Simulation for the prediction of fuel consumption, driving performance and lifetime

Overview

winEVA (EDP-programme for the analysis of consumption and driving performance) represents a software package for the powertrain simulation on computers running the Windows operating system. It can be linked to the programmes for the data logging - e.g. winADAM (automated data logging in the mobile operation) - and to the programmes for the fatigue life calculation - e.g. winLIFE (fatigue life information by finite elements). Very diffrent types of vehicles (cars, trucks and special vehicles can be simulated on various roads (street, terrain). In addition to the pure computer simulation winEVA is also used for:

  • real-time control of dynamic test rigs
  • driving simulator

 

The vehicle model

The vehicle model is built up in a block-oriented way (see figure) like it is usual today and already realised in many software products.

Up tp now, this system has been used for vehicles with the following drive types:

  • gearbox with constant ratio in each gear (max. 20 gears, to be switched manually or automated)
  • automatic transmission
  • series-hybrid, parallel-hybrid
  • catenary drive (trolley bus)
  • electric drive
  • continuously variable transmissions (electrically, hydrostatically)

Furthermore, almost any power train configurations can be modelled.

Environment and traffic

The inclusion of the environment (topography, curvature, sight distance) is a decisive parameter for realistic simulations (see figure below).

Model of vehicle and track

 

The boundary conditions and hindrances, caused by traffic regulations and by the traffic itself, play a major role as well. The following picture shows which events are registered and included as data in the simulation.

 

Events registered during vehicle measurements (winADAM) and transposed in simulation conditions

 

Many customers prefer the automatic generation of a track simulation based on measurement data (winADAM) due to the low effort regarding the data input.

Behaviour of the driver

The behaviour of the driver can be plotted differently depending on requirements:

  • repeated run of a measured speed course
  • independent speed selection based on the measured environment data

The gear selection of the driver for manual transmissions is also realised for trucks up to 18 gears.

The actions of the driver are influenced by motivation factors related to person and event. The behaviour of the driver is also influenced by the question whether an event is predictable or unpredictable. The figure below shows a driver with very different motorised vehicles on the same track with an identical driver. As you can see, the driver model adapts very realistically to the vehicle.

Course of a simulated speed using different vehicles

Test rig control

When using winEVA for the control of test rigs for power trains also measured driving cycles can be simulated with same speed realistically and with little effort. winEVA defines the accelerator pedal position, the brake pedal position and the numbers of revolutions per minute and/or the torques at the electric motors of the test rig. At a major vehicle manufacturer it could be proved that the results for the stress received by this test rig control are much more realistic than by using conventional test rigs. The picture shows the priciple by means of a transmission test rig which simulates the data measured with winADAM with good approximation.

schematic representation of the data flow during test rig simulation

Solution strategy

Based on the equations of motion of the vehicle all essential values within the power train are calculated. You can work optionally with torsionally flexible components, mass inertia and the real excitation torque of the combustion engine which is necessary for the analysis of torsional vibrations.

In order to receive realistic simulations the driver model verified in test is also very important. This model allows the definition of a desired speed course which applies to a large number of vehicles. Furthermore, a measured speed course can exactly be simulated - considering physical limits.

Driver and vehicle model access the data of the simulation track which is defined by the speed course, topography, coefficient of friction, rolling resistance coefficient, curvature and visibility of each track interval. The length of each track intervals may be lying between a few centimetres and several hundred metres depending on the task.

The components of a vehicle must be described very exactly. To do that, extensive data are required. In case of analytically known relationships the equations are directly used (air resistance, hydrodynamic relationships, etc.), in case of components whose characteristics can only be represented by characteristic diagrammes (engine diagramme, efficiency characteristics of the mechanic parts of the transmission, etc.) these relationships are described in matrix form.

Validation

The comparison of test drives with simulation results was carried out and published by a customer. The figure shows the results of a test drive at the Nürburgring for different values of the power train based on measurement and calculation.

Results of test drives at the Nürburgring in comparison with simulation results

Interfaces and result representations

Stationary characteristics

The characterisation of the vehicle by the following more or less stationary and stationary characteristics helps in the assessment of the vehicle as well as in finding errors in the input data.

Tensile force-, number of revolutions per minute- and efficiency diagram of a city bus with lines of the same slope
Braking force diagram as function of speed
Acceleration diagram (road, acceleration, speed) as fuction of time

Time series representation

Simulated values as function of time

Classifications

Torque rev.-collective (gearwheel or bearing)
Frequency distribution in the motor characteristic field
Percentages of time slices of the motor torque
rpm spectrum
Percentages of time shares of the motor rpm

Transfer of data measured by means of wínADAM

Data measured by means of the measurement data collecting system winADAM (automated data collecting in the mobile application) can be transmitted and processed (smoothing functions, filtering). They are immediately available for a simulation. The data structure (ASCII-data) is documented so that the user can write his own interfaces to his own measurement system, if required.

