curriculum vitae
english version

Ph.D Thesis

Dottorato di Ricerca in Materiali e Tecnologie Innovative

Combustione e Conversione dell’Energia- XV ciclo


Sviluppo di una metodologia evolutiva per il progetto di motori a c.i. validata mediante rilievi sperimentali

Teresa Donateo

Tutor: Prof. Ing. D. Laforgia, Ing. A de Risi

Coordinatore: Prof. R. Cingolani


Extended abstract

The more and more restrictive legislation about pollution and the need of dramatically reducing the development time of new engines have introduced a new challenge in the design of internal combustion engines. In this  scenario, the shape of the combustion chamber plays a key role in that it strongly affects the combustion process and emissions mechanisms of formation. Designers have to abandon the traditional “rule of thumb” in the definition of the combustion chamber shape and search for innovative solutions.

The development of new combustion chambers using the trial and error approach is no longer acceptable due to the small number of chambers that can be analyzed and compared via experimental studies. On the contrary a wide parametric study can be easily performed with the use of simulation codes. However, even with the use of CFD codes the design of the combustion chamber is still bound to tradition if the choice of the solutions to be analyzed is committed to designers. 

In the optimization method developed and illustrated in the Ph.D thesis, the selection of the potential combustion chambers to be analyzed is assigned to a genetic algorithm which is able to find the best chamber configuration among millions of possible solutions with an automated procedure.  This method is quite generic but, to illustrate its capability, a direct injection diesel engines equipped with a Common Rail injection system was considered.  Due to the high flexibility of the new injection systems in fulfilling arbitrary injection strategies, direct injection diesel engines are now widely diffused in the automotive market and a great support has been given to research in order to optimize both injection systems and combustion chamber.

The optimization of this kind of engines is very challenging due to the high number of control parameters to be optimized together with engine geometry: injection strategy, EGR, injection pressure, boost pressure, etc.  Of course, to completely optimize engine performance and pollutant emissions, a unique GA run including all these parameters should be carried out and a very high number of operating modes should be included. This approach should be capable of finding a completely optimized configuration of the input parameters but it is unlikely to explain why this configuration works well. Moreover, by using this approach the complexity of the system strongly increases when different operating modes are taken into account; in fact, injection strategy and the other control parameters can change according to engine load and speed. For this reason, a two-step optimization method was developed and used to obtain diesel engines able to fulfill the future EURO IV regulations.

The first step aims at optimizing engine design for a fixed injection strategy without EGR with reference to several operating modes. Then, the shape of the combustion chamber is kept constant and the optimal values of the control parameters for each mode are identified.

For the first step, two different strategies were considered: a single-pulse injection strategy with a six-hole injector design  and an innovative injection strategy characterized by a split injection with the first pulse injected very far from the top dead center with a seven-hole nozzle. The second strategy was selected so that a partially homogeneous charge was obtained in the combustion chamber before ignition. The input parameters for the algorithm included five geometric parameters defining the combustion chamber shape. The optimization was simultaneously performed for different engine operating conditions, i.e. load and speed values, weighted according to their occurrence in the European Driving Test (EDT).

A two-pulse injection strategy was considered in the second step and both energizing time and advance of the first injection pulse were allowed to change. The main injection pulse was obtained with a fixed energizing time, while its advance was changeable. For simplicity, this strategy is henceforward referred to as “pilot injection” even if some levels of energizing time and advance of the first pulse are beyond the usual limits of pilot injection. Two levels of injection pressure were taken into account. The results of this step were useful to verify whether the chamber configurations selected in the first step keep their capability to reduce engine emissions regardless of the injection strategy.

The second step of the optimization method can be also performed via experimental investigation by directly optimizing the injection control parameters on a experimental test bench. An application of the method to a four cylinder diesel engine in order to reduce emissions and fuel consumption is illustrated in the thesis. The results proved the effectiveness of using experimental investigations as evaluation method for GAs optimization. In fact, optimized injection strategies were found for all tested operating conditions.

The design method illustrated in the present investigation has been validated by building and testing the optimized combustion chamber.  The experimental results showed that the optimized chambers developed with the methods are effective in reducing soot and HC emissions in the case without EGR and improve also NOx-soot trade-off when EGR is used thus confirming the capability of the method in developing low-emission diesel engines.

The PhD thesis consists of an introduction, four chapters and a conclusion section. A brief summary of each part is reported here.


It contains the historical development and the state of the art of the combustion chambers for direct injection diesel engines as found both in scientific and patent literature


This chapter is meant to illustrate the computational code used to simulate the engine behavior. A detailed description of the models developed during the PhD course and added  to the KIVA 3V code is reported.


The modified version of the KIVA3V code is validated through comparison with experimental data to assess the capability of the code to predict engine emissions and performance when changing load, speed, injection strategy and EGR. A sensitivity analysis of the influence of mesh size is also included.


The optimization methodology called “Genetic Algorithm” is described and compared with other optimization tools. The application of genetic algorithms to complex multi-objective problems is analyzed and performed via an innovative method.


The optimization method is applied to different engine configurations and its capability to aid the reduction of diesel engine emissions is proved by means of both numerical analysis and experimental validation.


It contains a summary of the thesis and underline the important achievements reached during the PhD course.



Home | papers | Hi.Pe.GEO | Ph.D Thesis

Ultimo aggiornamento: 17-09-08