CVP Powertrain Analysis - Engine and Vehicle simulation · PDF fileIntegrate high fidelity...
Transcript of CVP Powertrain Analysis - Engine and Vehicle simulation · PDF fileIntegrate high fidelity...
U.S. ARMY TANK AUTOMOTIVE RESEARCH, DEVELOPMENT AND ENGINEERING CENTER
ENERGY ANALYSIS OF A POWERTRAIN AND CHASSIS INTEGRATED SIMULATION ON A MILITARY DUTY CYCLE
GT Suite User’s Conference: 9 November 2015
1
Denise M. Rizzo US Army TARDEC, Warren, MI
Periannan Kumaran Pratt & Miller Engineering, New Hudson, MI
Jonathan Zeman Gamma Technologies, Westmont, IL
UNCLASSIFIED: Distribution Statement A. Approved for public release.
Current Project
• Goal: Integrate high fidelity models of powertrain and vehicle dynamics in a co-simulation environment to understand the influence of chassis dynamics on fuel efficiency and energy losses.
• Background: To date, powertrain and vehicle dynamics models make simplified assumptions about their respective interaction. While this is reasonable under steady state conditions, the transient effects are unknown.
• Applications: Soft soil mobility, obstacle crossing such as step climb, steering loading under a non-straight maneuver such as on a duty cycle course, and fuel economy computation
• Potential Benefits: Understanding of interaction between powertrain and chassis Enables chassis-powertrain system design optimization Framework for chassis-powertrain system optimal control design Delivers an integrated chassis-powertrain system toolset for modeling & simulation
2 UNCLASSIFIED: Distribution Statement A. Approved for public release.
Vehicle Dynamic Simulation Environment
• In vehicle dynamics simulations, vehicle models with detailed subsystems such as engine/powertrain along with detailed chassis are rarely developed to study the integrated behavior of the combined full vehicle systems.
• MSC.Adams is a widely used Multi Body Dynamics Software currently used at TARDEC to model military vehicles and to perform vehicle dynamic simulations.
• The transmission model currently used in these vehicle models are simplified and lacks details in system dynamic behavior. – Typical use of look up tables for torque-speed map – Lacks automatic transmission (version 2012) – No representation of turbo behavior – Compromises the accuracy of transient performance of engine and
transmission
4 UNCLASSIFIED: Distribution Statement A. Approved for public release.
Chassis Vehicle Dynamics Model
• High Mobility Multi-Purpose Wheeled Vehicle (HMMWV) used by US Army
• Fully nonlinear model with rigid
bodies and lumped masses
• Parts are connected through idealized joints and/or force elements
• Linear springs and nonlinear dampers
• Pacejka 2002 tire model
• Manual transmission
5 UNCLASSIFIED: Distribution Statement A. Approved for public release.
Adams HMMWV Model
Powertrain Simulation Environment
• GT-SUITE is a multi physics platform currently used to model and study engine/transmissions/vehicle behavior at TARDEC 1-D simulations Has the ability to model transmission with much more detail than what is available in
MSC.Adams
• The full vehicle model available in GT-SUITE is a simplified reduced order model
Absence of lateral dynamics Absence of detailed tire models Absence of detailed roads and
terrains, which are important for military vehicle development
GT Simulation Environment Example 6 UNCLASSIFIED: Distribution Statement A. Approved for public release.
Powertrain Dynamics Model
• Transient map based system • Engine characterized by torque output map and fuel map • Additional dynamics for turbocharger lag • 4 speed automatic transmission with lock-up clutch • Gear ratio, mechanical efficiency and friction loss included
GT Drive HMMWV Model
7 UNCLASSIFIED: Distribution Statement A. Approved for public release.
Co-Simulation Framework
• Improved hardware and computing power have made this framework a viable solution
• Ability to integrate high fidelity subsystems with detailed full vehicle models
• Common methodology in integrated simulations
• Benefit from the individual strengths of each simulation tool
• Avoid development of duplicate subsystems and components
8 UNCLASSIFIED: Distribution Statement A. Approved for public release.
Integrated Model
GT-SUITE
Co-simulation
• Full vehicle chassis dynamic model developed in MSC.Adams/Car
• Powertrain dynamic subsystem model developed in GT-SUITE
• Use of Simulink as the coupling environment
• Both Adams/Car and GT-SUITE are coupled to Simulink through S-function blocks
• Simulink behaves as the “Master” of both GT-SUITE and Adams/Car
• Communication interval = 1.0 ms
• Subject the integrated vehicle model through various handling maneuvers and terrains of interest
