Special AVL the Chassis Dynamometer As a Development Platform
Total Page:16
File Type:pdf, Size:1020Kb
Special AVL The Chassis Dynamometer as a Development Platform A Common Testing Platform for Engine and Vehicle Testbeds 40 Total Energy Efficiency Testing – The Chassis Dynamometer as a Mechatronic Development Platform 45 “Easy and Objective Benchmarking“ 49 Interview with Christoph Schmidt and Uwe Schmidt, AVL Zöllner Innovative Use of Chassis Dynamometers for the Calibration of Driveability 51 Offprint from ATZ · Automobiltechnische Zeitschrift 111 (2009) Vol. 11 | Springer Automotive Media | GWV Fachverlage GmbH, Wiesbaden SPECIAL AVL A Common Testing Platform for Engine and Vehicle Testbeds In order to achieve time-savings during vehicle development, companies are increasingly looking to run the same tests on the engine test bed as on the chassis dynamometer. The aim is to correlate results in order to highlight differences and their influencing factors, as well as to verify the engine test bed results, and to achieve the additional benefit of reusing existing tests. A common test automation and data platform is required. The article describes an application in which this has been achieved for emissions certification testing and discusses the value of upgrading the chassis dynamometer to a higher level of automation. 2 ATZ 11I2009 Volume 111 1 The Synergies between Engine, – the ability to accurately simulate the The Authors Driveline and Vehicle Testbeds missing vehicle components on the testbed (e.g.: the vehicle powertrain The product development processes at on an engine testbed) Charles Kammerer MSc. OEMs and the component development – the ability to reproduce environmen- is Product Manager process of tier 1 suppliers rely on exten- tal conditions realistically for Test System Auto- sive testing phases of the different pow- – the ability to support iterative devel- mation at AVL List ertrain components and of the complete opment loops efficiently between the GmbH in Graz (Austria). vehicle. The testing is carried out using different stages of the process. various types of test benches: hardware- This last point is the key for delivering in-the-loop benches, component testbeds, the desired efficiency improvements sin- Roland Schmidt engine testbeds, driveline testbeds and ce it must allow the correlation of simi- is Head of Application vehicle testbeds. Finally, fleet testing of lar testing tasks carried out at different Development for Test prototypes takes place on the test track stages in the process for the purpose of Instruments and Sys- or road for the final adjustments. validating results obtained using simula- tems at AVL Emission In each of these test environments, tion, or to comply with legislative re- Test Systems GmbH in various testing tasks are carried out. For quirements. Gaggenau (Germany). instance, an engine testbed is used during For example, emission tests are car- the development phase to verify the en- ried out on an engine testbed and repea- gine durability and its thermodynamics. ted on the vehicle testbed for certifica- Ing. Gerald Hochmann It is also used to set up the base calibra- tion; drivability assessment and pow- is responsible for the tion of the ECU and predict the engine ertrain calibration optimization are car- Application of Engine emission behavior using vehicle simula- ried out on engine, driveline and vehicle Test Systems at AVL List tion. testbeds and verified later in the vehicle GmbH in Graz (Austria). The OEMs are coming under ever-in- on the road; climatic testing takes place creasing pressure to reduce the time to on the engine testbed and again later on market of new vehicles while saving on the vehicle testbed; durability testing development costs. This translates into a takes place on the transmission testbed need for an efficient and shorter prod- and again later on the road or on the ve- uct development process. A key element hicle testbed, Figure 1. in reducing development time is front- One well-known way of facilitating the loading the testing of vehicle character- correlation is the use of a common simu- istics earlier in the process. Front-load- lation platform which also avoids the ing not only allows time-savings and a need to develop and maintain multiple reduction in the number of vehicle pro- models. totypes needed; it also supports a reduc- Another source of productivity gains is tion of development costs by addressing the use of a common automation plat- unplanned design changes much earlier form across all types of testbeds. in the process in a cost-efficient manner, as the later a component failure is de- tected the more expensive the fixing 2 The Common Automation Platform will be. However, this requires a state-of-the-art The automation test system can be split development tool chain that fulfils the into three different layers in terms of following requirements: software functions, Figure 2: Figure 1: Simulation across the product development process ATZ 11I2009 Volume 111 3 SPECIAL AVL ronment rather than on the road. One