institute of Mechanics and Mechatronics Customer Application Fig. 2: Airbus A320; structure of aircraft the main parts of the fuselage and the whole wings. Other elements, for example the foremost part of the fuselage in front of the nose landing gear, as well as the empennage, are considered rigid, but have the correct mass-properties and were attached to the corresponding elastic beam element. Implementing beam elements to generate elastic structures reduced the required time for simulation. Different beam elements were defined between certain markers and reflected a change in structural properties. In addition, those markers were set at positions to easily accommodate major mass concentrations and structural attachments, especially for the wings and the landing gear, see Fig. 3. To compensate for possible elements which have no structural influence, for example aircraft systems and fairings, the masses were equally distributed along the elastic beams between certain positions, and the mass was adjusted accordingly. Fig. 3: Arrangement of flexible beams of fuselage and wings SIMPACK News September

4 Customer Application Reinhard Lernbeis, TU Wien, Institute of Mechanics and Mechatronics 1 st Eigenmode 2 nd Eigenmode 4 th Eigenmode lift and drag forces were calculated using the basic aerodynamic equations acting on the corresponding surfaces belonging to each marker. The angle of attack is measured for each section of the wing considered, taking flexibility of the wings into account. With this aerodynamic model, neither static nor dynamic stability of the aircraft model is possible. This is accomplished by using artificial stability generated by a flight control model which simply generates forces on the horizontal stabilizer and elevator to produce flight stability and control. In the case of a real aircraft, the vertical speed upon touch-down is minimized. For the simulation, however, pre-determined vertical speeds are required for the intended parameter variations. Therefore, a schedule for the target vertical speed, dependent Fig. 4: Eigenmodes All other elements were added as rigid mass elements. This facilitated changing the loading of the aircraft and the amount of fuel for simulation of different landing mass and centers of gravity. Some of the resulting eigenmodes of the elastic aircraft model are shown in Fig. 4. LANDING GEAR A similar procedure was applied to establish elastic models of the landing gear system. The elastic structure of the landing gear model was comprised of a system of beam elements. The method of using distributed masses and mass concentrations, where needed, was also applied. To generate appropriate forces acting on the wheels and tires, a modified HSRI-tyre model was used. An automatic braking system in conjunction with an anti-skid system was used to achieve braking action during the rollout phase. The functions of these systems were basically reproduced by models programmed with MATLAB and Simulink using cosimulation. CONTROLLING OF THE LANDING MANEUVer To accomplish a realistic landing-flare, it is imperative to incorporate aerodynamic forces. However, one of the goals is to keep total simulation time low. Therefore, the aerodynamic model was kept as simple as possible and only those aerodynamic loads were applied which are essential to generate a nearly realistic landing maneuver and which have possible influence on the elastic structure with respect to the dynamic behavior of the landing gear. On the wings data aquisition SIMPACK output reverse thrust autobrake selection yaw & roll pre-selection Fig. 5 and 6: Landing simulation SIMPACK model aerodynamic wings flare trajectory aerodynamic hor. stabilizer engine thrust brakes & directional control data aquisition SIMPACK intput output 4 SIMPACK News September 2010

5 Reinhard Lernbeiss, TU Wien, Institute of Mechanics and Mechatronics Customer Application Fig 7: Nose landing gear; Airbus A320 upon height, was introduced. To simulate crosswind conditions upon landing and corresponding landing techniques used by the pilot or the automatic landing system, certain angles of roll and bank were selected. These angles were kept constant during flare until touch-down of the main wheels by a separate controlling system. SIMULATION SET-UP The model of the elastic aircraft structure together with the landing gear system in the MBS-software SIMPACK was simulated with the controlling system of the aircraft, programmed in Simulink using a co-simulation. Simulink was used for aerodynamic forces, anti-skid, autobrake system and steering on the runway during roll out. Thrust control during flare and reverse thrust were also provided, see Fig. 5 and 6. the design of the landing gear and for the flight test later in the development process. Therefore, it is imperative to conduct sufficient simulations of that test in advance to save resources and time. In addition, the simulation of drop tests of the landing gear with flexible structure models enables a touch-down with different side slips (yaw angles) and the motion of the de-rotation of the aircraft, which is the lowering of the nose after touch-down of the main landing gear. SIMULATION OF LANDING AN AIRCRAFT It is crucial to have sufficient model detail to gain insight into the dynamic behavior of landing gear. Fig. 8 presents a comparative study of the relative displacement and twisting of the wheel axis of the main landing gear using different modeling techniques of the elastic properties of the aircraft structure. It can be seen that the results may differ quite significantly. It is crucial to have sufficient model detail to gain insight into the dynamic behavior of the landing gear. CONCLUSION To facilitate the design process, it is advantageous to implement and use simplified models to simulate a number of operational aspects to prevent undesired and costly but necessary improvements during flight tests. It is of utmost importance to have an easy to use and changeable model at any stage of development to predict the behavior of the landing gear. It is then possible to modify the design or to make appropriate adjustments to a shimmy damper or similar device at an early stage of design. Implementing a flexible structure in the simulation model is essential. In addition, it is possible to produce a realistic landing maneuver using only limited application of aerodynamic calculations and save computational time to facilitate the design process by testing various structural configurations early in the design process. The model presented enables the use of real-time simulations used in flight simulators, and is necessary for aircraft that use an extensive amount of flexible composite components. DROP TEST SIMULATION During the design and development process, a so-called drop test conducted in a laboratory is used. Here a single real landing gear unit, loaded with the appropriate mass, falls onto a rotating drum which represents the moving runway surface. This test is a vital source of information and the data generated are used to optimize both Fig 8: Relative displacement of the wheel axis; different levels in modeling the aircraft structure SIMPACK News September

6 customer application Klaus Wüst, Daimler AG, CAE Commercial Vehicles Fig. 1: Mercedes-Benz Actros Safety Truck Electronic Stability Program (ESP) for Trucks on the Daimler Driving Simulator ESP for Commercial Vehicles The Electronic Stability Program (ESP) safety system is a common feature in passenger cars nowadays and has proven to be an essential component in the reduction of traffic accidents. Electronic "stability control" and "rollover control" have been available since 2001 for commercial vehicles. The system used for trucks and buses has two basic functionalities: The "stability control" functionality works similar to passenger car ESP systems. Sensors measure the steering wheel angle, the yaw velocity and the lateral acceleration of the vehicle and feed the values into a simple vehicle model. As soon as this model detects a big Fig. 3: Truck rollover oversteer Fig. 2: ESP stability control understeer difference between the desired and measured vehicle path, either in an oversteering or an understeering direction, individual wheels are braked to generate a moment around the vehicle s vertical axis, which stabilizes the vehicle movement (Fig. 2). For commercial vehicles, especially when loaded, these limits of adhesion are usually only reached on roads with a low coefficient of friction. For trucks and buses travelling on dry road surfaces, it is more important to prevent them from reaching the rollover limit (Fig. 3). The rollover control functionality limits the lateral acceleration of the vehicle to a pre-defined value which is dependent on the vehicle mass. When the vehicle reaches this maximum value, all the wheels of the towing vehicle and the trailer are braked to reduce speed. For Daimler trucks, the system is integrated into the electronic braking system and is sold as Telligent Stability System in a safety package together with other systems like lane keeping assistant and active cruise control. However, until now the system was only 6 SIMPACK News September 2010

7 Klaus Wüst, Daimler AG, CAE Commercial Vehicles customer application Fig. 5: Simulation-assisted ESP development available for tractor/semitrailer combinations, and the percentage of vehicles equipped with the system is still low. This is why the legislative committee of the European community has decided to make ESP systems mandatory for trucks and buses. The newly defined legislation ECE-R13 will go into effect this year starting with mandatory ESP systems for coaches and for tractor/ semitrailer combinations, where the system is already available. All other trucks and buses (apart from vehicles used for construction purposes), will have to follow by This makes it necessary for truck and bus manufacturers and braking system suppliers to develop the system for a wide variety of commercial vehicles. As the number of vehicles which can be used for proving the system functionalities in field testing is limited by material expense and time, it is necessary to use vehicle dynamics simulation to support the development of the system. A myriad of commercial vehicle parameters, such as different axle configurations, different wheelbases, tire variations, varied loading conditions, etc. can be incorporated into the simulation. Fig. 5: ESP field testing in reality and the respective SIMPACK model simulation SIMPACK News September

8 customer application Klaus Wüst, Daimler AG, CAE Commercial Vehicles ESP software-in-the-loop codes for SIMPACK vehicle models To study vehicle behavior with ESP in simulation, a code of the system has to be integrated into the vehicle dynamics simulation tool. The CAE analysis division of Daimler Trucks uses SIMPACK as a standard multibody simulation tool for different purposes. SIMPACK offers various possibilities for the integration of system codes into the simulation, e.g. the MatSIM interface for the integration of Simulink codes. The SIMPACK model consists of detailed component based models of axles and cab mounting. A detailed model was also used for the steering system. The frame includes torsional bending behavior. For the integration of truck ESP into SIMPACK, Daimler and the system supplier WABCO have chosen MATLAB and Simulink as an exchange and integration platform. The Daimler CAE analysis division uses the SIMPACK code export feature to generate a Simulink S-Function, which is combined with an S-Function of the ESP code, delivered by WABCO. The driving maneuvers and the evaluation of the results are generated within MATLAB and Simulink. Fig. 6: SIMPACK vehicle model maneuver SIMPACK Code Export ESP S-function results Fig. 7: Simulink simulation with exported SIMPACK model and ESP code Transfer of SIMPACK models to the Daimler driving simulator The Daimler driving simulator in Berlin has been in existence since 1995 and will be moved to Sindelfingen in A complete car or a truck cabin is installed on a hexapod which can additionally be moved in the horizontal direction. The movements of the simulator are generated by a vehicle simulation model. This simulator serves for investigations of the interactions between the driver and the vehicle. On one hand, it is used for subjective evaluations of driver assistance systems. For example, the brake assist that guarantees full braking application during emergency situations was developed based on simulator investigations. On the other hand, the simulator is also used for the subjective evaluation of parameter changes within the chassis layout during the development phase of new vehicles. The Daimler Trucks CAE analysis division has used the driving simulator since 2006, when the real-time capabilities of SIMPACK Fig. 8: Daimler driving simulator 8 SIMPACK News September 2010

