【作者】

Michael Engelmayer

【摘要】

该论文已在赫尔辛基举行的第28届CIMAC大会上发表，论文的版权归CIMAC所有。As a response to market demand and fluctuating fuel costs, engine combustion concepts and subsystems must be characterized by maximum flexibility. Researchers, manufacturers and suppliers are forced to develop and optimize combustion concepts for fuel-efficient multi-application engines in a shorter amount of time. With little change in hardware, these engines must be capable of running on different types of fuel with varying properties and meeting emission limits that depend on the area of application and national legislation. The well-known ‘Design of Experiments’ method, or DoE, can help meet these multiple challenges. Making use of a statistical approach is not an innovation itself, but today DoE can be employed to do much more than just optimize a test plan and minimize test bed time. By examining a variety of concrete cases, this article will explore the range of ways the DoE method can be used to advance engine development from its classic application to DoE based model development to purely simulation-based DoS (‘Design of Simulations’) system optimization. In the section on classic application of DoE, the advantages of a statistical approach are compared to a full factorial collection of measurement data in the examination of a high-speed diesel engine focusing on PM emissions. With this approach multiple applications (gen-set, locomotive, etc.) were optimized in terms of fuel consumption and emissions. Two further examples of areas of application to be discussed are the pre-optimization of efficiency and nitric oxides in the case of a dual fuel engine as well as the investigation of the influence of different fuel gas compositions on the operating range (knocking) in the case of a gas engine. In addition, the DoE method can also lay the foundations for model development, e.g. for describing the combustion process. Originating from experimental investigations, measurement-based functions provide the input required for system optimization. With this input, multi-cylinder engine simulation following a statistical approach can be applied to vary engine parameters in order to optimize efficiency while respecting the constraints of emissions, knocking and misfire. Within this paper this process is demonstrated for a pre-chamber gas engine. Finally, optimization work involving multiple parameters can also be conducted entirely with simulation without using any measurement data. Using simulation instead of measurements can result in long processing times, e.g. detailed reaction kinetic determination of ignition delay, laminar flame speeds, etc. Therefore, these simulation tasks are treated separately; the results can once again be described by functions and serve as input for system optimization.

【会议名称】

第28届CIMAC会议

【会议地点】

芬兰 赫尔辛基

【下载次数】

3

返回