Introduction
Education and Research at the ika
Development of a semitrailer truck simulation-model
Introduction
ADAMS has been used at the 'Institut für Kraftfahrwesen Aachen' of the Aachen University of Technology for several years. Various research-projects have increased the experiences with this simulation programme. From the beginning of the next term on we will also enable students to become familiar with ADAMS more intensively and we will start educating the students in working with this programme.
That's why the opportunity is taken within today's presentation to introduce the institute and its development- and education-programme. In the second part, some development-examples taken from the research programme will be shown by presenting a project which was realized in cooperation with
Continental AG.
Education and Research at the ika
This summer the 'Institut für Kraftfahrwesen Aachen' moved into new buildings in Aachen-Melaten. Thus, the employees of the institute find ideal work conditions concerning education and research.
Fig. 1: 'Institut für Kraftfahrwesen Aachen' - New buildings
The possibilities in research and development are of great variety at the institute and they cover the following fields:
• Chassis
• Body
• Drive-train
• Electronics
• Acoustics.
• and Traffic
Besides simulation-accounts at the computer, numerous experiments can be carried out. Moreover, the institute holds its own test track which can be used for driving tests in different conditions.
Fig. 2: 'Institut für Kraftfahrwesen Aachen': Elements of Lecture and Research
The design of the new buildings clearly demontrates the special relation towards education which led to the installation of a lecture hall, a PC-laboratory and a laboratory for psycho-acoustics. Offices and drawing rooms for students and employees, a technical library and CAD-facilities complete the installations refering to lecture and practice.
Fig. 3: ika teaching offer with external lectureships
At the Aachen University of Technology the 'Institut für Kraftfahrwesen Aachen' is responsible for the sector "Automotive Engineering", which is one of almost 20 extention subjects students at the faculty of mechanical engineering can choose. After having passed their pre-examination, the students are prepared for their profession as an Automotive Engineer through diverse lectures.
The combination of research and education at the Institute of Automotive Engineering can be found once more with regard to the use of the CAE-Tool ADAMS.
The assistants of the institute have been working with ADAMS for a longer period. Because of frequent requests by the automotive industry and also because of the own demand for engineers with experience in working with ADAMS, the education in using this simulation-programme is included explicitly in the ika teaching offer.
In principle, ADAMS is lectured in two presentations:
• the lecture "Computer Technology in Automotive Engineering"
• the PC-Laboratory "Introduction in the Multibody System Analysis ADAMS"
The lecture "Computer Technology in Automotive Engineering" is one of the compulsory subjects of any student of automotive engineering in 7th term. The essence of this lecture is demonstrated in Fig. 4.
Fig. 4: Essence of the lecture "Computer Technology in Automotive Engineering"
In the course of a general introduction to simulation techniques, this lecture also covers the multibody system analysis ADAMS in a separate presentation. The essential components of this presentation are:
• the history of multibody system analysis programmes and the creation of ADAMS
• description of the programme
- creation of models
- tire model
- steering manouevres and power transmission
- course of calculations
- possibilities for the presentation of results
• further elements of ADAMS
The important point of this lecture isn't the education in programming ADAMS but the transmission of the actual working method of the programme and the publishment of fundamental elements.
Fig. 5: PC-Laboratory at the 'Institut für Kraftfahrwesen Aachen
Fig. 6: Testequipment of the ika supporting vehicle-dynamics-simulation
Within the second lecture-presentation, which is the PC-Laboratory "Introduction in the Multibody System Analysis ADAMS", the student is able to become familiar with the programme. He himself carries out exercises at the computer with ADAMS. For this purpose, 12 Penthium-PC with 32 MB-Ram were donated by the companies "Kober" and "Lemförder Metallwaren" and are equipped with ADAMS-PC-versions.
Some students get the possibility to work as a "Student Assistant" beyond this laboratory and to assist directly in industrial projects. Just because of the equipment of the institute with test-benches, equipment for driving tests and computers for ADAMS, a complete examination of this subject matter is possible.
By the help of diverse equipments there is the possibility for the differents fields of vehicle-dynamic-examination:
• Data determination
- Tyre test bench for measurement of longitudinal and lateral forces as well as rolling resistance and noise
- Test bench for investigations on axle-kinematics and elasto-kinematics
- Test bench for definition of centre of gravity and moment of inertia of vehicle-parts and of total vehicles.
• Creation of models on diverse computers (workstation and PC)
• Validation by the help of the results from driving tests at the own test
An example which is typical of a research-project that was realized at the Institute of Automotive Engineering in close cooperation with the industry is the "Development of a semitrailer truck simulation-model". This project was executed in cooperation with
Continental AG and is presented in the following part.
Development of a semitrailer truck simulation-model
This part of the paper deals with the cooperation of the institute of automotive engineering with industrial partners. With the
Continental AG / Hannover we worked together at a 40 t semitrailer-truck model. Because of earlier experiences with truck models the ika built and validated the model for different load conditions and ride-maneuvers.
The measurements of the truck-geometry was done by
Continental. Some special quantities could be taken from earlier models we have built.
