News and Library > news 2009 > IP Applies Cutting Edge Skills to CR Diesel EMS >

---

IP Applies Cutting Edge Skills to CR Diesel EMS

Date: 10/03/2009

Integral Powertrain has, for many years, been well known for its advanced model based calibration (MBC) processes and their application to advanced gasoline EMS systems.

Over the last 5 years, an increasing share of IP’s workload has been on advanced and production versions of latest generation CR diesel systems. Due to the features and complexities of these systems, IP’s MBC processes have proven more relevant than ever.

Turning to some examples, recent projects have demonstrated the value of screening experiments, model based calibration, and the use of MBC techniques in conjunction with physical modelling. Tools have also been developed to assist with the collection and analysis of OBD data and its subsequent calibration.

In one project, a client had been advised by another consultant to carry out a fully Designed Experiment to generate an engine map for an advanced diesel application. The number of variables was high and the original consultant’s experiment attempted to model all the variables under the full range of engine operating conditions. This produced an experiment of impractical size and cost. Using the IP team’s understanding of the role of each variable, and by conducting screening experiments, it was possible to limit variables to their effective range, create models covering their operating influence, and eliminate certain combinations which are known not to occur in the final application. By using these approaches, the duration of the testing was reduced by several weeks at a significant cost saving to IP’s client.

In another case, IP was responsible for optimisation of a twin series turbocharged application using a current technology EMS. The system was difficult to optimise as a result of significant interactions between the combustion characteristics, twin turbochargers, EGR system and torque model. Using MBC toolbox and IP’s own engine modelling tools, it was possible to combine physical and non physical models (of the mapping data) in order to carry out a desktop optimisation of the boosting system. Once this had been dealt with, the engine mapping became a more manageable activity which was carried out using screening and model based calibration.

In another example, IP was responsible for the turnkey engine management application on a 4th generation common rail diesel application. The EMS includes multiple pilot injection strategies and latest torque based control, with decoupled AFR control. The engine included an electric valve EGR with cooler and bypass, electronic throttle for AFR control, fully variable swirl control valve, exhaust temperature and pressure sensing and cDPF system.

In order to simplify the mapping process, certain practical screening tests were carried out in order to reduce the number of variables. One outcome of this was that speed / load regions were identified with specific trade offs between combustion noise, fuel consumption, gaseous and particulate emissions. In each region, different criteria applied and the correct trade off was made. This enabled the injection strategy to be set and the full designed experiment run with this variable eliminated. This screening approach is powerful, particularly at the early stage of a calibration project. It results in a faster time to the availability of an engine map which can be used for early vehicle testing, with resulting savings in the overall project timing.

IP also uses its own toolset and processes for the calibration of OBD systems on both diesel and gasoline applications. The tools include MATLAB based tools which are used for processing the data collected, and allowing the engineer to visualise the ideal thresholds for detection of OBD faults. A good example is the calibration of misfire monitor, which demands the processing and analysis of a high volume of test data and where selection of the correct thresholds can be difficult. These processes improve the efficiency and robustness of the OBD calibration process, and reduce the engineering cost.

The calibration of misfire monitor is another task where physical modelling can be used to investigate the performance of the system. IP recently used its cranktrain model combined with measured combustion data in order to investigate poor detection signal to noise ratio. The results of this exercise showed that the detection window could be adjusted under certain conditions, and the concern was resolved.

Using screening, MBC, physical modelling and data analysis methods, IP’s advanced processes result in time and efficiency savings, together with higher quality and robustness, delivering excellent value on client projects.