9 Klaus Wüst, Daimler AG, CAE Commercial Vehicles customer application allowed the transfer of simulation models used for the chassis layout directly to the driving simulator (see SIMPACK News 2/08). Since then, several simulated driving tests have been conducted with SIMPACK models. The results were used to define target values for the vehicle dynamics of new truck or van generations and for the subjective evaluation of many chassis variants before field testing. ESP investigations on the driving simulator A first test with ESP for trucks on the driving simulator was conducted in November The test served as a basis for investigating if the simulator could give additional value to the simulation-supported development of ESP for trucks. The basic questions were if the simulator would be able to realistically reproduce the interventions of an ESP system for trucks, and if the simulator could be used for optimizing the system, e.g. for defining the intervention threshold values for different vehicles and various vehicle parameters. As the Windows-based MATLAB and Simulink Software-in-the-Loop (SiL) environment was not suited for the UNIX operating system of the driving simulator, a decision was made to use a different approach for this first test. A Fortran code, which simulates the basic ESP functionalities of stability control and rollover protection, was integrated by SIMPACK into a realtime vehicle model as a user routine. The model was then exported into the simulator environment together with the user routine, Fig. 9: Transfer of SIMPACK models to the driving simulator which then could be parameterized after the code export via the SIMPACK subvar file. This has the advantage of a flexible parameterization during the simulator tests. In addition to the parameters of the ESP system, vehicle parameters were also varied during the simulator test, i.e. loading conditions, steering parameters and tire characteristics....the combination of SIMPACK vehicle models with a Software-in-the-Loop code of the ESP system is able to realistically reproduce the stability interventions of the system... To evaluate the ESP interventions on the driving simulator, different driving situations were chosen, all based on straightahead driving on a highway track. Pylons were used for various lane-change and slalom maneuvers. With the aid of different sets of MF-Tyre tire parameters either a dry road surface or a low friction road could be simulated. As a main result, it was proven that the interventions of the ESP system on the driving simulator produced a realistic feel of the system. In addition, the influence of varied vehicle parameters and loading conditions on the interventions of the systems were shown to be realistic. Yet, for a complete evaluation of the ESP system on the simulator, the simple code used for the first test is not sufficient. For this, it will be necessary to integrate the complete system functionality, i.e. the interaction of the ESP system with the vehicle drivetrain. Results and Future work During a first test of a truck ESP system on the Daimler driving simulator, it could be shown that the combination of SIMPACK vehicle models with a Software-in-the-Loop code of the ESP system is able to realistically reproduce the stability interventions of the system and thus can be used for optimizations of the system for a wide variety of vehicles and vehicle conditions. In the next step, a complete WABCO code of the truck ESP will be compiled on a Linux system and integrated into SIMPACK as a user routine to obtain the complete functionality of the system on the driving simulator. Fig. 10: Lane change and slalom maneuver on the driving simulator SIMPACK News September

10 CUSTOMER APPLICATION Alexandra Lehnart, Florian Fleißner, Peter Eberhard, Institute of Engineering and Computational Mechanics, University of Stuttgart Simulating Tank Vehicles with Sloshing Liquid Load The cargo, particularly a liquid cargo, of a transport vehicle can have a significant influence on the driving characteristics of a vehicle. The design of the tank is of upmost importance as it greatly affects the dynamics of the cargo and, therefore, the dynamics and stability of the vehicle. These issues have to be taken into account when performing a simulation. A cosimulation approach is proposed coupling a multi-body system simulation using SIMPACK for the vehicle with a particle-based fluid simulation using the software PASIMODO for the cargo. INTRODUCTION SIMPACK is a very powerful tool in its application to coupled multi-body system (MBS) simulations of various vehicles. But when considering tank vehicles, the motion of the cargo and the tank design (including the number of compartments or the general shape) need to be accounted for, too. One way to approach this challenge is to use simple non-physical pendulum models for the cargo. However, this is just a crude approximation with little regard to the reality of the tank s shape and cargo dynamics. Much better results can be obtained by coupling the MBS simulation with another simulation method specifically designed to simulate the liquid cargo material interacting with the tank. In our approach, we couple SIMPACK with PASIMODO (PArticle SImulation and MOlecular Dynamics in an Object oriented fashion), a Lagrangian simulation framework for the 3D simulation of granular materials and fluids, developed at the Institute of Engineering and Computational Mechanics at the University of Stuttgart. Fig. 1: Tank vehicles with sloshing liquid load Fig. 2: Simulink model of a PASIMODO-SIMPACK co-simulation loop with fixed synchronization time interval 10 SIMPACK News September 2010

11 Alexandra Lehnart, Florian Fleißner, Peter Eberhard, CUSTOMER APPLICATION Institute of Engineering and Computational Mechanics, University of Stuttgart Fig. 3: Model of the truck with 17 degrees of freedom DYNAMIC CO-SIMULATION The co-simulation delegates the simulation of the tank and silo vehicles to the two specialized programs. PASIMODO calculates the forces on the tank geometry resulting from the sloshing liquid or moving solid cargo, while SIMPACK calculates the behavior of the vehicle. For this purpose, SIMPACK needs to know the particle forces and torques on the tank to account for their influence on the vehicle s dynamics. On the other hand, PASIMODO needs the recent tank states to be able to calculate the resulting particle forces. This exchange is established via MATLAB and Simulink as SIMPACK already provides a co-simulation interface that allows for data exchange. To couple PASIMODO with MATLAB and Simulink, we took advantage of PASIMODO s plug-in interface, which enables users to implement their own custom subroutines in C++. These can be used to reposition geometries during a simulation, e.g. the tank s internal surface geometry according to the recent tank states received from SIMPACK. Then, the plug-in sends the calculated particle forces and torques with respect to the tank s center of gravity back to SIMPACK. The data exchange is carried out via a TCP/IP interface both between PASIMODO and Simulink as well as between Simulink and SIMPACK. The exchange always takes place after a fixed time interval, but in between, each simulation program can use its own variable time stepping. This is especially relevant for the particle simulation as the particle dynamics solver usually has a much smaller time step than the dynamics solver of the truck. For these time steps, the position of the tank geometry is updated via extrapolation from the already received tank states. Fig. 2 depicts the co-simulation procedure as implemented in the Simulink model. PASIMODO calculates the forces on the tank geometry, while SIMPACK calculates the behavior of the vehicle. VEHICLE MODEL For the simulation of the vehicle, a classical MBS approach is used. The model of the truck has 17 degrees of freedom, schematically shown in Fig. 3. The computation of the tire forces follows the Tire Similarity Model from Pacejka. To account for a driver, two additional degrees of freedom are added which are influenced by a feedback controlled driver model. The truck has a total dead weight of kg, a wheelbase of 3.1 m and a track of 1.66 m. PARTICLE MODEL Considering the MBS model and the particle model, different aspects of the vehicle are of interest. In SIMPACK, the forces and torques on the tank resulting from the moving cargo are supplied by the particle simulation. They can be treated as applied forces on the tank s center of gravity and no further information on the tank s geometry is needed. The particle simulation, on the other hand, needs a detailed internal tank geometry and the states of the tank to place it appropriately, but doesn t need any information about the remaining MBS. The program PASIMODO follows the Discrete Element Method approach for the simulation of granular systems and the Smoothed Particle Hydrodynamics (SPH) method for fluids. SPH is a discretization method in space for partial differential equations, where the discretization points, called particles, carry the important information like velocity, density and pressure as mean values over the volume surrounding them and move along the velocity field of the material they represent. The discrete values at the particle positions are then smoothed by means of a differentiable kernel function, often similar to a Gaussian function, and summed up to yield an approximation of the corresponding function that is also differentiable. Using these approximations in the partial differential equations gives ordinary differential equations in time that can be solved with a given time stepping scheme. Fig. 4 shows some moving particles with their corresponding kernel functions. Boundaries and non-spherical objects are defined by triangular surface meshes. To prevent particles from passing through the mesh, a penalty force is evaluated and applied for each pair of particle and triangle. If the mesh is not fixed, kernel function Fig. 4: Exemplary particles with their corresponding moving kernel functions SIMPACK News September

12 CUSTOMER APPLICATION Alexandra Lehnart, Florian Fleißner, Peter Eberhard, Institute of Engineering and Computational Mechanics, University of Stuttgart them a full braking with different tank designs. As cargo, water is used with a total mass of 4700 kg. Regarding the evaluation of the simulation results, the co-simulation approach offers all the possibilities that SIMPACK offers, together with those from PASIMODO. The exchanged data can be stored by MATLAB and loaded later to compare results from different simulations. This includes the states and forces for every body in the simulation. PASIMODO offers the additional ability to track the cargo s center of gravity. Fig. 5 depicts some snapshots taken from a full braking with an undivided and tripartite tank, respectively. Please note that in the visualization particles moving within the liquid are displayed. The corresponding displacement of the cargo s center of gravity is shown in Fig. 6. In Fig. 7 the tire locking that occurs during full braking with the undivided tank can be seen, while the compartments in the other tank prevent it. Regarding the evaluation of the simulation results, the co-simulation approach offers all the possibilities that SIMPACK offers, together with those from PASIMODO. CONCLUSION The results obtained show that the SIMPACK-PASIMODO co-simulation is suitable for predicting the stability of driving maneuvers of transport vehicles. It can be used to investigate the impact of different tank designs on the behavior of the vehicle system, easily taking into account compartments or curved walls. Comparisons between different tank designs show the significant influence that the movement of the cargo and the tank design have on the driving characteristics of the vehicle. A positive effect of subdivisions in the tank in longitudinal directions in terms of braking stability can be seen, as they reduce the sloshing motion in that direction. For details about PASIMODO please see pasimodo/pasimodo_en.php. More information about the methods and simulations is provided in the authors publications. Fig. 5: Snapshots during a full braking maneuver with a tank with an elliptical cross section and one and three compartments, respectively Fig. 6: Normalized displacement of the cargo s center of gravity in longitudinal direction it is defined relative to a moving coordinate system that lies in the center of gravity and accumulates all the forces and torques on the triangles. The position and orientation of the coordinate system and with it the mesh geometry can then be either integrated in time or, as in our case, taken from the tank states supplied by SIMPACK. The accumulated forces and torques are sent to the MBS simulation. RESULTS To investigate the influence of moving cargo on driving dynamics, classical benchmark driving maneuvers are performed, among Fig. 7: Relative angular velocity at the rear tire 12 SIMPACK News September 2010