9 UNCLASSIFIED: Distribution Statement A. Approved for public release.
Integration Validation: Acceleration Event
• Perform full load acceleration event to validate the model integration Standalone GT-SUITE Standalone Adams Integrated Model
• Use GT-SUITE full load acceleration results as the baseline. GT-
SUITE is currently used as the simulation tool for automotive performance evaluations
• Update the following parameters in standalone Adams model to match GT-SUITE point mass model Aggregate mass Unloaded tire radius Tire rolling resistance Aerodynamic drag force Gear ratio and shift RPM
10 UNCLASSIFIED: Distribution Statement A. Approved for public release.
Integration Validation: Acceleration Event
11 UNCLASSIFIED: Distribution Statement A. Approved for public release.
Integration Validation: Acceleration Event
12 UNCLASSIFIED: Distribution Statement A. Approved for public release.
• Shift RPM Comparison: Model Comparison – RPM limits for the integrated model are same as GT-SUITE
standalone – Adams/Car standalone model uses single shift RPM for all gears
Integration Validation: Conclusions
• Integrated model speed profile from the co-simulation correlates reasonably well with standalone GT-SUITE model speed profile
• Top speed difference < 1.0mph
• Differences between the standalone GT-SUITE simulation and integrated
simulation results could be attributed to – Tire model differences – Aero dynamic model difference – More details in the integrated vehicle model compared to standalone
GT-SUITE point mass vehicle model
• Standalone Adams simulation speed saturates early due to engine revolution limit in the powertrain subsystem
• Integrated model is well suited to perform additional correlation studies
13 UNCLASSIFIED: Distribution Statement A. Approved for public release.
Additional Simulations
• To demonstrate the advantages of the integrated co-simulation environment over the standalone GT-SUITE and Adams dynamic models
• Events selected – Typical military vehicle requirements 18 inch step climb 60% graded road
14 UNCLASSIFIED: Distribution Statement A. Approved for public release.
18 inch Step Climb
• Determines the required torque to climb a step of specified height
• Performed at WOT
• Starts from standstill position
• Not supported in standalone powertrain dynamic model
• Transfer case at low position for increased torque
15 UNCLASSIFIED: Distribution Statement A. Approved for public release.
18 inch Step Climb: Vehicle Speed and Torque
Simulation Peak Torque (Normalized)
Adams Standalone WOT 1.0
Integrated WOT 0.45
16 UNCLASSIFIED: Distribution Statement A. Approved for public release.
18 inch Step Climb: Conclusions
• Initial speed drop in the standalone Adams model is due to vehicle bouncing off the step before climbing
• The high torques in standalone Adams model results are due to clutch slippage at low engine RPM
• Integrated model speed controller takes longer to settle down to target
speed. This may indicate a need to tune the controller to better follow the target speed at low speeds
• The ability to evaluate a step climb transiently in a simulation environment has now been realized with the integrated model
• The integrated model is more suited to evaluate torque requirements compared to the standalone Adams model for step climb simulations
17 UNCLASSIFIED: Distribution Statement A. Approved for public release.
60% Graded Road
18
• Determines the ability of the vehicle to climb specified inclination at a constant speed
• Performed at low speed
• Starts from a flat road
• Transfer case at low position for increased torque
UNCLASSIFIED: Distribution Statement A. Approved for public release.
60% Graded Road: Vehicle Speed
19 UNCLASSIFIED: Distribution Statement A. Approved for public release.
60% Graded Road: Left Front Halfshaft Torque
20
UNCLASSIFIED: Distribution Statement A. Approved for public release.
60% Graded Road: Conclusions
21
• The 60% grade maneuver with vehicle dynamics fully
integrated has been shown, however steady state speed control at low speed was elusive and further investigation is required
• Standalone Adams model shows a speed drop when the vehicle was coming off the graded road. This in turn resulted in clutch slippage as seen in the torque results
UNCLASSIFIED: Distribution Statement A. Approved for public release.
Fuel Economy and Energy Loss
22
• Current fuel economy studies are performed in standalone powertrain environment with the reduced DOF vehicle model
• Studies are constrained to straight line simulations
• Appropriate for courses with minimal turns and not much influence from lateral dynamics
• Cross country and secondary road have very steep elevation changes and sharp turns
• Need to understand the influence of lateral dynamics on energy loss related to these courses
UNCLASSIFIED: Distribution Statement A. Approved for public release.