Data transfer for fatigue life calculations

A data transfer from the programme winEVA to winLIFE is possible without any problems. A value obtained by the simulation (for example torque, force) can be transferred to winLIFE and a fatigue life calculation can be carried out afterwards .

Driving cycles of traffic flow simulation

The simulation of traffic flow is integrated in winEVA. Realistic driving cycles can be generated based on a microscopic model for highways. These driving cycles can be the basis for the simulation of a power train. In this way, it is possible to take the traffic flow into account.

Publications

[1] Willmerding, G.: Ein Simulationsmodell für den Autobahnverkehr, Teil 1. ATZ Heft 5 1992

[2] Willmerding, G.: Ein Simulationsmodell für den Antriebsstrang, Teil 2. ATZ Heft 6 1992

[3] Willmerding, G.: Jehlicka, E.: Verbrauch-suntersuchungen an Stadtbussen; Der Nahverkehr Heft 5 1986

[4] Willmerding, G.: A simulation system to study the working conditions of vehicles and to develop fuel efficient drivetrains. publication on the FISITA-congress 1992,  Inst. of mech. Eng. 1992

[5] Willmerding, G.: Ein Simulationsmodell für Kraftfahrzeuge im Verkehrsfluß, publication on ASIM-congress TU-Berlin 1993, Fortschritte in der Simulationstechnik, 8. Symposium 1993, Band 6, ISBN 3-528-06555-9

[6] Böhm, Jehlicka, Willmerding: research of a modern freight traffic-system by using computer simulation, publication for the FISITA-congress in Peking 1994

[7] Willmerding, G; Trübswasser, F; Häckh. .: A simulation system to predict fuel consumption and emissions considering the traffic flow. 5. Aachener Kolloquium Fahrzeug- und Motorentechnik 1995. Tagungsband.

[8] Häckh, Willmerding, G.: Handbuch zum Computerprogramm winLIFE. 155 Seiten, Steinbeis-TZ-Verkehrstechnik Ulm, 1.10.1997

[9] Häckh, J; Willmerding, G.: Handbuch zu winE-VA, 199 Seiten, TZ-Verkehrstechnik. 1998

[10] Willmerding, Dietzel, Körner: Rechnergestützte Entwicklung von Schaltprogrammen für Automatik-getriebe, 3. Stuttgarter Symposium Kraftfahrwesen und Verbrennungsmotoren 1999, Seite 843 - 856, ISBN-Nr. 3-8169-1751-8

[11] Willmerding, G.; Häckh, J; Schnödeweind, K.: Fatigue Calculation using winLIFE, Vortrag auf dem  NAFEMS-Seminar Fatigue Analysis am 8.11.00 in Wiesbaden, Herausgeber: The International Association for the Engineering Analysis Community, www.nafems.de

[12] Willmerding, G; Häckh, J;  Berthold, A.: Vor-hersage von Fahrzyklen, Belastungen und Lebens-dauer auf der Basis von gemessenen Streckenda-ten, Vortrag auf der Tagung Fahrwerktech 2001 am 8. und 9. März in München; Veranstalter: TÜV München Tagungen und Kongresse, Westendstr, 122

[13] Willmerding G.: Vorhersage der Lebensdauer dynamisch belasteter Bauteile durch Kombination von Lebensdauerberechnung mit Finite Element Methode; Vortrag Nr. 14 auf der Tagung Fahrwerk-Tech 99 Adaptive Fahrwerksysteme, 4 bis 5.3.1999, München, Tagungsunterlagen TÜV-Akademie München.


[14] Willmerding, G.: Lebensdauerberechnung dynamisch belasteter Bauteile für den multiaxialen Fall in Kombination mit Mehrkörperdynamik und Finite Element Methode; Vortrag auf der MSC-Anwenderkonferenz am 21. und 22.6.1999 in Weimar, Herausgeber Mac Neal Schwendler Corporation, München

[15] Willmerding, G; Häckh, J; Berthold, A: Driving Cycle, Load and Fatigue Life Predictions based on measured Route Data; Vortrag auf der ATT-Tagung in Barcelona 2001, SAE-Paper 01ATT120


[16] Häckh, J; Willmerding, G; Kley, M; Binz, H; Körner,T.: Rechnerische Lebensdauerabschätzung von Getriebegehäusen unter Einbeziehung realer multiaxialer Belastungen; DVM-Tagung Fulda vom 5. bis 6.6.2002,VDI-Berichte N2. 1689, 2002 Seite 303 - 317