13 Gero Zechel, Michael Beitelschmidt, Dresden University of Technology; Customer Application Helmut Netter, Bombardier Transportation Fig. 1: MBS model of a five car 100% low floor tram Development of a SIMPACK User Routine for Dynamic Light Rail Vehicle Gauging Simulations More and more cities are seeking public transportation systems that offer safe, comfortable and effective mobility integrating seamlessly into the urban landscape. To meet this demand, made-to-measure vehicles have to be developed that allow high capacity cars on narrow spaced infrastructure. In a joint venture between the Chair of Vehicle Modeling and Simulation at the TU Dresden and LRV (Light Rail Vehicle) Vehicle Engineering at the Bautzen site of Bombardier Transportation, a SIMPACK User Routine has been developed that obtains the dynamic vehicle envelope for arbitrary train/track combinations with the push of a button. LIGHT RAIL VEHICLE GAUGING The complex kinematic and highly dynamic behavior of modern articulated light rail vehicles require elaborate research of the resulting dynamic vehicle envelope and the structure gauge needed to rule out collisions of vehicles with other vehicles and trackside objects. The variety of vehicle configurations, the uniqueness of infrastructures, and the generally low vehicle quantities common in light rail demand flexible and efficient gauging methods. While kinematic studies are crucial for the early conceptual design, all dynamic factors have to be included for design verification at the end of the development phase. Conducting early gauging with simplified kinematic models and integrating gauging simulations into full-scope multi-body simulations, like the one shown in Fig. 1, throughout development of a vehicle is highly beneficial. INPUT CONCEPT Even for sophisticated MBS models, the outer shape of the vehicle can be completely ignored. For gauging, however, the precise vehicle contour is vitally important. To make it available in the models without relying on CAD data, SIMPACK Input Functions have been chosen. They can be used to describe the contour as the car width over the car center plane. To define the origin of these contours on the car, body fixed markers can be created separately. This way, the contours can be created in Fig. 2: The vertical grid (blue) on top of the track centerline (blue, dotted) used as the reference system for all track related data collected during a run SIMPACK News September

14 Customer Application Gero Zechel, Michael Beitelschmidt, Dresden University of Technology; Helmut Netter, Bombardier Transportation Gauge +Y / m result.kintest_wgt1.gauge +Y over Track result.kintest_wgt2.gauge +Y over Track result.kintest_wgt3.gauge +Y over Track s_track / m y_contour / m Gauge +Y / m result.kintest_wgt1.contour +Y over Contour X result.kintest_wgt1.gauge +Y over Contour X x_contour / m Fig. 3: The maximum lateral distance of the first three cars to the track centerline, plotted over the track length Fig. 5: Grey: The left side contour of the first car (view from above, car pointing to the right). Red: The dynamic envelope of the car during the whole run (relative to the car reference system) Gauge +Y / m result.kintest_wgt1.gauge +Y over Track result.kintest_wgt1.causative Contour Point for +Y s_track / m Fig. 4: The first car s maximum distance along the track (black), along with a curve indicating the contour point responsible for the maximum value (blue) substructures, varied using substitution values, stored in databases, and reused over several cars or models. For gauging simulations using these contours a SIMPACK User Routine has been developed. IMPLEMENTATION To achieve a seamless integration into the SIMPACK user interface, the concept was realized by means of a SIMPACK User Result Element written in Fortran using the built-in editing and compiling tools with a third-party compiler. This User Result Element can be integrated in any vehicle model by using the Model Setup GUI. In the Result Element window, the Input Function and the Marker can be picked from the list of available MBS elements and output options can be set. Although one Result Element represents only a single contour, it can be added as many times as needed to cover the whole vehicle. In general, Result Elements are evaluated offline when Measurements are performed. They are called by SIMPACK beforehand for initialization, then at every communication point, and afterwards to write the output vectors. While the measurements are running, four data collections are built up and stored in memory allocated by SIMPACK using the available Access Functions. The first data collection keeps state information x_contour / m The concept was realized by means of a SIMPACK User Result Element. that is needed for every time step, like the number of contour points, the track length, and a time step counter. The second data collection is more complex it stores all values related to the infrastructure. To do so, a grid is spanned over a vertical surface positioned on top of the track centerline, as shown in Fig. 2 in blue. Every point in the area around the track can be assigned to a unique grid point as long as the distance from the track is less than the smallest track radius this covers the relevant area at all times. The maximum centerline distance any contour point reaches within each of the resulting volume segments (Fig. 2, shown in red) is gathered during the measurement run. To get this information, an algorithm is run for each time step that circles through all contour points one by one and calculates their distance to the grid using an iterative method. If a new maximum distance is found along the track, this value, as well as the index of the contour point responsible for the new maximum and the absolute position of the contour point, is stored in the memory. The third data collection is filled in a similar way but stores values in reference to the contour itself. At the end of the run, it therefore contains the maximum distance from the track for each contour point along with the information at which grid point this maximum occurred. Finally, the fourth data collection gathers time-domain data like the absolute position of each contour point to certain time steps. While memory usage stays reasonably low, these four data collections are not suitable for plotting as is, so they are evaluated at the end of the run to gain comprehensive one-dimensional output vectors. This data is stored in the resulting standard result-file and can be easily plotted using the SIMPACK PostProcessor. DATA EVALUATION AND PLOTTING First and foremost, the track related data collection is evaluated to get the maximum distance of the vehicle to the track centerline over the track length. Two output vectors are created for the left and for the right hand side of the track. Fig. 3 shows the results for a kinematic vehicle model passing through a right turn. Three Result Elements where defined for the first three cars of the vehicle. In the plot, the results for the left-hand side of the track for each car are overlaid by drag-and-drop in the SIMPACK PostProcessor. The plot shows that the maximum distance from the track at curve entry and 14 SIMPACK News September 2010

15 s of Track Gero Zechel, Michael Beitelschmidt, Dresden University of Technology; Customer Application Helmut Netter, Bombardier Transportation Gauge -Y over Track windloadonstraight.result.gaugingexampleq.trackslice s= H(Y) nowindloadonstraight.result.gaugingexampleq.trackslice s= H(Y) TopDownView X(Y) result.gauge_wgt1.topdownview Track Centerline X(Y) result.gauge_wgt1.topdownview Movement Visualiz. X(Y) result.gauge_wgt2.topdownview Movement Visualiz. X(Y) result.gauge_wgt3.topdownview Movement Visualiz. X(Y) result.gauge_wgt1.topdownview Envelope -Y X(Y) result.gauge_wgt1.topdownview Envelope +Y X(Y) result.gauge_wgt2.topdownview Envelope -Y X(Y) result.gauge_wgt2.topdownview Envelope +Y X(Y) result.gauge_wgt3.topdownview Envelope -Y X(Y) result.gauge_wgt3.topdownview Envelope +Y X(Y) TopDownView Indep Y Fig. 6: Structure gauge for a vehicle on a straight track with and without crosswind load Fig. 7: Top-down view of an S-bend track showing the track centerline, the position of three car contours at two different time steps, and the kinematic envelopes of each car projected to the ground part of the curve exit is caused by the second car (red), while the first car (black) leads to the maximum distance along the track between curve entry and exit (1.54 m). To provide information about what part of a contour is responsible for maximum displacement, the track related data collection is also filtered to produce the output vector shown in Fig. 4. The plot indicates that the contour point located 3.76 m in front of the bogie (where the body fixed marker is placed) results in the maximum distance for the given example. Finally, the evaluation of the contour-related data collection provides the outputs shown on Fig. 5. The left side contour of the first car is depicted in grey, the dynamic envelope of the car during the whole run is shown in red. This plot confirms the values found above, and shows that the head contour needs to be changed to lower the maximum distance to 1.4 m from the track centerline (everything in the range of 3.5 m to 4.3 m in front of the bogie). By applying more advanced evaluation techniques to the track related data collection, plots are made possible that show the structure gauge needed by the vehicle along the track. This can be helpful when analyzing the rolling behavior in curves, the clearance at railway platforms and the influence of crosswind loads, Fig. 6, for example. If gauging data is needed in the context of the simulated scenario, absolute positions stored in the track-related and the time-domain data collection can be combined to get a general idea of the vehicle s kinematic and dynamic gauging behavior. Fig. 7 shows an example plot of three cars passing an S-bend. In addition to the track centerline and the car contours at two different time steps, the kinematic envelope of each car projected to the ground can be seen....simpack s built-in tools and functionality can be used for gauging simulations with high accuracy... CONCLUSION When combined in the right way, SIMPACK s built-in tools and functionalities can be used for gauging simulations with high accuracy, even for highly dynamic situations like the one shown in Fig. 8. The complexity of the Gauging User Routine can be reduced to consider just one single contour, because it can be utilized many times in a single model at the same time and the results can be overlaid in the PostProcessor. Its use, however, is not restricted to the car body it can also be applied to pantographs, bogie parts, or even markers fixed to the inertial system to include objects of the infrastructure like railway platforms or masts of the catenary wire. This extreme versatility is possibly the biggest advantage compared with many other gauging software tools as it allows for the analysis of a huge variety of vehicle designs and infrastructure systems which LRV-engineering has to cope with every day. Fig. 8: Dynamic model of a 5-car tram passing a gate after a curve (top view) SIMPACK News September

