Predrag Mrdja, MSc, PhD student, Internal Combustion Engines Department, Faculty of Mechanical Engineering, University of Belgrade, Serbia
Developing and building internal combustion (IC) engine test bench submodules for test bench control, digital data acquisition, and online look-up table calibration on the engine control unit (ECU).
Using off-the-shelf NI LabVIEW software and PXI hardware with modular instruments to rapidly solve numerous challenging data acquisition and control problems on the IC engine test bench.
Predrag Mrdja, MSc, PhD student - Internal Combustion Engines Department, Faculty of Mechanical Engineering, University of Belgrade, Serbia
Nenad Miljic, MSc, Teaching Assistant - Internal Combustion Engines Department, Faculty of Mechanical Engineering, University of Belgrade, Serbia
Slobodan Popovic, MSc, Teaching Assistant - Internal Combustion Engines Department, Faculty of Mechanical Engineering, University of Belgrade, Serbia
Marko Kitanovic, MSc, PhD student - Internal Combustion Engines Department, Faculty of Mechanical Engineering, University of Belgrade, Serbia
The Internal Combustion Engines Department (ICED) at the University of Belgrade, Faculty of Mechanical Engineering is the largest and oldest institution devoted to IC engine research in Serbia (including former Yugoslavia). Dealing with one of the most dynamic objects in mechanical engineering, our research requires intensive lab work and cutting-edge acquisition systems. Our in-house-developed, VME Bus-based engine test lab data acquisition and measurement systems date back to the early 1980s and we started to use NI solutions almost a decade ago.
Observing IC engine development reveals two major trends: increased engine performance and harmful pollutant emission reduction requirements. Staying on both trends requires implementing complex engine control algorithms from a series of input parameters. IC engines are complex, dynamic systems affected by many control parameters (aside from environmental conditions and geometric characteristics) such as air-to-fuel ratio, throttle position, spark advance, intake- and exhaust-valve timing, and exhaust gas recirculation (EGR) valve position. Engine torque, fuel consumption, and pollutant emission parameters must be optimized to obtain peak performance throughout the entire operating range of the engine. It is impossible to have optimal ECU look-up tables without performing extensive engine testing, no matter how reliable and advanced IC engine simulation software packages are.
Students have shown great interest in engine management problems and have devoted themselves to developing an open ECU to use in our IC engine test labs. Students built all the components—from engine wiring harnesses, to injection and ignition system power driver modules, to signal conditioning modules. Because the system, based on a Freescale MPC565 microcontroller, is inherently modular, we can easily modify it to use in both lab and testing conditions.
While prototyping the ECU, we performed hardware-in-the-loop (HIL) testing to verify the developed ECU software prior to using it on a real engine. We found NI hardware and LabVIEW ideal for this task. We used the NI PXI multifunction data acquisition devices for simultaneous engine sensor signal simulation and verification. While the 32-bit NI function generator features, such as direct digital synthesis mode, and the NI PXI digital multimeter modules suit the ECU validation process, it is easier to generate engine signals using NI field-programmable gate array (FPGA) hardware, which is the next step in upgrading our HIL system.
The automotive industry typically uses a controller area network (CAN) for in-vehicle data transmission and communication. One CAN communication protocol devoted to ECU calibration is the CAN Calibration Protocol (CCP). We easily established two-way communication using a high-speed NI PXI CAN interface card and the NI ECU Measurement and Calibration Toolkit (see Figure 1).
We used the CCP to fetch data from the ECU software and to calibrate tuneable parameters defined in the ECU development stage. The CCP also read numerous engine sensors (engine speed; intake collector pressure and temperature; throttle position; air mass flow; intake air temperature; and wide-band lambda). Through parameter tuning, we can modify the air-to-fuel ratio or influence ignition timing in real time while the engine is running. This gave us complete insight to optimize the developed ECU software and its basic environment.
Testing an IC engine involves acquiring data from various engine-mounted sensors, as well as the test bench system and IC engine control. We used LabVIEW to develop the entire engine test bench monitor and control system. We relied on NI PXI hardware and PXI modular instruments to comfortably control the engine load, cooling, and fuel supply system. With this system, we can clearly visualize sensor information regarding the following:
Because these readings come from sensors mounted in parallel with the engine’s original automotive sensors accessed by the ECU, we can use this application to test ECU signal conditioning capabilities and software functionality.
Optimizing certain parameters requires a detailed engine work-cycle analysis based on angle-resolved, in-cylinder pressure measurement (engine-indicating). To do this, we use an encoder-triggered NI PXI multifunction I/O module. To reduce the PXI system CPU load, we implemented network-published shared variables to transfer acquired data to another PC with a specially developed LabVIEW application devoted to analyzing and executing indicated parameters. This analysis application relies on in-house-developed script nodes implemented using MathWorks, Inc. MATLAB® software. The nodes saved time and helped us achieve fast analysis. For example, we implemented online calculation of the mass fraction burned (MFB) and rapid determination of its 50 percent angle position. Because this position is strongly related to engine efficiency and is mostly influenced by ignition-timing parameters, we immediately used it to calibrate the ECU.
Thanks to NI PXI hardware and LabVIEW software, we used the engine test bench to calibrate ECU control parameters within one programming environment, which successfully integrates data acquisition and control as well as CAN and LAN communication. The versatility of LabVIEW also helps us easily visualize the acquired data and analysis results. This concept is especially valuable for student engineers building their IC engine testing and development skills.
MATLAB® is a registered trademark of MathWorks, Inc.
Predrag Mrdja, MSc, PhD student
Nenad Miljic, MSc, Teaching Assistant
Slobodan Popovic, MSc, Teaching Assistant
Marko Kitanovic, MSc, PhD student
Internal Combustion Engines Department, Faculty of Mechanical Engineering, University of Belgrade
Kraljice Marije 16, 11120 Belgrade 35, Serbia
Tel.: +381 11 3302 416