With respect to behavior following collision of the spray with the wall surface, comparison was carried out between the wall jet model of Naber and Reitz^[6] and the liquid film model of Wakisaka et al.,^[7] with results shown in Fig.2. This shows that the liquid film model is closer to actual data, and thus the liquid film model was adopted for use in the computational method presented here.

　

Fig.1 Comparison of calculated and experimental results for free spray

Fig.2 Comparison of calculated and experimental results for impinging spray on the wall

　

2-3. IGNITION & COMBUSTION MODEL

The Livengood-Wu self ignition model^[8] as proposed by Wakisaka^[9] was used for estimation of self ignition. In this model, self ignition is considered to have occurred when the time-integrated value of the reciprocal of the ignition delay exceeds 1.0. For our purposes, this integration value was considered to move in response to flows between computational meshes. In addition, the laminar turbulent characteristic time combustion model of Abraham and Bracco^[10] was used as the combustion model.

　

2-4. HEAT TRANSFER MODEL

The model of Launder and Spalding^[11] was used to predict turbulent flux on the combustion chamber walls from high temperature gas. The various models used in the current context are summarized in Table 1.

　

3. EVALUATION FOR CALCULATION

　

In order to verify the appropriateness of the computational method that was developed, combustion conditions were observed in an observation engine, and combustion calculations were performed accordingly.

　

Table 1 Calculation models

　

3-1. OBSERVATION TEST

Test equipment specifications and test conditions for the observation test are shown in Table 2. In order to model combustion in a low speed 2-cycle engine in the observation test, side injection was adopted as illustrated in Fig.3.

　

Table 2 Test equipment specifications and test conditions

Fig.3 Combustion chamber

BACK　　　CONTENTS　　　NEXT