example is the shift quality and drivabil- ity assessment application. Being able to automate such an application on a vehi- cle testbed may bring significant produc- tivity gains and address the increasing effort required for transmission calibra- tion. All this requires the vehicle testbed TAS to be able to: transfer testruns and correlate test results between the engine Figure 2: The architecture of a common automation platform and the vehicle testbeds (e.g.: for emis- sion testing); transfer testruns and corre- late test results between the driveline and the vehicle testbeds (e.g.: for power- – the controller – for testbed and unit- cle parameters for the simulation on an train calibration optimization); transfer under-test control engine testbeds and the parameters for road profiles recorded in the vehicle and – the test automation system (TAS) – for the chassis dynamometer). replicate them on the vehicle testbed data acquisition and test automation In addition, significant benefits can (e.g.: for endurance testing); provide the – the application (e.g. an emission test be derived from the effects of scale and frameworks and interfaces to the calibra- run such as FTP 75). standardization, which contribute to re- tion tool chain (e.g.: automatic calibra- If the controller is specific to the differ- ducing the total cost of ownership of the tion software, direct access to the xCU ent types of testbed, the TAS is not. There- TAS. and the application systems); provide the fore, provided consistent system architec- For instance, the “one window” ap- connectivity to the vehicle testbed sub- ture on engine, driveline and vehicle proach reduces training needs and main- systems (e.g.: dynamometer, emission testbeds, the same TAS can be used on all tenance costs; a common data handling benches, driver-robots, measurement de- types of testbeds and, if it successfully ab- and IT infrastructure is used across the vices, in-vehicle buses, facility and condi- stracts the application from the control- whole product development process; tioning systems). ler, the application is then portable from common process management tools are Up to now, automation of the chassis one type of testbed to another. used across the whole product develop- dynamometer was often limited to data Being able to port the application ment process for scheduling, equipment acquisition while the testbed control was from one testbed type to another allows management and supervising. left to the chassis dynamometer control- the synergy potential between the differ- While the need for vehicle testbeds ler and the human driver or robot sys- ent testing steps of the product develop- may have been challenged by front-load- tem. However, to fulfill the above require- ment process to be efficiently leveraged. ing many testing tasks, its use remains ments, one requires a TAS for the vehicle Result data can be correlated, test specifi- more crucial than ever for verifying the testbed which is able to fully automate cations can be reused (e.g.: shift quality results obtained earlier in the process the testbed and provide the required and drivability assessment on a power- while also enabling the execution of tra- compatibility with the TAS of other test- train testbed and on a vehicle testbed). ditional in-vehicle activities in a more bed types in terms of architecture and Parameters can also be reused (e.g.: vehi- deterministic manner in a testbed envi- data formats. 3 The AVL Solution AVL’s solution is to use PUMA Open as the automation platform for rig testing: from the component testbed to the en- gine, powertrain and vehicle testbed. While PUMA Open is well-known and well-established as the TAS for compo- nent, driveline and engine testbeds, its extension to vehicle testbeds was re- quired to fulfill the requirements of a common automation platform. Thanks to its modular design and its Figure 3: Architecture scalability, PUMA Open can be tailored of the software solution for a wide range of vehicle testing appli- for vehicle testbeds cations, from simple data logger to ad- 4 ATZ 11I2009 Volume 111 Table: Automation of the vehicle test bed with PUMA Open Open through a real-time interface al- lows the same degree of automation and Data acquisition PUMA Open supports industry-standard bus technology such as CAN, dynamic control performance to be Profi bus or FireWire and state-of-the-art ASAM compliant interfaces for achieved on a vehicle testbed as on an en- ECU/TCU application and diagnostic systems (MCD3MC and MCD3D) or gine or driveline testbed – especially rais- for automatic calibration tools (ACI). ing the repeatability and reproducibility Real-time execution The parameterization of a test run is made easy through a graphical of tests on the vehicle testbed. These per- block sequence editor, thus avoiding the need for programming skills. formance gains are a must for new inno- Even complete road profi les can be directly imported in the test run vative applications on the vehicle testbed: enabling a real-time replication of road data.