16 Customer Application Berthold Schlecht, Thomas Rosenlöcher, Institute of Machine Elements and Machine Design, Chair of Machine Elements, Dresden University of Technology Determining Reliable Load Assumptions in Wind Turbines using SIMPACK INTRODUCTION The Chair of Machine Elements at Dresden University of Technology does research which, for the past several decades, has focused on machine elements like shafts, gearings and bearings. Since 2001, MBS software has played an integral role in the development of guidelines, standards and verification of the dynamic behavior of drivetrains. MBS software has been used to develop modeling strategies that realistically represent the dynamic behavior of machine elements and have a high correlation with measured data. The main focus of research has been the investigation of the dynamic behavior of large drivetrains which can be found in roller mills, compressors, ships, fans, shearers, cranes and wind turbines. The intensified efforts to provide alternative and renewable sources of energy led to a substantial worldwide growth in the wind turbine industry over the last few years. The advantages of this clean energy source and the recent successes in increasing power output rates for on- and offshore wind turbines is extremely encouraging. However the need for more durable wind turbines must still be addressed. The operation of anchored flexible light weight constructions under high dynamic stochastic loads was a relatively new challenge that arose with the wind turbine. These challenges are being met with the aid of the multi-body system (MBS) software SIMPACK. Fig. 2: Gear box substructures MOTIVATION In comparison to the wind turbine design software used by wind turbine manufacturers, multi-body system simulation software allows for a more precise modeling of the drivetrain components. Instead of a detailed rotor model and a simple 3-mass torsional vibration model for the gear box and the generator, all components can be represented with up to six degrees of freedom and with coupling stiffnesses. The resulting simulation model of the wind turbine consists of the substructure's rotor, drivetrain, coupling, generator, supporting structure, tower and foundation (Fig. 1). It allows the determination of the natural frequencies of the complete structure, and additionally, shows the mode shapes of all drivetrain components. To detect possible ranges of resonances, the effects of the rotor, the gear meshing frequencies, and rotation speeds of the components can be compared to the calculated natural frequencies. Simulation in the time domain enables the determination of torques, forces, displacements, velocities and accelerations for the modeled components and degrees of freedom. The resulting values at the different load conditions can be used for the design of the components, bearings and gearings, as well as the gear box housing and supporting structure. Fig. 1: Flexible multi-body system model of a wind turbine drivetrain BASICS OF DRIVETRAIN SIMULATION The analysis of drivetrains operating under high dynamic loads presupposes the assembly of a detailed simulation model which is able 16 SIMPACK News September 2010

17 Berthold Schlecht, Thomas Rosenlöcher, Institute of Machine Elements and Machine Design, Customer Application Chair of Machine Elements, Dresden University of Technology Fig. 3: Modeling of shafts by discretisation, SimBeam model and FEM-method to represent the dynamic behavior of the drivetrain in the frequency and time domain. Even if high performance computers are available the level of detail of the simulation model has to correspond to the formulated question to ensure a feasible The degrees of freedom of each substructure can be adjusted to accommodate varying degrees of detail for the overall model. Submodeling enables easy verification of the appropriate level of model detail. MODELING OF Drivetrain COMPONENTS The dynamic behavior of a drivetrain results from the gear ratio and the distribution of the mass, mass of inertia and stiffness. The determination of mass parameters is possible with three-dimensional CAD models or by using simple analytical approaches. Great effort is required to accurately calculate the various stiffness parameters for all of the drivetrain components. The torsional stiffness of the drivetrain is mainly characterized by the stiffness of the shafts. Special consideration must be given to slender shafts whose elastic properties need to be accounted for. Additionally, the bending stiffness of such shafts may have considerable influence on the dynamic behavior as well as the resulting displacements. The required simulation model can be assembled in three ways: a) by using the method of discretization, b) via the beam approach or c) by implementing modally reduced finiteelement models (Fig. 3). For models that include the axial and radial motion of the shafts, the properties of bearings must be accounted for. Essentially, the modeling of the bearings is realized by a force element which introduces the reaction forces in the axial and radial directions as well as the reaction moments, if necessary. The bearing properties can be described by average bearing stiffness, characteristic curves or complex models imported as DLLs. The transmission ratio between the rotor and the high speed generator can be realized by a gear box consisting of a set of planetary and helical gear stages (Fig. 4). The changing speeds and torques in a gear box as well as the varying gearing stiffness resulting from the total overlap ratio have an important influence on the dynamic behavior of the drivetrain and must be considered in the simulation model. SIMPACK offers the special force element Gear Pair (FE 225) The degrees of freedom of each substructure can be adjusted to accommodate varying degrees of detail of the overall model. Fig. 4: Gear box of a 3 MW wind turbine calculation effort. Based on the available data and the experience of the engineer, a discrete simulation model can be assembled. A successive and modular assembly of fully parameterized simulation models allows a clear and reproducible modeling process. The modular concept requires decomposition of the drivetrain into its substructures. Using this approach, a simulation model of a common wind turbine consists of the following substructures: rotor, main shaft, coupling, generator and an additional subdivided gear box. For the gear box, a combination of helical/spur and planetary gear stage substructures is necessary (Fig. 2). Each substructure consists of model components which can be subdivided into shafts, gear stages, bearings and supporting structures. This combination of single substructures leads to the complete simulation model of the wind turbine. Fig. 5: Variable gearing stiffness over the contact path using Fourier coefficients SIMPACK News September

18 Customer Application Berthold Schlecht, Thomas Rosenlöcher, Institute of Machine Elements and Machine Design, Chair of Machine Elements, Dresden University of Technology which enables a comfortable modeling of gearing. An alternative modeling approach offers the mathematical description of the resulting forces in the gearing by user routines. Based on the calculation of the tooth normal force in the ideal pitch point, the complete tooth contact is simplified and described in one point. The tooth normal force consists of stiffness and damping dependent parts. Information about the displacements and velocities in tangential, radial and axial directions resulting from the relative position of the gears can be determined from the joint states and the corresponding trigonometric relationships. The gearing stiffness can be considered as average contact stiffness according to DIN 3990 and variable gearing stiffness over the path of contact using Fourier coefficients (Fig. 5). In order to improve the model even further, modeling of the coupling and generator is necessary. The most important influence on the dynamic behavior results from the rotorblades. A simplified approach for modeling of the rotorblade stiffness can be done by splitting the blades into mass segments coupled by spring damper elements to represent the bending stiffness (Fig. 6). Therefore, the information of the mass and stiffness distribution as well as the natural frequencies in edge- and flapwise direction are sufficient. In addition, if the information of every profile section is available, the rotorblade generation in SIMPACK offers an automated procedure. A detailed modeling of the shafts, bearings, gearings, coupling, generator and rotor allows the representation of the dynamic interaction between the components and the determination of displacements, velocities, accelerations, forces and torques. Even if spring-damper elements are used to support the components, the surrounding structure, e.g. the gear box housing or the main frame, is assumed to have a stiff coupling to the global reference system. Thereby the influences resulting from the flexible structure of a wind turbine are neglected. The enhancement of the stiff multi-body system model using modally reduced finite-element structures allows the consideration of these effects. The implementation of a flexible structure in SIMPACK is based on a meshed finite-element model of the component geometry and the definition of the material properties. The connection points to the SIMPACK offers the special force element Gear Pair (FE 225) which enables a comfortable modeling of gearing. supporting spring-damper elements can be modeled by means of multipoint constraints (MPC). Due to the resulting complexity and degrees of freedom of the FEM-models a reduction of the structure is required. The application of the approach according to Craig-Bampton requires the definition of the connection points between the flexible structure and the rigid bodies. The mode shapes of the reduced model are used to determine the deformation under load. The number of natural frequencies chosen for the modal reduction defines the valid frequency range and the accuracy of the model, which is also influenced by the choice of frequency response modes in the SIMPACK Add-On Module FEMBS. The implementation of flexible structures allows representation of the flexibility of the supporting structure as well as the consideration of the stiffness of the drivetrain components, e.g. shafts and planetary carriers, with a higher degree of accuracy. ANALYSIS OF NATURAL FREQUENCIES AND EXCITATIONS The resulting flexible multi-body system model allows the determination of the natural frequencies and can take into account various degrees of freedom. The resulting frequencies can be compared to the excitation frequencies to determine possible resonances. Relevant excitations are the first, second, third, sixth (ninth, twelfth) order of the rotor rotation frequency. Additionally, the first and second orders of the rotation frequencies of all drivetrain components, as well as the meshing frequencies of the gear stages, have to be considered. The comparison of natural frequencies and excitations by means of a Campbell diagram reveals possible resonances. The analysis of the corresponding mode shapes allows further statements as to whether the excitation of a natural frequency can cause critical operation points or not. Fig. 6: Discretisation of a rotorblade ANALYSES IN THE TIME DOMAIN The detailed flexible multi-body system model also offers the possibility of determin- Fig. 7: Wind loads on flexible rotorblades 18 SIMPACK News September 2010

19 Berthold Schlecht, Thomas Rosenlöcher, Institute of Machine Elements and Machine Design, Customer Application Chair of Machine Elements, Dresden University of Technology ing resulting displacements, deformations, velocities, accelerations, forces and torques under the dynamic loads resulting from the stochastic wind field. To model a realistic wind field, different approaches are available. The common wind turbine design software tools like Bladed, Flex5 and AeroDyn are mainly used to define the design loads. An interface between AeroDyn and SIMPACK is available (FE 237: Wind AeroDyn*)which offers the possibility of determining wind loads based on the rotorblades, tower and wind turbine parameters. If all the required information is not given, different simplified approaches based on measurements can be used. A wind model based on the information of the wind speed, power coefficients and assumed load distributions over the blade length and the height can also be used. Using nine segments for each blade and by superimposing a stochastic wind field, the tower shadow can be calculated. An enhancement of this approach replaces the assumptions for the wind field and load distribution by measurement results captured as torque and bending moments at the main shaft. An algorithm adjusts the single blade forces to achieve the measured results on the main shaft in the simulation model so that the measured states of the wind turbine can be represented by the model. The changing forces at the flexible modeled rotorblades are shown in Fig. 7 as arrows and scaled deformations. The operation of the wind turbine under different load conditions and extreme load cases like emergency stops can also be calculated. The resulting speeds, torques and forces in the gearing of the first planetary gear stage are shown in Fig. 8. In addition, the information of displacements of the main shaft, gear box and gear box components like the sun shaft (Fig. 9) can be obtained. This data can lead to a deeper understanding of the dynamic behavior of the system and the resultant loads under different load conditions. Fig. 8: Emergency stop simulation, speed, pitch angle and force CONCLUSION Wind turbines are anchored flexible complex drivetrains which operate under highly dynamic stochastic loads. For onshore wind turbines, and especially for offshore wind turbines, the request for high reliability requires comprehensive knowledge of dynamic behavior already in the design phase. This includes information about possible excitations and natural frequencies which can cause resonances in the operational speed range. Additionally, the displacements, deformations and resulting forces in the drivetrain, as well as the influence of wind turbine control under maximum loads during normal operation and emergency cases, has to be analyzed. The multi-body method offers the ability to realistically model a wind turbine while considering all relevant components and degrees of freedom. This approach enables the required knowledge to be obtained in order to fully understand the dynamics of wind turbines. Fig. 9: Displacement of the sun SIMPACK News September

