Flow Pattern Simulation Computational Model

Figure 4. Computational Region and Grids (CASE 1)

Boundary Conditions

As shown in Figure 4, the inlet boundary consists of several interface planes representing MSV exit, the fan outlet towards FVABI and the core engine ex­haust, respectively, whilst the outlet boundary is the plane facing to the ram combustor downstream. The rest is solid and viscous sidewall surfaces of the ram inlet/bypass duct and the rear mixing region. The total pressure and tem­perature at each of the inlet planes, as well as the static pressure at the outlet boundary were determined from the transient engine thermo-cycle simulation results. Table 2 gives a list of those parameter values corresponding to CASEs 1-5 for the present fbw pattern visualization.

Results

Figure 5 is a summary result of the flow pattern simulation, which shows the iso-contours of Mach number (left) and static pressure (right) in the com­putational domain.

CASE1: During the turbo mode, MSV is firmly closed, and CCE is working just as a turbo fan engine. In this status the opening area of FVABI is large and the cross-sectional stream area of fan exit fbw is continued to increase in the bypass duct following FVABI, with the result of static pressure rise due to fbw diffusion. It becomes necessary, therefore, that the opening area of RVABI be kept small to accelerate the bypass duct flow to reduce its static pressure for a balance with that of the core flow meeting at the rear mixing region. It causes a particularly large velocity gradient in the radial direction, leading to non-uniformity at the ram combustor interface.

CASE2: When the ram combustor is ignited, it is predicted by the engine thermo-cycle performance analysis that the static pressure in the ram combus­tor becomes raised a little bit, wherefrom transmitted to the upstream steadily. Presently, no such instantaneous pressure sign can be expected traceable, but the pressure level in the bypass duct showed slight elevation. It must be rather pointed out that there is very little difference in the flow patterns between CASE1 and CASE2, which indicates continued flow pattern stability against the ram combustor ignition without giving a fatal damage to the engine opera­tion.

CASE3: To open MSV for transition into the ram combustion mode, the static pressure in the bypass duct should be decreased in order to prevent the recirculation phenomenon. The latter is explained that the fan exit flow through FVABI be reversed upstream as MSV is opened. To avoid such a dangerous phenomenon, a precaution is taken to reduce the opening area of FVABI and to increase that of RVABI. The latter operation is necessary for the pressure balance between the bypass duct and the core engine exhaust. The static pres­sure at FVABI exit becomes lower enough than the total pressure in front of MSV inlet, now ready to open MSV. It is thus important that the static pressure balance in the ram inlet duct be carefully checked at any time during the mode transition. CFD results yielded rather unsatisfactory flow pattern in the bypass duct that the main stream is seen to separate from the hub wall in the region near RVABI inlet to form a backlbw in the bypass duct. As a result, two big vortexes are observed to indicate large ft>w inefficiency. Several reasons for the latter might be explained, for instance, that the back pressure of the RVABI is fixed according to the fight altitude, nozzle area ratio and ram combustion status, and that the downstream of the FVABI tends to be little diffused in the bypass duct just because of the narrowed opening area of FVABI. It is apparent that the predicted fan exit flow area be much less than the actual one, so it has to be rapidly diffused into the whole duct in conformity with the higher back pressure at RVABI, leading to a back ft>w in the rear part. There seemed to be another reason that the slope of the hub wall near RVABI is steep, which can cause the ft>w separation. Since no turbulence model was introduced in the present CFD calculation, no further discussions may be available, but it be­came clear that there is unfavorable tendency of flow separation in the bypass duct due to a lack of flow diffusion.

CASEs4-5: MSV is gradually opened, so air is introduced into the ram inlet duct. In CASE4 (MSV opening area is small), there appears also a separation like in CASE3, however, in CASE5 (MSV opening area is large), there is no disturbance near RVABI and the flow becomes very smooth. That means the introduction of a ft>w in the shroud wall region yields favourable ft>w diffusion in the whole bypass duct, whence it may be concluded that the flow instability in the bypass duct is influenced much due to the degree of flow diffusion.

To sum up, all the five cases representing control sequences during mode transition showed no sign of recirculation or fbw reversal towards MSV in the ram inlet duct, indicating a possible stable transition assignment according to the present variable geometry control procedure. The bypass air jet is seen to be more diffused in the ram air duct as MSV becomes open wider, yielding better and clean ft>w pattern downstream. In this way, the present CFD simu­lation method will provide a very helpful tool to predict the effects of variable geometry control upon CCE flow patterns.