Influence of Lateral Dynamics on Energy Loss
23
• Perform a straight line simulation for the length of Churchville profile on a flat road
• Repeat the simulation with Churchville profile on a flat road.
Churchville road is considered a cross country terrain
UNCLASSIFIED: Distribution Statement A. Approved for public release.
Influence of Lateral Dynamics on Energy Loss
24
• Lateral dynamics does have an impact on the fuel economy mainly due to tire/road interaction.
• Fuel economy for Churchville profile is 30% less than straight line.
• Vehicle speed has an influence on the fuel economy. Higher speeds
resulted in larger difference in fuel economy.
Road Profile Fuel Economy (normalized)
Straight Line 1.00
Churchville 0.70
UNCLASSIFIED: Distribution Statement A. Approved for public release.
Influence of Lateral Dynamics on Energy Loss
25
• Noticeable speed drop during turns
• Increased power requirement to maintain required speed
UNCLASSIFIED: Distribution Statement A. Approved for public release.
Influence of Lateral Dynamics on Energy Loss
Increased longitudinal tire forces resulting from increased lateral tire forces
26 UNCLASSIFIED: Distribution Statement A. Approved for public release.
Influence of Lateral dynamics on Energy Loss
27
• Tires are subjected to simultaneous lateral and longitudinal slip conditions in a turn
• Application of lateral slip at a given slip ratio tend to reduce the generated longitudinal forces resulting in reduced speed
• More power is required bring the
vehicle back to target speed
Reference : Thomas D. Gillespie “Fundamentals of vehicle dynamics”, 1992
UNCLASSIFIED: Distribution Statement A. Approved for public release.
Analysis of Energy Losses
28
• Focus energy loss analysis on following areas – Engine and Accessories – Transmission – Driveline
• Engine and Accessories losses
– Accessories • Alternator and Fan
– Exhaust • Indicates the fuel burned
UNCLASSIFIED: Distribution Statement A. Approved for public release.
Analysis of Energy Losses
29
• Transmission losses: – Larger losses in the transmission
and torque converter for the Churchville course.
– Torque converter losses are especially apparent due to the longer time in the inefficient unlocked position.
• Driveline losses:
– Brakes are not applied during the event
– Larger losses are shown in the driveline components for the Churchville course.
UNCLASSIFIED: Distribution Statement A. Approved for public release.
Road Test Comparison on Churchville-B
30
• Simulate the integrated model on actual Churchville-B profile
• Churchville-B profile contains steep grades
• Road coordinates were measured by profilometer in order to quantify the road roughness
• Create MSC.Adams road profile from measured data • Model Updates
– Aerodynamic drag coefficient – GVW – Unloaded tire radius – Tire rolling resistance – Power loss
• Event performed at a constant speed
UNCLASSIFIED: Distribution Statement A. Approved for public release.
Road Test Comparison on Churchville-B
31
• Fuel economy value compared very well with the road test results
• Integrated model captured all of the energy losses
Model Configuration Fuel Economy(mpg) - normalized
Road Test 1.00
Simulation 0.99
UNCLASSIFIED: Distribution Statement A. Approved for public release.
Summary and Conclusions
32
• A transient map based powertrain dynamic model built in GT-SUITE was successfully coupled with a detailed chassis dynamic model in MSC.Adams/Car with MATLAB/Simulink as an intermediate coupler
• Specific simulations pertaining to military vehicles were performed to confirm that the integrated model will be successful in these simulations
• A detailed study was performed to understand the influence of lateral
dynamics on fuel economy and energy loss
• A study comparing the fuel economy in integrated vehicle model to road test on Churchville terrain resulted in good correlation providing high confidence in the integrated vehicle model and co-simulation frame work.
• Simulations such as step climb and soft soil (not presented here) provides an early understanding of power and mobility requirements that was not possible in the past
UNCLASSIFIED: Distribution Statement A. Approved for public release.
Future Work
33
• Use FMI framework or direct GT/Adams co-simulation and eliminate the
intermediate Simulink layer
• Replace the constant speed with limit handling which mimics the real world driving scenario
UNCLASSIFIED: Distribution Statement A. Approved for public release.
Disclaimer
34
Reference herein to any specific commercial company, product, process, or service by trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or the Department of the Army (DoA). The opinions of the authors expressed herein do not necessarily state or reflect those of the United States Government or the DoA, and shall not be used for advertising or product endorsement purposes.
UNCLASSIFIED: Distribution Statement A. Approved for public release.
Churchville-B Animation
35
UNCLASSIFIED: Distribution Statement A. Approved for public release.