20 CUSTOMER APPLICATION Markus Kochmann, Milan Ristow, Germanischer Lloyd Industrial Services GmbH Simulation of Drivetrains on Wind Turbines within the Framework of Certification with SIMPACK Simulation with multi-body systems is relatively new in the emerging wind energy industry. Representative tests on a constructed model are practical only on a limited basis because of the long draft lifespan of 20 years. Realistic simulation is of paramount importance in product development. Programs such as Bladed or Flex5 are well suited for calculating loads in order to verify the general stability of a wind power plant. These programs have their strengths in the investigation of the dynamic behavior of the overall system. However, the analysis is usually limited to a small range of frequencies. The investigation of phenomena in a larger range of frequencies requires a more detailed representation of the system. SIMPACK simulation software provides an excellent platform for this detailed model and analysis. The drivetrain of a wind turbine is an integral and expensive assembly component of the plant. Manufacturers of transmissions, manufacturers of wind turbines and certification bodies (such as Germanischer Lloyd Industrial Services GmbH, Renewables Certification (GL)) engage in expansive efforts to put a reliable design into practice. One critical aspect of the design is the dynamic behavior of the drivetrain, which plays an important role in the determination of local loads and the physical integrity of the mechanical components. type certification For several years, the investigation of the dynamic behavior of the drivetrain has been part of the Type Certification of wind turbines. A type certification is the confirmation of conformity for adherence to fixed requirements (e. g., guidelines and standards) for certain types of wind turbine. An important component of a type certification is a thorough design assessment. The certification procedure is based on the international standard IEC [1], [2] and/or the GL guidelines for the certification of wind turbines [3], [4] and [5]. component of the design assessment As a component of the design assessment, an independent evaluation of the dynamic behavior of the drivetrain is conducted by SIMPACK is used by GL in order to investigate the dynamic behavour of the drivetrain. GL. The necessary model data are derived from technical drawings, CAD and FEM models. On the basis of such information, a realistic representative multi-body simulation model can be prepared. SIMPACK is used by GL in order to investigate the dynamic behavior of the drivetrain. In July 2010, the revision to guideline [4] was published as "GL 2010" [5]. Experiences from various certifications, research projects, discussions between GL and external experts and, above all, the technical specialized committee (with many specialists from the wind energy industry) have led to a new issuance which considers state-of-theart knowledge regarding the development of wind turbines. In the "GL 2010" guidelines [5], an entire application-oriented appendix is dedicated to the investigation of dynamic behavior of the drivetrain. The recommendations assume investigations with the use of multibody simulation systems. The study of the dynamic behavior of the drivetrain on the basis of "GL 2010" will lead to the following changes: Multi-body simulation modeling Multi-body simulation models are frequently used for the determination of natural frequencies and the investigation of the dynamic behavior of the drivetrain. The possible degrees of detail for such models range from simple mass shock absorber systems, with rotatory degrees of freedom, up to very complex systems with flexible bodies and super-elements for the consideration of the flexible housings and support structures. Often a simple model provides sufficient information regarding the dynamic behavior of a complex dynamic system and helps with the investigation and understanding of dynamic phenomena. However, for special problems, more detailed and more complex models are necessary. Fig. 1: 3D MBS model of wind turbine (including flexible blades, detailed drivetrain and generator) 20 SIMPACK News September 2010

21 Markus Kochmann, Milan Ristow, Germanischer Lloyd Industrial Services GmbH Customer application One emphasis is the measurement of a drivetrain of a wind power plant of the megawatt class. At the same time, results arising from simulations with SIMPACK are compared with the measurement results. Further information is available at the Internet site for the project (see Conclusion In order to obtain a more in depth view of the dynamic behavior of complex systems (e.g., the drivetrain of a wind power plant), a clear trend to more complex and validated simulation models is imperative. SIMPACK is high-performance and reliable simulation software is an important tool for achieving this goal. Fig. 2: Detailed SIMPACK model Torsion, bending and axial degrees of freedom must be considered. Simulation models with only rotational degrees of freedom can be used (if the calculation results are verified with measurements). For certain analyses, the simulation of a run-up must be conducted in the time domain. In part, such new requirements are arising from current EU cofunded research projects that are concentrating on the validation of simulation models through measurement. Together with manufacturers of wind power plants and transmissions, research institutions and universities, GL is involved with the PROTEST research project. Information For more information about GL and renewables please see: References [1] IEC IEC Wind turbine generator systems Part 1: Safety requirements, nd edition, February [2] IEC Wind turbines Part 1: Design requirements, 3rd edition, August [3] Germanischer Lloyd, Hamburg, Germany: Guideline for the Certification of Offshore Wind Turbines, 2005 Edition. [4] Germanischer Lloyd, Hamburg: Guideline for the Certification of Wind Turbines, Edition 2003 with Supplement [5] Germanischer Lloyd, Hamburg: Guideline for the Certification of Wind Turbines, 2010 Edition. SIMPACK AG SIMPACK NEws SIMPACK Academy Events in 2010 MBS Numerics Academy, , in Andechs, Southern Germany Prof. Dr. Martin Arnold, Institute of Mathematics, Martin Luther University Halle-Wittenberg and Dr. Gerhard Hippmann, Solver Technology, SIMPACK AG Multi-Body System Numerics aims to enable the calculation engineer to understand the capabilities and limits of numerical methods in multi-body dynamics. Elaborate presentation of the theoretical background is combined with software issues as well as practical examples and tips. Wind Turbine and Drivetrain Academy, , in Gross Schwansee, Northern Germany Part 1: Dynamics and System Design of Wind Turbines, Prof. Martin Kühn, ForWind University of Oldenburg and Stefan Hauptmann, University of Stuttgart This course combines theoretical background with recent industrial experience to provide a comprehensive introduction to state-of-the-art wind turbine dynamics and design from a system s view-point. Part 2: Design and Analysis of Drivetrains in Wind Turbines and Other Large Industrial Applications, Prof. Berthold Schlecht and Thomas Rosenlöcher, Technical University of Dresden This course covers the dimensioning and design of drivetrain systems and their components. In addition, the advantages and accuracy of multi-body simulation (MBS) will be shown. For more information and registration please visit: SIMPACK News September

22 Customer Application Jürgen Arnold, Wolf Krüger, Gunnar Einarsson, Deutsches Zentrum für Luft- und Raumfahrt, Göttingen Coupling of MBS and CFD: an Oscillating Aeroelastic Wing Model Multi-body simulation has been shown to be a valuable software tool for virtual aircraft design. It is a standard approach for the analysis of landing gears, of aircraft on the ground, and for the design of high-lift systems. The medium level of complexity of typical multi-body models also makes it a suitable tool for the application of flight mechanics in combination with elastic deformations. The development of reliable aerodynamic models, in addition to the existing interface for complex elastic structures, has been a major activity in the DLR Institute of Aeroelasticity during the past several years. The coupling procedure of SIMPACK to CFD is shown in this article for the application of the Aeroelastic Model Programme (AMP) wind tunnel wing simulated with SIMPACK and the DLR TAU code. Fig. 1: CFD model of the AMP wing AEROELASTIC SIMULATION USING SIMPACK AND CFD Aeroelastic simulations in terms of pure fluid-structure interaction have reached a satisfactory level of maturity for both steady and unsteady problems. A step beyond this classical scope is the additional consideration of large motions superimposed by flight maneuvers. In DLR, such a coupling has been developed based on elastic multi-body systems coupled with CFD calculation. The intention of this article is to describe the model set-up used for the coupled calculations, as well as to describe the options to introduce MBS-generated motions into the CFD calculations. Results are given for the so-called AMP wind tunnel model for heave and pitch oscillations at a Mach number of 0.6. A more detailed illustration of the work as well as a list of references is given in [1]. GENERAL SET-UP For the simulation of a complete elastic aircraft using MBS, the flight mechanics (FM)are represented as non-linear MBS joints. This approach is possible for transport aircraft, where the flight mechanics can be a 6-degree-of-freedom joint, but also for wind tunnel models with a reduced number of degrees of freedom or for helicopters, where the complex kinematics of the system (e.g. the rotor hub) can be introduced. The elastic members are included in modal form 76 elastic markers wing, MBS joint fixed support / adjustment of angle of attack / heaveand pitch exitation, MBS joint for the definition of wind tunnel co-ordinate system (Co-ordinate system for the exchange of loads / deformation) Fig. 2: MBS model of the AMP wing 22 SIMPACK News September 2010

