**5. Effect of the fins on cavitation noise reduction (Comparison with constant pressure loss coefficient)**

The results shown in Fig. 3 were obtained by measuring cavitation noise at each flow velocity, while flow velocity was increased using a pump controlled by an inverter. The upstream pressure was kept at atmospheric pressure in each experiment, and the effect of the saturated steam pressure was minimal. Therefore, the flow velocity was also almost the same when the cavitation number was same. However, in many cases, the flow rate was not controlled by changes in the frequency of the pump, and the flow rate was controlled by adjusting the opening of the valve. Additionally, the flow rate in the actual plant was determined by the head curve of the pump and the pressure loss of the piping system as s whole. Accordingly, the effect of cavitation control should also be investigated under the constant pressure loss coefficient condition.

When a fin is attached to a valve body, the flow rate may be changed by the variation in the pressure loss of the valve. Figure 4 shows the pressure loss coefficients for each type of valve. For an example, the pressure loss of TYPE-C was larger than that of the normal valve at the same Reynolds number and at the same valve opening. When the butterfly valve is used to control the flow rate, the valve opening of TYPE-C must be larger than that of the normal valve in order to obtain the same flow rate. Therefore, the experimental results must be compared with a constant pressure loss coefficient. The following discussion is based on a comparison of cavitation noise under the condition that the pressure loss coefficients of the test valves are nearly equal.

Fig. 4. Pressure Loss Coefficients for each Valve Type.

Noise Reduction in Butterfly Valve Cavitation

maximum sound pressure was suppressed.

even taking into account the accuracy of the data.

**6. The visualization of the effect of the fins** 

Fig. 7. Top Views of Cavitation Conditions.

**6.1 Visualization experiments** 

around almost all the edges of the valve body at about 25

by Semicircular Fins and Visualization of Cavitation Flow 489

level of the cavitation began to rise abruptly. Accordingly, the inception of cavitation for TYPE-B was earlier than that for the normal valve. However, the results shown in Fig.6 indicate that the maximum cavitation noise value was suppressed in TYPE-B by the fins. Cavitation noise was reduced by about 5 dB when . The noise reduction effect in TYPE-B was remarkable just before flashing condition. Though cavitation occurred a little earlier than with the normal valve, the increase in the noise was milder in TYPE-B and the

The accuracy of the data mentioned above was ±1-1.5 dB near the inception of cavitation and ±0.5dB during cavitation growth and flashing. Though the differences among the valves were slight as shown in Fig.6, the noise reduction effect was clear for TYPE-B in flashing

Figs. 7 and 8 show the photographs of the cavitation conditions of each valve taken with a high-speed camera. Figure 7 was taken through the upper surface of the transparent pipe (top view). Figure 8 was taken from the side surface (side view). These photographs were not taken simultaneously because only one camera was available to be used. As shown in Fig. 7(a), vortex cavitation clouds were clearly visible on the orifice side of the normal valve and 1 Dia and further downstream from the stem axis. These cavitation clouds were identified as vortices of an intensity which brings about the cavitation damage (Tani et al.,1991). Similar vortex cavitation also occurred in the TYPE-A valve with one fin. It was impossible to suppress flow interference using only a single fin. Moreover, it is obvious from Fig.7 (b) that the cavitation was further intensified since the fin existed in the part of the orifice side where the contraction flow was severe. However, such a vortex cavitation cloud is not clear around the TYPE-B valve as shown in Fig. 7(c). The interference of the flow from the orifice side with the flow from the nozzle side seemed to be suppressed by the fins. Accordingly, it can be said that the cavitation around the valve body in TYPE-B was

(a) Normal Valve (θ=45°,σ=17,ζ=8) (b) TYPE-A (θ=50°,σ=17,ζ=8)

(a) Normal Valve (θ=45°,σ=17,ζ=8) (b) TYPE-A (θ=50°,σ=17,ζ=8)

, where the sound pressure

Figure 5 shows the pressure loss coefficients for each valve when was about 7. The pressure loss coefficient was almost constant for each valve within the range of the Reynolds number of the experiment. However, when the Reynolds number was <sup>5</sup> Re 2.25 10 in the case of TYPE-C, the pressure loss coefficient began to increase and the noise began to decrease corresponding to flashing. According to the results shown in Fig.5, the noise levels around each valve were compared with a pressure loss coefficient of about 7.

Fig. 5. Pressure Loss Coefficients (ζ≒7).

Figure 6 shows the effect of the fins on cavitation noise. The pressure loss coefficient for Type-B was about 7 at the valve opening of 45° and almost the same as that for the normal valve at the valve opening of 45°. Cavitation began to occur at about 30 in the case of TYPE-B. On the contrary, in the case of the normal valve, light cavitation noise occurred

Fig. 6. Effect of Fins on Cavitation Noise under (ζ≒7).

pressure loss coefficient was almost constant for each valve within the range of the Reynolds number of the experiment. However, when the Reynolds number was <sup>5</sup> Re 2.25 10 in the case of TYPE-C, the pressure loss coefficient began to increase and the noise began to decrease corresponding to flashing. According to the results shown in Fig.5, the noise levels

> **Valve opening 45**°**Normal Valve opening 50**°**TYPE-A Valve opening 45**°**TYPE-B Valve opening 50**°**TYPE-C**

Figure 6 shows the effect of the fins on cavitation noise. The pressure loss coefficient for Type-B was about 7 at the valve opening of 45° and almost the same as that for the normal

1 2

Reynolds Number (Re)

TYPE-B. On the contrary, in the case of the normal valve, light cavitation noise occurred

**Valve opening 45**°**Normal Valve opening 50**°**TYPE-A Valve opening 45**°**TYPE-B Valve opening 50**°**TYPE-C**

10 20 30

Cavitation Number σ

valve at the valve opening of 45°. Cavitation began to occur at about 30

[105 ]

in the case of

was about 7. The

Figure 5 shows the pressure loss coefficients for each valve when

Fig. 5. Pressure Loss Coefficients (ζ≒7).

0

10

20

Pressure Loss Coefficient (

30

40

50

ζ)

Fig. 6. Effect of Fins on Cavitation Noise under (ζ≒7).

60

70

Sound Pressure Level (dB)

80

around each valve were compared with a pressure loss coefficient of about 7.

around almost all the edges of the valve body at about 25 , where the sound pressure level of the cavitation began to rise abruptly. Accordingly, the inception of cavitation for TYPE-B was earlier than that for the normal valve. However, the results shown in Fig.6 indicate that the maximum cavitation noise value was suppressed in TYPE-B by the fins. Cavitation noise was reduced by about 5 dB when . The noise reduction effect in TYPE-B was remarkable just before flashing condition. Though cavitation occurred a little earlier than with the normal valve, the increase in the noise was milder in TYPE-B and the maximum sound pressure was suppressed.

The accuracy of the data mentioned above was ±1-1.5 dB near the inception of cavitation and ±0.5dB during cavitation growth and flashing. Though the differences among the valves were slight as shown in Fig.6, the noise reduction effect was clear for TYPE-B in flashing even taking into account the accuracy of the data.
