**5. References**

204 Mass Transfer - Advanced Aspects

In conclusion, this fluctuation of cavitation size is almost cyclic, as observed previously in the analysis of the 2D unsteady cavitating flow. As can be seen in the figure, the cavitation length is maximal at the instants *t=135·tref* and *t=144·tref* on the blade **2** and at the instants

The period of the cavitation fluctuation is *Tcav=0.0675 s* and its frequency *fcav=14.8 Hz*. The fluctuations can be driven by the cavitation torch formed upstream of the inducer. The torch turns in the same direction, but with a lower rotational velocity of the inducer. So, the torch turns 1 time while the inducer turns 9 times, as can be observed on the Fig. 23. The unsteady

calculation corresponds about *10%* of the inducer performance drop curve. In this case, one cavitation fluctuation occurs each 22 inducer turns. Thus, the cavitation fluctuation period is

> σ*=0.043*.

Steady and unsteady numerical simulations were carried out in many configurations: first, in a venturi duct, next in two blades cascades with different characteristics, and then, in

In order to understand the cavitation behaviour in the inducers and to validate the steady results obtained numerically, many experimental tests were developed, in steady state, for

In general, a good agreement between experimental and predicted results was found for a range of flow rates and cavitation behaviour, i.e., overall performances, cavities sizes and cavities location, etc. This study shows that the optimal inducer design depends mainly on three criterions to choose the best design for the three studied inducers: The critical cavitation coefficients corresponding to 5% and 15% of head drop, the head and efficiency produced in cavitating and non-cavitating conditions, and the vibrations generated by the

The numerical results, in a simple geometry, suggest a strong interaction of the turbulence

b. Stable behaviour with non–symmetrical cavitation length, observed only for a two-

rates. Alternate blade cavitation was observed numerically for partial flow rates, when the *l/h* ratio was higher than about *65%*. This phenomenon was observed only on the two-blade inducer. Finally, the rotating cavitation was observed for lower flow rates, on both studied blades cascades. The calculations were carried out using RNG κ–ε model and RNG κ–ε

Unsteady numerical results showed three different mechanisms of cavitation instabilities: - Self–oscillation of the cavitation due to the interaction between the recirculation flow

blade inducer, according experimental observations (Tsujimoto, 2001) c. Cyclical unstable behaviour with non–symmetrical cavitation length. Symmetrical cavitations on all the blades were observed for high values of

σ

σ

*=0.051*, and then, for

σ

and flow rates, predicted

σ

and high flow

*=0.064*. Then, from this

*=0.043*. This last

σ

*t=131·tref* and *t=140·tref*, and *t=149·tref* on the blade **1**.

unsteady results, the calculations were realized for

*Tcav=0.0600 s* and its frequency *fcav=16.7 Hz*, for

**4. Conclusion** 

three axial inducers.

inducers operations.

and the unsteady cavitation.

the three studied inducers.

cavitating calculations were performed, in the first place, for

Cavitating flow in the blades cascades, for various values of

three types of cavitation behaviour on the blades cascade: a. Stable behaviour with symmetrical cavitation length,

modified model which provide different results.

and the cavity surface (intrinsic instability).


**9** 

*Taiwan* 

**Mass Transfer Within the Location** 

Jen-Horn Yang2, Yean-Ren Hwang3 and Chuan Li4

Jing-Chie Lin1, Ting-Kang Chang2,

**Where Micro Electroplating Takes Place** 

*1Institute of Materials Science and Engineering, National Central University, 2,3,4Department of Mechanical Engineering, National Central University,* 

Electroplating and electroforming are the two electrochemical processes extensively used in metal fabrication. Electroplating provides a thin metal film to bestow the surface with desired property such as abrasion and wear resistance, corrosion protection, lubricity and aesthetic qualities; electroforming leads to a deposition of metal skin onto a mandrel which is then removed and then the metal deposit was thickened to obtain precise fabrication of molds. Both the electrochemical processes are carried out in the bath where sufficient concentration of metal salt is supplied in presence of an electric field. The electrochemical kinetics is determined not only by the strength of electric field but also by the mass transport phenomenon of the electrochemical active ions. The electric field employed in the electroplating is relatively lower and the field distribution is homogeneous. In contrast, the electrical field exerted in electroforming seems to be much stronger and the field

In 1995, a novel localized electrochemical deposition (LECD) process was pioneered by Hunter [1] to fabricate three-dimensional (3D) metal microstructures. The LECD brings the electrochemical process to a new era. However, in the LECD process, the electrical field exerted at the electroplating site is super high and the distribution of field strength is ultra heterogeneous. The phenomenon of mass transport in such a strong field distributed in extremely heterogeneous is the case which we have never encountered in doing usual electrodeposition. In the process of micro electroplating, the site where LECD taking place was experimentally controlled to along the track guided with a microanode. Accordingly, the micro metallic features could be fabricated electrochemically along the motional track guided by the microanode [2]. Due to this fact, LECD was also named as microanode guided

electroplating (MAGE) process. The schematic diagram of MAGE is shown in Fig. 1.

A platinum wire (diameter in the range from 25 to 125 μm) was fixed coaxially, and cold mounted with epoxy resin in polymethylmethacrylate (PMMA) tube (inner and outer diameters are 3 and 5 mm, respectively) to expose a disk (25 ~ 125 μm in diameter) acting as the microanode. The micranode was driven to move by a stepping motor in an electroplating bath thus guiding the micro electroplating way according to the program built in the micro-CPU. The micoanode assembly and microanode was driven to move by a

**1. Introduction** 

distribution becomes less homogeneous.

Yakhot, V. & Orszag, A.S. (1986). *Renormalization group analysis of turbulence–Basic theory,* J. Scientific Computing. Vol. 1(1), pp. 3–51.

Wilcox, D.C. (1998). *Turbulence Modeling for CFD*, DCW industries. La Canada, California.