[17] Körner, T; Depping, H; Häckh, J; Willmerding, G; Klos, W.: Rechnerische Lebensdauerabschätzung unter Berücksichtigung realer Belastungskollektive für die Hauptwelle eines Nutzfahrzeuggetriebes; DVM-Tagung Fulda vom 5. bis 6.6.2002, VDI-Berichte N2. 1689, 2002 Seite 275 - 285


[18] Körner, T; Depping, H; Häckh, J; Willmerding, G.: Fatigue Life Prognosis for Transmissions based on critical Component Spectrum; World Automotive Congress FISITA 2002, Helsinki, Paper Nr.F02V091

[19] Willmerding, G.; Häckh, J.; Körner, T.: Ein durchgängiges Antriebsstrang-Simulationssystem im Entwicklungsprozess von Automatikgetrieben; Vortrag auf der Tagung: Dynamisches Gesamtverhalten von Fahrzeugantrieben, Haus der Technik, 11. bis 12.3.2003 in München.

[20] Häckh, J.; Willmerding, G.: Untersuchung der Einsatzverhältnisse unter realistischen Bedingungen durch Kombination von Messdatenerfassung und Computersimulation; Vortrag auf der VDI-Getriebetagung 2006: VDI-Getriebetagung 2006.

[21] Häckh, J.; Willmerding, G.: Load spectrum prediction for Transmissions und realistic use combining tests and computer Simulations, Vortrag auf der FISTA-Tagung Budapest 2007: FISTA-Tagung 2007.

[22] Seifert, C.;Willmerding, G.: Fatigue Life Prediction of Automotive Drive Trains By Combination of Drive Cycle Measurements and Simulation Using WinLife, Vortrag auf der NAFEMS Tagung Vancouver , Mai 2007 NAFEMS-Tagung 2007.

[23] Willmerding, G.; Häckh, J.: Ein System zur Untersuchung von Fahrzeugantrieben zur Vorhersage von Fahrleistung, Kraftstoffverbrauch und Lebensdauer unter realistischen Einsatzbedingungen, Vortrag auf der VDI-Tagung Erprobung und Simulation in der Fahrzeugentwicklung, 24.6.-25.6.2009, Würzburg Erprobung_Simulation_VDI_Würzburg_2009.

Services / Projects

We investigate your powertrain problems on behalf of you. Examples of completed projects:

  • Prediction of fuel consumption for different passenger car powertrain concepts (3-litre car)on a test track and analysis of the driver influence
  • Savings potential by targeted management of auxiliary consumers
  • Savings potential by engine shutdown in passenger cars
  • Savings potential by brake energy storage
  • Savings potential by tyres with reduced rollling restistance coeffizient
  • Simulation of the powertrain management to optimize the brake energy storage
  • Comparison between different transmission concepts with regard to fuel consumption
  • Influence of the transmission efficiency on fuel consumption
  • Comparison between electrically powerd trucks and the railway with regard to a use at NEAT (new european alpine transversal)
  • Fatigue life prediction of retarder shaft based on measured driving cycles
  • Influence of the shift process during manual shifting on the fatigue life of the transmission
  • Investigation of innovative shift strategies for automatic transmissions by SIL (Software in the Loop-Simulation)
  • Fatigue life prediction of transmission components under realistic load spectra
  • Investigation of continuously variable transmissions on the basis of power split hydrostatic transmissions
  • Driving simulation of CVT-concepts with the aim of developing a control strategy
  • Simulation of different powertrain concepts with the aim of a transmission selection
  • Temperature development in electric motors under realistic driving cycles in commercial vehicles
  • Influence of the traffic density on the speed cycles of commercial vehicles and their influence on the fuel consumption

Further products for the powertrain analysis

winADAM: Aquisition of operation data

winADAM is a mobile measurement and analysis system for vehicles. It collects - without the need for a cable connection to the measuring object - comprehensive operating data like driving speed, lateral acceleration, yaw rate, course angle, geodetic height and longitudinal incline of the road.

These values can be exported to winEVA and used as a basis for a simulation.

winLIFE: Fatigue life calculation in combination with finite elements

winLIFE BASIC
Enables basic methods of the fatigue life prediction including the connection to finite element programmes such as MSC/NASTRAN for Windows.
 
winLIFE MULTIAXIAL
Calculation of special tasks in which the direction of the principal stress changes significantly.This module is a supplement to the BASIC-module which covers also the most dificult questions.
 
winLIFE GEARWHEEL&BEARING
Calculation of gearwheels and bearings according to the standard calculations methods without finite elements.