23 Jürgen Arnold, Wolf Krüger, Gunnar Einarsson, Deutsches Zentrum für Luft- und Raumfahrt, Göttingen Customer Application CFD is performed via the Python program and a socket connection using the SIMPACK IPC co-simulation interface. Fig. 3: Temporal coupling scheme via the FEMBS interface, the equations of motion solved by the MBS tool. The CFDbased aerodynamics are calculated by a dedicated CFD solver and coupled to the MBS system via co-simulation. CFD The behavior of the flow around the wing dz [m] (CSS) MBS Runge-Kutta time stepping scheme is used. For time-accurate computations, an implicit dual-time stepping approach is used. Fig. 1 shows the aerodynamic model of the AMP wing used in the work. The TAU Code modules have been wrapped with Python interfaces, and can thus be used as library functions from within a Python is simulated with the script. To couple TAU TAU Code, a CFD The data transfer is to SIMPACK, the TAU tool developed by the performed via the Python Solver is called from DLR Institute of Aerodynamics and Flow using the SIMPACK IPC The flight mechanic program and a socket connection a coupling script. Technology. The TAU co-simulation interface. data calculated by Code solves the compressible, three-dimensional, time-accurate the Motion module of the TAU Code, which SIMPACK is sent to Reynolds-averaged Navier-Stokes (RANS) builds the required transformation matrices equations using a finite volume formulation. used by the Solver module. The data transfer The TAU Code is based on an unstructured grid approach, capable of using hybrid grids. The TAU Code functionality is organized into modules. The following modules have been used for the process described in this paper: the Preprocessor module, which uses the information from the initial grid to create a dual-mesh; the Solver module, which performs the flow calculations on the dual-mesh and applies guided rigid body motions when specified; the Deformation module, which propagates the deformation of surface coordinates to the surrounding grid; and the Postprocessing module, which is used to convert TAU Code result files to formats readable by popular visualization tools. The Solver module can be executed in Euler mode, or using Navier-Stokes (RANS) equations with 1-Equation or with 2-Equation turbulence modeling. The results shown in this paper are all based on the Euler mode. This is mainly done for reasons of computation time. The coupling procedure as such is identical for RANS calculations. For steady calculations, an explicit multistage Fig. 4: Results of quasi-steady coupling STRUCTURAL MODEL The structural model of the wing has been set up in the FE code NASTRAN and has been subject to a modal analysis. The results have been exported to SIMPACK using the FEMBS Interface. The model used in SIMPACK consists of a wing model and model support represented by 76 markers on the elastic structure and 20 elastic modes. Fig. 2 shows the MBS representation of the structural model including the used reference frames and joints defined for prescribed motion. SPATIAL AND TEMPORAL COUPLING Due to the different discretizations of the CFD and the elastic MBS model, dedicated routines for spatial coupling have to be used. For time-marching simulation, a temporal coupling scheme has to be employed. Spatial coupling of SIMPACK to TAU is realized via the DLR inhouse development PyCSM. The approach makes use of node-based, conservative interpolation methods to map aerodynamic forces between structural and aerodynamic grids, and a non-conservative interpolation to map deformations. CFD and MBS codes exchange their results (forces/deformation) at each simulated time step in co-simulation through a TCP/ IP socket. The communication scheme is No. of Iterations [-] SIMPACK News September

24 Customer Application Jürgen Arnold, Wolf Krüger, Gunnar Einarsson, Deutsches Zentrum für Luft- und Raumfahrt, Göttingen Fig. 5: Comparison of pressure distribution Fig. 6: Wing tip deflection for wing heave motion the so-called 'Conventional Serial Staggered CSS', a first-order scheme, see Fig. 3. This approach allows TAU to run on a high performance computing cluster using highly parallel computation, if required. SIMULATION RESULTS AND COMPARISON The test cases used for comparison are simulations for a Mach number of 0.6, a pressure of 0.9 bar, and an angle of attack of Three different configurations are regarded, first a steady state solution, second a sinusoidal heave oscillation at the model support of f = Hz and an amplitude of z = ± m, and third a pitch oscillation at the same frequency with a pitch of α = ±0.5 around the same point. For the pitch and the heave case, the resulting motion of the wing tip will be a combination of rigid body motion of the excitation plus an elastic structural deformation. STATE OF EQUILIBRIUM The state of equilibrium for the deformed AMP-Wing is computed with two different approaches, both starting from the undeformed wing shape. First, a quasi-steady coupling procedure, taking five coupling steps into account (see Fig. 4), second, a transient simulation for the physical time t of 1.0 s using time-accurate unsteady aerodynamics and 500 co-simulation steps. The resulting static deformation at the wing tip in the z-direction and the constrained force in the z-direction at the support are the same for both approaches with steady and unsteady aerodynamic forces. Values of m and 1853 N are obtained in the wind tunnel coordinate system. To validate the result, pressure distributions of the experiment have been compared to the results obtained with CFD/MBS. Numerical and experimental results correspond very well; see Fig. 5 for data at a location of 69.2 % wing span. HEAVE AND PITCH OSCILLATIONS Two different approaches have been investigated to find the response of the elastic AMP-Wing to the forced heave and pitch excitations at the model support. The first approach represents the rigid body motion due to heave excitation and the elastic wing deformation from aerodynamic loads together in the TAU Deformation module. The second approach uses the TAU Motion module to represent the heave or pitch, and the Deformation module to represent the elastic deformation only. In the latter approach, the Motion module is supplied with twelve flight mechanics (FM) parameters. The FM-parameters are comprised of three angles and angular rates, each in the body coordinate system and three translations and translational velocities each defined in the geodesic coordinate system. They are measured as MBS sensor data and communicated through a Pythonshell to the TAU Motion module. The oscillating deformation at the wing tip in the z-direction and the constrained force in the z-direction at the model support are the same for both approaches. This is true for the heave as well as for the pitch case. Fig. 6 shows the corresponding time histories of the wing tip deflection for the heave case. The red line represents the total deflection; the blue line the purely elastic part of deflection. Pitch oscillation data look very similar and are given in [1]. Unfortunately, no experimental results for direct comparison are currently available for these cases. OUTLOOK The work described has been a test case for the interface of SIMPACK to CFD aerodynamics, i.e. the TAU Code. This coupling is of great interest for aeronautical applications both in the area of fixed wing aircraft and for helicopters. The implemented approach forms the basis of an extensive application of MBS/CFD coupling pursued by the DLR Institute of Aeroelasticity. REFERENCES [1] Arnold J., Einarsson G., Krüger W. R. (2009): "Multibody simulation of an oscillating aeroelastic wing model." NAFEMS International Journal of CFD Case Studies, Volume 8, pp SIMPACK News September 2010

25 Steven Mulski, SIMPACK AG Software Gear Pair Enhancements with SIMPACK Version 8904 Several new functionalities are available with the SIMPACK Gear Pair module in SIMPACK version Not only has the visualization and handling of bevel gears and force arrows been vastly improved but major new functionalities (e.g. tooth profile and flank modifications, easy modelling of non-parallel axes, etc.) have been added. HISTORY Initially developed for Formula 1 high performance engines back in 2003 (by Lutz Mauer, an executive board member of SIMPACK AG), the SIMPACK Gear Pair functionality has since been used in a large variety of industrial sectors, e.g. automotive, wind, rail, shipping, aerospace, concrete mills, material handling, etc. GENERAL In SIMPACK, a large variety of elements are available for the simulation of torque converters. Depending upon the task at hand, elements of various level of detail may be used for achieving the optimum balance between solver speed and accuracy. For example, simple one-dimensional elements may be used for torsional analyses whereas gearbox elements (e.g. planetary gear stage) may be used for more detailed analyses when reaction moments on the housing are required. For simulations where individual tooth contact forces are required, the SIMPACK Gear Pair force element, FE 225, may be used. This element enables the additional analyses of the meshing forces and moments, shaft bending, bearing Fig. 2: Gear box with Gear Pair forces and other resultant forces Fig. 1: Bevel gear with crowning forces, and a host of other pertinent analyses (Fig. 2). The gear pair FE 225 is an analytical element, and therefore, extremely fast simulation times can be achieved. Graphical primitives are defined for the gear wheels which are subsequently used for the force calculations. This results in accurate animation of the gear tooth contacts and play. The element includes the following functionality [1, 2]: Involute spur, helical and bevel gears Internal and external gears Profile Shift Backlash and friction Single and multiple tooth contact (internal excitations due to tooth meshing) Dynamically changing gear pair center distance and backlash (particularly important for floating suns (Fig. 3) and elastically mounted shafts) The major gear pair enhancements with version 8904 are: Rack and pinion gearing Bevel gear primitive Tooth modification Flank modification Shuttling forces Easy slicing for non-parallel gear wheels and gear wheels with flank modification Easy handling of output values and animation of contact forces GEAR PAIR PRIMITIVES MAJOR ENHANCEMENTS With bevel gears, a new parameter, the Rim thickness, has been added for a more realistic graphical representation (Fig. 4). For all gear pair types, tooth and flank modification has been added. The modifications are primarily used for smoothing the non-linear internal excitations due to the continually changing number of teeth in contact. The following modification types have been added: Tip (Fig. 5) Root Circular Left and Right Side Lead Crowning (Fig. 6) Lead Angular Bias (Twist) Input Function Array All modification types can be input for the right and left flanks or for both together. Fig. 3: Motion of floating sun within a planetary stage ( IMM, TU Dresden) SIMPACK News September

26 Software Steven Mulski, SIMPACK AG Fig. 4: Bevel gear primitives GEAR PAIR FORCE ELEMENT MAJOR ENHANCEMENTS For simulating gear pairs with non-parallel axes, slicing of the gear wheels is necessary [3]. Previous to version 8904, extra gear wheel primitives and force elements had to be used for this purpose. This functionality is now achieved by setting single parameter (i.e. Number of slices ) within the gear pair force element. The handling of the offset angles for helical gears is now fully automatic. Slicing is also necessary if flank modification is used. Shuttling forces, i.e. the axial displacement of the contact forces, has now been implemented. In the case of helical gears, this will result in an additional tilt moment. The graphical representation of rack gears has been available for a long time now. With Fig. 5: Tip profile modification version 8904, the calculation of the rack and pinion forces has also been implemented (Fig. 7). With version 8904, the user can now easily switch on and off, and choose between, various output value types. This enables easier handling and a more efficient use of data storage space. The different types of output are described below. GEAR PAIR DATA CHECK In order to check the input parameters and initial conditions of the gear pairs within a model, a user can perform a Test Call. This will result in a list being generated for each gear pair consisting of important input parameters and calculated data. Information such as the theoretical center distance, radial offset, axial offset, transverse contact ratio, overlap ratio, and total contact ratio will now be readily available. GEAR PAIR OUTPUT VALUES By way of parameterization, a user can choose for which gear pairs the Basic Output Values will be generated. These values include such data as the relative Fig. 6: Crowning, left and right flank angles and angular velocities, total normal contact stiffness and the dynamic transmission error. Similarly, a user can also choose which Advanced Output Values are to be saved (Fig. 8). These values are primarily used for analyzing the coupling forces of the gear pairs, either for the sum of all teeth in contact or the individual tooth-pair contacts. In addition the Advanced Output Values enable easy animation of the force arrows in the PostProcessor (Fig. 9). After an integration run is complete a user can subsequently choose which output values to generate. Re-running the time integration is not necessary. Only reperforming measurements is required. CONCLUSION SIMPACK version 8904 represents a major milestone in the development of the gear Fig. 7: Rack and pinion gear Fig. 8: User choice for advanced output values pair module. New functionalities such as tooth and flank modification and automatic slicing and force arrow visualization, enable not only easier and faster model generation but also improved accuracy and quicker analyses. Although a significant development step has been achieved, the demanding and varied applications of the gear pair element will continue to result in further, more advanced requirements. The development of the gear pair element does not have one static functionality goal, after which the development can be seen as being completed, but rather a continually Fig. 9: Animation arrows of normal loads Indiv. load (fl_n) i,k advancing goal to which subsequent further SIMPACK development will ensure that the gear pair element can accompany industrial users long into the future. REFERENCES [1] L. Mauer, GearWheels in SIMPACK, SIMPACK News, July 2004 [2] L. Mauer, Modeling and Simulation of Drive Line Gears, SIMPACK News, July 2005 [3] E. Pfleger, Simulation of the Dynamic Behaviour of Nose Suspension Drives for Rail Vehicles Using SIMPACK-GearWheel, SIMPACK User Meeting 2006 All articles and presentations available at 26 SIMPACK News September 2010

