**8. Conclusions**

498 Mechanical Engineering

(a) t=1/20000(s)

(b) t=2/20000(s)

(c) t=3/20000(s)

(d) t=4/20000(s)

(e) t=5/20000(s)

(a) t=1/10000(s)

(b) t=2/10000(s)

(c) t=3/10000(s)

(d) t=4/10000(s)

Fig. 19. Cavitation Bubbles at the Position B (valve opening= 45 , 51.7 ).

Fig. 18. Cavitation Bubbles at the Position A (valve opening= 45 , 48.6 ).

The ability of fins to reduce the cavitation noise around a butterfly valve was investigated in this study. When two semicircular fins were attached to the downstream side of a valve body (TYPE-B), the experimental results indicated that the inception of cavitation was earlier than that found with a normal valve. This was discovered via comparison of noise given a constant pressure loss coefficient. However, cavitation noise increased gradually after inception and the maximum value of cavitation noise just before flashing was shown to be suppressed for the TYPE-B valve by the fins. Cavitation noise was reduced by about 5 dB when a valve opening of 45degrees was used. The effect of noise reduction around the TYPE-B valve was most remarkable just before flashing. In this study, it was found that fins are an adequate method for the reduction of cavitation noise. The optimum size and position of the fins should be investigated in future studies.

Visualizations created by a high-speed camera showed that intense vortex cavitation clouds were not clear in the TYPE-B valve with two fins. The interference of the flow from the orifice side with the flow from the nozzle side seemed to be suppressed by the fins. This was also confirmed by numerical analysis.

As for the visualization of cavaitation bubbles by using high-speed camera, it was found in this study that cavitation bubbles occur at the position before the leading edge and that the cavitation bubbles grow larger just behind the leading edge. It was also found that the maximum diameter occurs near the leading edge and that the bubbles become smaller through the pressure recovery of the flow. In this observation, the maximum diameter of the bubbles was 500 m near the leading edge. The diameter of the bubbles ranged from 20 m

**Part 4** 

**Simulations and Computer Applications** 

to 200 m at position B and position C. It is considered that the cavitation occurrence is dominant at the leading edge by the contraction flow and that the bubbles which occur at the leading edge become smaller in size through pressure recovery.

As for the effect of velocity distribution, it was clarified in this study that the cavitation noise around a butterfly valve becomes larger when the upstream velocity distribution was different from the normal velocity distribution. It was clear that the interference of the flow from the orifice side and the flow from the nozzle side was suppressed by the fins under not only the normal velocity distribution but also under the biased velocity distribution. The cavitation noise of Type-B was smaller than that of the normal valve in each upstream velocity distribution. Moreover, from the visualization results, it was found that the cavitation bubble diameters ranged from about 20 m to about 200 m and that the numbers of cavitation bubbles in Type-B was less than that of the normal valve.
