The knowledge of the loads on a component is a basic prerequisite for stress related design. In many cases, the multibody simulation (MBS), sometimes in combination with the finite elements method (FEM), is the only reasonably applicable method in order to identify the component loading.

Depending on the application, the excitation of the system can be defined for example by setting certain motions or based on real car measurements to derive the unknown quantities to be calculated.

The sufficient modeling of force elements and their properties is important, e.g. for air springs which are playing a major role in the field of commercial vehicles.

Simulating large mechanical systems like test rigs or complete cars, the challenge is not only to model the interaction of many moving parts correctly, but also the behavior of complex combinations of elements like sensors, actuators together with controls and hydraulics.

In real live limits are given due to time needed, the available hardware resources and especially the challenge to find correct parameters. A key task is therefore to reduce model complexity by keeping the physically "correct" behavior of individual components with the aim of not changing the overall prediction quality of the whole model too much.

In the context of the widely used multibody simulation (MBS-modeling) for example the elastic behavior of most components is neglected and a rigid model with inert mass, ideal joints and force elements is build. In some cases resulting errors can be compensated by a skilful choice of model parameters and nonlinear characteristics for friction and compliant bearings. Using the technique of modal MBS-FEM coupling, the non rigid behavior of flexible key components can be taken into account in the overall model.

Complex force or control elements often contain additional fractions beyond the pure mechanical domain (hydraulics, pneumatics, electronics ...). Special software tools for the modular or one dimensional system simulation not just allow their efficient implementation with a high degree of flexibility for the level of detail of the resulting sub-model, but also support the integration into the overall mechatronic model (Cosimulation).

Quantitative modeling requires synchronizing with direct or indirect measurements at system and component level (parameterization of the model). Generally speaking, therefore, the prediction quality of the simulation results not only depends on the details of modeling, but also on the quality of input data (method of measurement, sampling, filtering ...).

Projects

• Multibody simulation of a field crop chopper
• System Level Durability - integrated multibody and durability simulation
• Load data analysis and simulation for excavator development
• Determination of section forces based on strain measurements
• Detailed modeling of an air spring

The modeling of servo hydraulic test rigs, also including the complete controls and hydraulic system, forms the basis for projects with the following objectives... Further information ... The modeling of servo hydraulic test rigs, also including the complete controls and hydraulic system, forms the basis for projects with the following objectives:

• conception and design in the development of the entire test rig,
• checking the suitability of existing test rigs for new test scenarios/items,
• studies on effects through modification of test parameters,
• determination of stresses in the test item during testing and
• pre-iteration of drive signals to save time on the real test rig.

Projects

• Multibody simulation of a suspension test rig including controls and hydraulics
• Multibody simulation of a full vehicle test rig
• Simulations for improving the design of a truck trailer test rig

The development in the field of real-time simulation on ITWM pursue two main objectives:

• on-board calculation of durability related, but difficult to measure values, from easy to measure data that are already available and
• real-time simulation of the environment (e.g. excavator & ground) for the integration of humans as part of the system.

The on-board load calculation of durability related quantities is closely related to customer usage profiles and targets. The knowledge of operational loads allows comparing component loads on duty with loads assumed designing a test scenario on the test track or testing rig. On the other hand, with a broad knowledge of actual stresses in the field, the design targets (for test track, test and simulation) are better adapted to the reality.

Therefore real-time models are developed in a convenient prototyping environment (dSpace ds1006/MicroAutobox, Matlab/Simulink, Simpack etc.), as well as they are directly implemented on on-board microcontrollers (e.g. AT91SAM7LA2 Atmel, Infineon TriBoard TC 1796). As a test environment (next to the vehicles of the participating OEMs) ITWM uses a realistically equipped model truck as development platform (scale 1:10, wire CAN bus, motion and acceleration sensors, etc.).

In many cases a human is a part of the control system and his response directly depends on the current system behavior. Examples include the earth moving activities or the unloading of material. To include a real person into the system real-time models are absolutely necessary to provide the essential and realistic reactions of the system to the people. Working in this environment includes:

• the creation of real-time system models,
• the simulation of the behavior of soils or the interaction of the system with other materials (e.g. boulders) and
• the development of a VAR system in order to give realistic feedback concerning motion and scene view to the people acting.

Projects

• Design of a mechatronic development platform (truck model)
• Design and construction of a VAR system

The extreme usage variability and versatility for commercial machinery in construction and agricultural applications leads to many restrictions and demands for mechatronical systems. In the early stages of product design, knowledge about realistic load profiles for a work machine increases the efficiency of the overall design process by reducing the technical risk in developing new systems and variants.

The solution developed and implemented at ITWM is to include the human operator in the simulation, such that the product, e.g. a commercial vehicle, can be experienced virtually. Based on this principle, a novel driving simulator is developed and built at ITWM. In comparison to the widespread simulators based on a 6-axis parallel kinematic Steward Gough platform the ITWM driving simulator has much more clearance to combine translational and orientation tasks effectively.

In addition to motion feedback covering cybernetic requirements, visual, vibrational and acoustic feedback are available. The realistic layout and feel of the simulator cabin supports the immersive character of the simulation. Simulation data and operator actions can be saved and stored for subsequent detailed and computationally intensive simulations.

In the future, the soil-tool interaction models developed at ITWM will be adapted to the simulator and will increase the realistic impression further.

Contact

• 

  Dipl.-Ing. FH Oliver Hermanns

  • Phone +49 631 31600-4362

  • Send an e-mail
  • more Info

top of page

For the computer simulation of dynamical systems, a mathematical model of the real system is needed as well as an input- or excitation-model that describes properly the interface between the considered dynamical system and its environment.

In the context of vehicle engineering, a typical excitation-model is a tire-model together with a road profile. Another possibility is to excite a vehicle model with measured forces and torques that act on the vehicle's spindles.

But there is a crucial difference between the two mentioned approaches: on the one hand, tire-model and road profile constitute an invariant excitation-model, i.e. the input-model does not depend on the dynamical system, the vehicle, but it can be used to simulate different vehicle variants. On the other hand, however, measured wheel forces and torques highly depend on the vehicle that was used for the measurement; they are not invariant, they can only be used to excite a computer model of the measurement vehicle.

Obviously, for reasons of efficiency, invariant excitation-models are very desirable, measurements and experiments with many vehicle variants are costly and time-consuming. Unfortunately, the invariant counterpart, i.e. tire-model and digital road profile, is hard to obtain and often not available.

In this context, the so-called I6D-method has been developed over the last years and applied for a patent at the Fraunhofer ITWM. On the basis of measured wheel forces and torques (not invariant, but comparably easy to obtain) at one (!) reference vehicle, a 6D virtual road profile is identified. To this end, a 6D tire-surrogate model is introduced using a specific subsystem approach that leads to a mathematical control problem. The latter can be solved numerically by appropriate algorithms very efficiently and without any iteration. The resulting excitation-model is defined by the pair that consists of the tire-surrogate model and the computed virtual road profile. This pair possesses specific invariance properties and can be used to simulate other vehicle variants.

The I6D-method, primarily developed to compute invariant excitation-models for full-vehicle simulation, can also be applied to compute appropriate invariant excitation models for other dynamical systems.

Contact

• Dr. Michael Burger

  • Phone +49 631 31600-4414

  • Send an e-mail
  • more Info

top of page