27 Stefan Dietz, SIMPACK AG; Wolfgang Erhardt, SAFE-FEM GmbH Software Fig. 1: Normal stress in the x-direction (axial) of the crank shaft. Fatigue Analyses with FAT4FEM in SIMPACK's PostProcessor Since the release of SIMPACK version 8900 in 2007, the computation and graphical representation of component stresses has been available in the SIMPACK PostProcessor. With the forthcoming SIMPACK version 8904 this feature is expanded with the ability to perform fatigue analyses. These may be configured and started in the PostProcessor. After the computation, the results and safety factors may be displayed in the PostProcessor as a contour plot. The fatigue analyses are performed by FAT4FEM which is an easy-to-use and easy-to-learn fatigue tool that is based on the critical plane approach. THE EXAMPLE MODEL The FEM model of the crankshaft, which was generated for the stress calculation, was created in NX NASTRAN. The model has approximately 3 million degrees-of-freedom that, in the interest of computational time, must be reduced to a lower number of degrees-of-freedom for a multi-body simu- Fig. 2: FAT4FEM GUI for defining material properties to be used in fatigue analysis SIMPACK News September

28 Software Stefan Dietz, SIMPACK AG; Wolfgang Erhardt, SAFE-FEM GmbH relief modes and eigenmodes can accurately represent the stresses due to static and dynamic loading conditions. When simulating an engine running at constant speed (rpm), forces, modal coordinates, etc. are issued for approx. 720 communication points in time. With this modal superposition, results of static and time-dependent simulations can be efficiently and accurately reproduced in SIMPACK. Fig. 3: FAT4FEM GUI for defining failure criteria for fatigue analysis lation. The crankshaft in SIMPACK is defined by approx. 50 modal degrees of freedom and does not deviate by more than 1 percent from a static solution in NX NASTRAN. The reduced flexible crankshaft is coupled into the multi-body system at the main and large end bearings with the assistance of force elements in order to incorporate the stiffness and attenuation properties of the bearings. The connections of the crankshaft with the flywheel and the torsion vibration damper are modeled with the help of joints and/or constraints (i.e. rigidly connected). In the multi-body model, the dynamics of the conrods, pistons and the engine with the engine mounts are all taken into consideration. Gas forces are applied as external loads. Consideration of the hydrodynamics in the main bearings and large end bearings, as well as the flexibility of the engine block, is also possible; however, the corresponding influences are not illustrated in this model. STRESS CALCULATION For modal superposition of stress, unit stress vectors are needed in SIMPACK. These are the stresses of the eigen vectors as well as stresses of so-called "inertia relief modes". The eigen vectors are used to show stress as a result of free vibrations. In this model, only 40 eigen vectors are needed to illustrate vibration behavior in the relevant frequency range with sufficient accuracy. The stresses, caused by coupling forces in the bearings, are represented by "inertia relief modes". An inertia relief mode is the result of a static FAT4FEM is an easy-to-use and easy-to-learn fatigue tool that is based on the critical plane approach. calculation in the FEM program. Here, the induced forces are compensated by inertia forces from the rigid body motion, i.e. the external load and the inertia forces form an equilibrium of forces and moments. Consistent with the mathematical formulation of linear elastic bodies, deformations are not considered in this equilibrium of forces. A unit load (force and/ or moment) is a load in the direction of the coordinate axes that is scaled to unity. Only the relevant unit forces were considered for the bearings. These are the forces in x- (longitudinal direction) and y- and z-directions (radial) at the axial main bearing and at the interface points to the connecting rods. Forces in the y- and z-directions were considered at the main bearings, which do not transmit axial forces, and the forces and moments in all directions at the joints to the flywheel and the torsion vibration damper. In total, 35 inertia relief modes have been used in this case. This corresponds to a total of 75 unit stress vectors. The modal coordinates of the eigen vectors considered in the SIMPACK model, as well as the forces in the coupling elements of the flexible crankshaft, are the result of a multi-body simulation that is required for superimposing the unit stress vectors. The stresses of the eigen vectors are scaled with their respective modal coordinates and the stresses of inertia relief modes with the corresponding forces. After the scaling of the unit stress vectors, their superposition and representation in the PostProcessor follows; see Fig. 1. The combination of inertia LIFESPAN ESTIMATION WITH FAT4FEM Lifespan estimation is carried out with FAT4FEM ("Fatigue for FEM") available from Safe-FEM GmbH, This software has interfaces to Abaqus/CAE, ANSYS and FEMAP. FAT4FEM is integrated in to SIMPACK and can started from the PostProcessor. FAT4FEM is based on the critical plane approach. The critical plane approach looks for every comparison of two load cases and for the level at a node which yields the most damaging equivalent stress. This technology is suited particularly (but not solely) to problems concerning rotating principle stress directions (which is the case here). The critical plane approach is extremely computation-intensive and very precise in principle. Still, we can only speak of a fatigue estimation. Keep in mind that with today's technology, computer-aided fatigue prediction can always only be a fatigue estimation and cannot give absolute data, e.g. maximum number of years of lifetime, etc. Nevertheless, computer-aided fatigue prediction is a valuable tool for the engineer to quickly compare variants and to identify "weak points" of a mechanical design during the design process. During the development of FAT4FEM, primary focus was given to the user-friendliness of the graphical user interface. Operating the user interface of FAT4FEM is something you can learn in a matter of hours, which makes the program very suitable for non-experts, see Fig. 2 and 3. LIFESPAN ESTIMATION THE MOST IMPORTANT ISSUES Roughness effect: The crankshaft is made of cast iron (GGG-60). Therefore, roughness effect does not necessarily apply because GGG-60 is a material with internal notches, and surface is smoothed at critical places so no surface effect is taken into account. Temperature effect: At oil temperatures up to 140 C, the temperature effect must not be considered for GGG-60. The following factors are of importance for the correct assessment of the crankshaft with regard to lifespan: 28 SIMPACK News September 2010

29 Stefan Dietz, SIMPACK AG; Wolfgang Erhardt, SAFE-FEM GmbH Software 1. Correctly determined stresses via MBS and FEM It is assumed that the stresses calculated with SIMPACK and NX Nastran are sufficiently accurate. 2. Correctly defined S/N curve The material properties for GGG-60 were taken from the publicly accessible website of the TU Darmstadt. These properties are based on data pools of Ch. Boller and T. Seeger and are publicly accessible on the internet: 3. Suitable equivalent stress hypothesis GGG-60 has a breaking strain A5 of 7 %. This means that the equivalent stress hypothesis according to Van Mises can be applied. 4. Mean stress correction Mean stress sensitivity is approx. 0.3 %, which indicates a ductile material. 5. Size effect This takes the size of the component into account. The size effect was not considered in the present study. Therefore, we get more conservative values for safety. With comparative calculations, as is also customary for crankshafts, it is, of course, necessary to always use a specific method, either with or without notch sensitivity. ILLUSTRATION OF THE RESULTS After the lifespan estimation, the results are automatically uploaded to the SIMPACK PostProcessor by FAT4FEM. Fig. 4 shows the mean stresses and Fig. 5 the factor of safety for the examined rpm. SUMMARY The SIMPACK FAT4FEM functionality described here is available as a licensed module in addition to the SIMPACK Stress module with version The seamless integration of MBS, FEM and Fatigue comes with clear time and cost advantages for the experienced user. The process Operating the user described here is available interface of FAT4FEM is for automated calculations something one can learn in batch operations. The in a couple of hours. fatigue estimation only causes minimal extra time effort in the overall process chain. Defining the FAT4FEM project takes just a few minutes. Regarding CPU times, the calculation of FAT4FEM for the areas shown in Fig. 4 and 5 and for 360 load steps is about one hour. If only every third load step is activated, one gets nearly the same results for a fatigue estimation, but evaluation time amounts to only about six minutes. This means that lifespan estimation for dynamically stressed components, even for a great number of operating modes, is possible in an extremely fast and simple manner. Fig. 4: FAT4FEM results in the SIMPACK PostProcessor Equivalent stress amplitude (red corresponds to large stresses) Fig. 5: FAT4FEM results in the SIMPACK PostProcessor Safety factor (green indicates critical spots) SIMPACK News September