The truck runs with a Pacejka tire-model, that was built by
Continental. Also the measurements of the tire and the curve-fitting for the Pacejka-parameters were done by
Continental. Coupling with the truck-model is made by the TIRE-statement and the tirsub-subroutine, that calls some other user-written-subroutines, e.g. for initialisation.
In order to connect tire-model and vehicle-model with the right signs and to investigate the tire-model, a tire-tester was used. In this model, the tire is coupled with a part, that is lead in translational x-direction. After loading the system, with stationary or quasistationary rotation of the tire about the slip-axle or the camber-axle, the behaviour of the tire can be investigated.
Fig. 7: Tire-tester
The truck-model consists of a tractor that is devided in two body parts and a semi-trailer, that was first build as a one-body-part and then devided in up to five parts to investigate the according influence. The complete model has 110 degrees of freedom for the two-part semitrailer
The tractor is devided into a front-part, that summarizes driver-cabine, drive-unit and front-axle-mountings as well as a rear-part with rear-axle-mountings and couple-unit. Front and rear body of the tractor are connected by a revolute-joint so that a torsional displacement is possible. The torsional stiffness is modelled by a rotational-stiffness parallel to the revolute-joint. The torsional stiffnes of the tractor-frame was measured by the vehicle-manufacturer.
The front-axle leaf-spring was modelled by a translational stiffnes in front and behind the axle at each side. The front part of the leaf-spring, that takes the guide function of the axle, was replaced by a stiff part with the same length and the same connecting-points at frame and axle like the leaf-spring. It is coupled with revolute joints to the frame at both sides and coupled to the axle with a spherical at one side and an inline-primitive at the other side. The anti-roll-bar for the front-axle is modelled by five beam-statements.
Fig. 8: Tractor-model
The rear-axle is supported by two air-springs for each side. This air-springs were modelled as stiffnesses in the working point and they are different for the unloaded and the loaded truck-model. The guiding of the rear-axle is done by two trailing-arms that are mounted deeper than the tire-center and have bushings for the frame-connection and sphericals for the axle-connection. A 'V-formed' arm takes the lateral control for the axle in reality. In the model a kinematic equal ability can be reached by a trailing-arm, that is connected to the frame with a revolute-joint. At the rear-axle there is also an anti-roll-bar, that consists of five beams.
Coupling of tractor and semi-trailer is realized by a universal-joint, that allows the relative motions 'yaw' and 'pitch' between the two vehicles.
The three-axle semitrailer consists of a body, that has torsional stiffnesses realized like those of the tractor. Different bodies were investigated such as total stiff bodies up to bodies with four torsional degrees of freedom. The semitrailer with four degrees of freedom has revolute joints between the king-pin and a middle-part, the middle part and the first axle and between first-second and second-third axle.
The semitrailer air-springs are guided by trailing-arms, that are formed of spring-steel and act as anti-roll-bars for the semitrailer. In the model, this drivers are build by beam-statements with revolute-joint-connection to the frame and fixed-joint-connection to the axle.
Fig. 9: Roll-movement of the semitrailor-axles
To use the model in a modular way, that means in this case to change the wheel-base simply, the rear axle of the tractor and the axles of the semitrailer aren't connected direct to the frame but connected to a subframe. Springs, shock-absorbers and joints work between axle and subframe. Subframe and mainframe are connected by fixed joints. All part-coordinate-systems of parts belonging to one axle have their origin in the x-z-coordinates of the tire-center. The Marker-coordinates of the parts belonging to this axle are always related to the tire-center. To change the wheelbase, only the part-coordinate-systems have to be changed to another x-coordinate. By this way, the changing of about 10 coordinates can move the whole axle.
Fig. 10: Modular construction of the model
For static investigations of its behaviour, the model can be build up on a test-rig. The tires are replaced by simple springs with the according radial stiffness. Depending on the investigation, the body-frame can be mounted to the ground und e.g. the tire-patches can move in z-direction, specially for toe-curves of passenger-cars. With this test-rig the steering-kinematic between right an left front wheel was investigated.
For controlling the different driving-maneuvers, the adm-file contains the steering-gear-motions for each maneuver. With the according acf-file, all motions become deactivated except that one which is needed for the desired maneuver. So with one model-file all maneuvers can be executed.
For holding the desired course, the steer-angle can be given as a time function (e.g. sinusoidal steer) or be computed by a controler (e.g. stationary state cornering) so that the course-radius keeps constant. The controler for stationary state cornering used in the model is a serial connection of a PD- and a PT1- controler based on the comparison between expected and real yaw rate for the current velocity.
The input-variables are known in the simulation. So the variable u can be computed. The controler-output-variable y is transferred as input for the steering motion.
Fig. 11: Feedback control system for stationary state cornering
The driving moments are also selected in the acf-file. Depending on the driving maneuvers, the different possible driving moments for a rising (stationary state cornering) or a constant velocity course (e.g. sinusoidal steer) are selected by deactivating the unwanted moments. The desired moment is determined by comparison of expected and real velocity.
Dipl.-Ing. Peter Holdmann
