the experiments, as shown in Figure 5. These are called (a) “Original Type Impeller” taking Figure 4 as it is, (b) “Alternate Type Impeller” where the leading and the trailing edge positions of the original blade are alternated in the tangential direction, and (c) “Reverse Type Impeller” reversing Original Type Impeller. The blade setting angle were also adjusted between ±7 degrees from the normal dimensions given in Figures 4 and 5 (the adjusting angle is shown by Δβ and the positive value is

in the clockwise).

3. Effects of Impeller Profile

Figure 6 shows effects of the blade setting angle on the fan efficiency with Long Tail Hub, where η is the jet fan efficiency [=QΔP_T/L, Q: the discharge, ΔP_T: the difference of total pressure between Sections S1 and S6, L: the shaft power]. Original Type Impeller takes the maximum efficiency, and the efficiency of Alternate Type Impeller is between those of Original and Reverse Type Impellers. Besides, the impellers with the normal dimension, Δβ = 0 degree, have the maximum efficiency regardless of the impeller types.

To know the acoustic noise from the model jet fan, the microphone of the sound-level meter was set at 1.5 m front of the inlet on the centerline of the shaft as shown in Figure 1. The measured data was accumulated momentarily in the computer and the characteristics were processed by the FFT analyzer. The level and the frequency of the acoustic noise were evaluated without the background noises emitted from the motor, pulley system and the ambient circumstances. To normalize these noises, the specific sound level L_SA [=SPL – 10log + 20] is shown in Figure 7. The increase of the blade adjusting angle, namely the impeller giving the higher head and/or discharge, has the acceptable effect to suppress comparatively the noise. Reverse Type Impeller generates the higher noise than the other type impellers, because not only the cusp leading edge causes the stall on the large scale but also the blunt trailing edge brings markedly the velocity defect. Alternate Type Impeller can reduce the effects of the stall at the leading edge and the wake flow from the trailing edge, but the noise is unfortunately larger than Original Type Impeller, as it is.

Figure 8 shows the effects of the hub profiles on the jet fan efficiency η at Δβ = 0. The efficiency is better with the moderate increase of the tail corn length. It can hardly ask the suddenly enlargement hub to take the high efficiency.

Figure 9 shows the specific sound level L_SA derived from Figure 8. The level seems to be in counter propor-

tion to the efficiency in Figure 8, as the flow separation increases the mixing losses in the downstream passage

and promotes the noise.

4. Effects of Air-Cooling Fan and Vanes

The air-cooling fin may work partly to turn the flow near the hub wall to the axial direction as if the guide vane, but such an effect is tiny. Table 1" target="_self"> Table 1 shows the effects of the air-cooling fin on the jet fan efficiency η and the specific sound level L_SA at Δβ = 0. The fin scarcely deteriorates the efficiency but makes the specific sound level increase more or less.

The effects of the stay vanes as shown in Figure 2, which are the flat plates with rounded leading and trailing edges, on h and L_SA are shown in Table 2 while Δβ = 0 with Long Tail Hub. The stay vane causes the separation on the large scale at the leading edge and disturbs the flow discharged from the impeller. Then, it is desirable to equip with the cambered vanes or the slender bodies like circular cylinders in place of the flat plates.

5. Modification of Blade Profile

5.1. Numerical Simulation

The blade profile was modified on trial with the numeri-

cal simulations by the commercial CFD code of ANSYS-CFX ver.12 with SST turbulent model, to provide the single stage impeller for the bidirectional type jet fan. The domain of the simulation is shown in Figure 10. The domain was divided into the rotational domain with the impeller, upstream and downstream stationary domains. The number of grid is about 800,000 for the rotating domain and about 800,000 for the stationary domains. These domains were connected with the frozen-rotor interface, while the boundary conditions are at the inlet and the constant static pressure at the outlet.

5.2. Flow Conditions around Original Impeller

Figure 11 shows the axial V_A and the swirling velocities V_Q at Sections S3 and S4, where the velocities are divided by the impeller tip speed u_T, R is the dimensionless radius with Long Tail Hub divided by the duct diameter. The commercial CFD code may simulate the flow around the impeller. Table 3" target="_self"> Table 3 shows the fan efficiency h and the average axial flow velocity V_AM (divided by u_T) at Section S6. The predicted results are in agreement with the experimental results.

5.3. Blade Profile Suitable for Bi-directions

The blade suitable for the bi-direction was designed to get the axial velocity V_AM as the same as V_AM of Original Type Impeller at the same rotational speed, with the aid of the numerical simulation whose reliability was verified above, and shown in Figure 12 (call Modified Impeller). The blade has no camber and the same profile at the leading and the tailing edges. Figure 13 shows the partial head H in the radial direction in comparison with H of Original Type Impeller, where H_th is the theoretical head averaged in the cross section of Original Type Impeller. The blade twists to increase the head in nearly proportion to the radius, though Original Type Impeller works mainly at the larger radius. Figures 14 and 15

show the effects of the blade adjusting angle Δβ on the average axial velocity V_AM and the fan efficiency η, in comparison with the values of the tandem impellers with Original Type Impeller (the bidirectional jet fan, call

Tandem Impeller). Tandem Impeller takes the maximum velocity V_AM regardless of the rotational speed. As for Modified Impeller, the normal adjusting angle, Δβ = 0, gives the higher velocity with the higher efficiency h as compared with one of Tandem Impeller. Modified Impeller with Δβ = 0 satisfies the rated velocity, V_AM = 35 m/s, at the rotational speed n = 1816 min⁻¹.

6. Concluding Remarks

To simplify the jet fan profile for reducing the initial cost without the performance deteriorations, the effects of the hub profiles on the performances and the flow conditions were investigated experimentally and numerically, using three type impellers/cascades. The efficiency at the reverse rotation of the original impeller, namely at the flow condition running from the outlet to the inlet, is doubtlessly lower than one at the original rotation. The unique cascade, where the leading and the trailing edges of the blade are alternated in the tangential direction, was prepared in anticipation of improving the performances, and gives the expected results. The hub with the long tail corn takes the best efficiency, but the stay vanes deteriorate the efficiency due to the flow separation at the leading edge. On the contrary, the air-cooling fins scarcely affect the efficiency and the flow condition. Based on these discussions, the blade profile was modified so as to take the acceptable performances.

7. Acknowledgements

The authors wish to thank Mr. S. Harada graduated from Kyushu Institute of Technology for helping the experiments, Mr. H. Kawakami and Mr. Y. Mitsuzuka of Fuji Electric Systems Co. Ltd. for fruitful discussions about the prototypes. Some parts of this research were thankfully co-sponsored by the 29th Harada Memorial Foundation.

REFERENCES