30 Software Wolfgang Trautenberg, SIMPACK AG 200 New Features and Improvements introduced with SIMPACK Version 8904 wheel shoulder, the chain plate can have contact with these additional rings. Fig. 1: Belt drive DYNAMIC BUSHING PARAMETER FITTING AND NEW OPERATING MODE A new operating mode was added to the dynamic bushing force element to enable a hysteresis computation that provides for a memory effect typically seen in bushing measurements. Finding the right set of parameters for the dynamic bushing was made much simpler and quicker by introducing a parameter fitting utility. This utility allows the user to tune the force element parameters interactively via sliders while a plot shows the comparison of measured and computed data. The SIMPACK development team is strongly focused on creating SIMPACK 9000, the next major SIMPACK release due in Nevertheless, a lot of attention and development effort was put into SIMPACK 8904, the latest SIMPACK version, released in September Almost 200 new features and improvements have been added to SIMPACK These include brand new SIMPACK modules such as the belt drive module or the Bio-Mechanics module, as well as major new features for existing modules, such as the newly added Gear Pair Force Element capabilities. The most important of these new features and improvements are summarized in the following sections. For in-depth information on all changes, please refer to the release notes of SIMPACK 8904 which are available on the SIMPACK Download Server. GEARWHEEL SIMULATIONS GOING 3D The fast and accurate SIMPACK Gear Pair Force Element was greatly enhanced in numerous ways. Functionalities such as automatic slicing and effects like force shuttling can now be used to include 3D effects into the gearwheel simulation. Additionally, different methods were added for taking into account profile and flank modifications of the gears. The list of supported geometry pairings was expanded to facilitate the simulation of rack and pinion pairs. To enable easier postprocessing and greater insight, the generated outputs were completely reworked, including drastically improved 3D visualizations of the forces acting in a gear pair. Another important new capability is the data check facility that lets the user see all important gear pairing data in a central location. For more details, see the separate article on gearwheel simulation in this edition of the SIMPACK news (see page 25). BRAND NEW BELT DRIVE MODULE The list of the major drivetrain coupling elements of gearwheels and chains was complemented by adding the Belt Drive module (Fig. 1). Starting with SIMPACK 8904, detailed and fast belt drive simulations can be performed via a modal belt description. A belt drive is defined by the pulleys, the optional tensioner system and the belt force element that connect the former elements to a belt drive. The belt itself is defined via its geometry and material properties. From these inputs, SIMPACK automatically computes a modal description of the belt. Different friction models are available. CHAIN DRIVE ENHANCEMENTS The Chain Drive module, the third of the three drivetrain coupling modules, was enhanced with the capability to include an additional centrifugal ring when using smart chains (Fig. 2). Now, in addition to the contact of the chain plate with the chain Fig. 2: Centrifugal contact ring for chains TIRE NEWS CDTire (Fig. 3), a tire model for comfort and durability applications, owned and now developed by the Fraunhofer Institute for Durability and System Reliability LBF, is now available as one of the tire modules shipped with SIMPACK. CDTire can automatically adapt to changing road situations by switching between different internal representations. These include detailed models for large deformations and allow for tire rim contact. In addition, the CDTire implementation in SIMPACK can take full advantage of multi core and multi CPU computers by utilizing one CPU core per tire if requested by the user. The DELFT Tire module integrated in SIMPACK is now version In addition 30 SIMPACK News September 2010

31 Wolfgang Trautenberg, SIMPACK AG Software Fig. 3: CDTire to many improvements, support for a new road type was added with this DELFT Tire version. The curved regular grid road (or CRG road), which describes the road via a grid of measured road heights lateral to a spine, can now be used with this tire model. NEW RAIL NEWS The New Rail module was expanded to enable simulations of untrue (out-of-round) wheels. In addition, the simulation of large yaw angles was reworked and significantly improved leading to greater accuracy and robustness. MORE SIMULATION POWER FOR WIND ENERGY In addition to many new features offered for drivetrain simulations which are also applicable for wind turbines, more wind specific functionality was added to this SIMPACK release. This includes a SIMPACK native integration of NREL s aerodynamic library AeroDyn for simulating the wind forces acting on rotorblades of horizontal axis wind turbines. For simulating large non-linear deformations in rotorblades an option was added to SIMPACK s rotorblade generator to automatically split the rotorblade into different flexible bodies. Fig. 4: Biomechanics BIOMECHANICS POWERED BY BIOMOTION SOLUTIONS For analyzing biomechanic systems such as a human being interacting with a power drill, a seat restraint system or a motorcycle, a completely new module was added to SIMPACK (Fig. 4). This module consists of special force and control elements for modelling muscles, tendons and their respective controllers, spinal discs and so called wobbling masses. These elements are used in full or partial models of human beings with an adaptable level of detail. The models can be generated by a model generator that works off a database of typical human beings. The Biomechanics solution is contributed by Biomotion Solutions ( a company that specializes in measuring and simulating the biomechanics of human beings. TIME EXCITATIONS SPED UP The time excitation types 15 - for importing measured data on either position, velocity or acceleration level and 20 for importing a speed profile given over position were significantly sped up and reworked to offer more options on dealing with the measured data. In addition, the time excitation type 17 for importing velocity dependent periodic signals was expanded to enable the definition of the Fourier Series via Re and Im as well as magnitude and phase. 3D PRIMITIVES New primitives were added for displaying belt pulleys and the belt path. The gear wheel primitive was reworked to provide for a much more realistic display of bevel gears, and cylinder primitives can now have a hole. FLEXIBLE BODIES AND FATIGUE The flexible body interface FEMBS now supports cdb files of ANSYS 12 models for the graphical representation of flexible bodies. Also, the solver performance for flexible bodies with a large number of modes has been greatly improved. A solution for performing fatigue analysis was added directly into SIMPACK. This solution is based on the FAT4FEM technology (Fig. 5) developed by SafeFEM GmbH ( Please see the separate article about this module in this SIMPACK news edition (see page 27). STRESS AND STRAIN COMPUTATION SIMPACK 8904 now offers the possibility to compute and display strains for selected flex body nodes. The strains can be plotted as standard 2D curves and can be exported in various formats. The stress display capability of SIMPACK has been greatly sped up for ABAQUS and improved so that much larger FE-result files can now be processed. CODE EXPORT The Code Export module was reworked to provide a simpler and cleaner interface by supplying the runtime library as convenient DLL. Also, the Code Export licensing is now completely based on OLicense. Fig. 5: FAT4FEM POSTPROCESSOR A new spectrum filter was added to the PostProcessor, providing many interesting features such as overlapping averaging windows. The integration and differentiation filters were both redesigned and extended and can now also handle signals with nonequidistant x-axis values. Displaying and scaling 3D force arrows was improved by adding a smarter scaling algorithm and more scaling options. New commands were added to the scripting engine. The script execution now only requires a license if PostProcessing functionality is triggered by the script. In addition, the size of curve markers is now user definable. MATLAB AND AMESIM INTERFACES The MATLAB interfaces SIMAT and MatSIM now support MATLAB version 2009b and 2010a. The AMEsim interface SIMAS now supports AMEsim version 8 and 9. PLATFORM NEWS Windows 7 is now a fully supported and certified SIMPACK platform. Official support for RedHat Enterprise 5.3 was also added with this SIMPACK version. On 64 bit operating systems, SIMPACK 8904 supports up to 4 GB of memory even with 32bit binaries. This is useful for simulation tasks that require huge amounts of memory, e.g. due to large result files that should be generated. It is therefore highly recommended to always install SIMPACK on machines with a 64bit operating system. SIMPACK News September

32 SIMPACK SIMPACK News News Inside this issue 1. Simulation of the Dynamic Behavior of Aircraft Landing Gear Systems 2. Electronic Stability Program (ESP) for Trucks on the Daimler Driving Simulator 3. Simulating Tank Vehicles with Sloshing Liquid Load 4. Development of a SIMPACK User Routine for Dynamic Light Rail Vehicle Gauging Simulations 5. Determining Reliable Load Assumptions in Wind Turbines using SIMPACK 6. Simulation of Drivetrains on Wind Turbines within the Framework of Certification with SIMPACK 7. SIMPACK Academy Events in Coupling of MBS and CFD: an Oscillating Aeroelastic Wing Model 9. Gear Pair Enhancements with SIMPACK Version Fatigue Analyses with FAT4FEM in SIMPACK's PostProcessor New Features and Improvements introduced with SIMPACK Version 8904 Contacts Germany WORLDWIDE HEADQUARTERs SIMPACK AG Friedrichshafener Strasse Gilching Germany Phone: +49 (0) Fax: +49 (0) info@simpack.de USA SIMPACK US Inc Commerce Center Drive Plymouth Michigan 48170, USA Phone: Fax: Mobile: RobertSolomon@SIMPACK-US.com France SIMPACK France S.A.S. Immeuble "Le President", 4eme étage 40, Avenue Georges Pompidou Lyon, France Phone: +33 (0) Mobile: +33 (0) info@simpack.fr Great Britain SIMPACK UK Ltd. The Whittle Estate Cambridge Road Whetstone Leicester LE8 6LH UK Phone: +44 (0) Fax: +44 (0) Mobile: +44 (0) info@simpack.co.uk Japan SIMPACK Japan K.K. 5F Okubo Bldg Yotsuya Shinjuku-ku Tokyo Japan Phone: +81 (0) Fax: +81 (0) info@simpack.jp China Global Engineering Technology Group (GET Group) Tri-Tower, C Building, Room 1802, 1804 Zhongguancun East Road 66 Haidian District Beijing , P.R. China Phone: +86 (0) Tom.Fu@get-technologys.com Brazil VirtualCAE Serviços de Sistemas Ltda. Rua Tiradentes, 160 Sala 22 São Caetano do Sul São Paulo CEP: , Brasil Phone: +55 (0) Mobile: +55 (0) Leandro@virtualcae.com.br india ProSIM R&D Center #21/B, 9th Main, Shankarnagar Mahalakshmipuram Bangalore , Karnataka, India Phone: +91 (0) / Fax: +91 (0) / info@pro-sim.com / Shama@pro-sim.com SIMPACK News Impressum: Publication period: Editorial, Design & Layout: Steven Mulski, Nicole Blum SIMPACK AG SIMPACK AG Friedrichshafener Strasse Gilching, Germany Phone: +49 (0) Fax: +49 (0) info@simpack.de Circulation: All previous SIMPACK News articles can be found on downloads, newsletters/articles If you would like to sign up for free delivery of the SIMPACK News please visit: downloads, newsletters/articles, subscription