2 replies to this topic

[carbon60]

Hello all,

I am currently a senior ChemE on internship and one of my assignments is to calculate the evaporation rate of water in an ultrasonic cleaning station.

The knowns are:

• Dimensions of the water basin
• The relative humidity of air blowing across the basin
• The temperature of the air (assumed to be ambient)
• The flow rate of air over the basin
• the temperature of the water bath

my first thought was to use a mass transfer/diffusion method assuming that the basins were flat plates and accounting for the transition from laminar to turbulent flow.

In this case the mean mass transfer coefficient would be

K_c=(0.664(D_ab)(Re_t)^1/2Sc^1/3+0.0365(D_ab)Sc^1/3(Re_L^4/5-Re_t^4/5))/L

and the average flux would be

W_A=K_c(P^*_A-P_Abulk)


Where

• Kc=mean mass transfer coeff
• D_ab=diffusion coefficent
• Re_t = the transition Reynolds number
• Re_L=the Reynolds number at the end of the pan
• L=critical length of basin
• Wa= average flux leaving the surface
• P^*_A=vapor pressure of water at the basin temperature
• P_Abulk = partial pressure of water in Air

However as I thought about it more I realized that the mass transfer correlation would most likely not hold up under the ultrasonic conditions since the "wavy" motion of the water would greatly enhance the mass transfer for a number of reasons (in my mind there would be more surface area, more energy supplied to the water, and I guess more turbulence at the interface of air/water).

I was wondering if someone could point me a the right direction for either a better mass transfer coefficient correlation that would apply to ultrasonic flow over a flat plate, or a better method that I sound be using.


Thanks a lot for all your help

[kkala]

I do not have much knowledge on the subject, apart from that ultra sounds (US) seem to promote boemite precipitation in Bayer liquor (alumina process), as indicated by lab experiments (~1997). The lab tube of liquor accepted US from up downwards intermittently. My understanding was that US created stationary waves in the liquor and heat was produced, since power of US passed into liquor.
Probably you could also consider power given by US to the water, which is converted into heat and must be more or less known. According to my understanding water bath temperature remains constant (steady state), so water vapors entrained by air are actually vaporized by the US power given. This could define rate of water vapor transferred into air, a supplementary relation to your equations.
As if I had heard (1997) of potential liquid surface disturbance on US application in industrial scale (it was not implemented), apparently this problem did not exist in lab tubes. But (to my understanding) US power would define water vapor transfer into air irrespectively of surface waving.
One organization dealing with US applications is "Instituto de Acustica", Madrid, Spain, http://www.ia.csic.es/?Lang=EN, probably they can supply some information on the matter. European Acoustics Association http://www.acta-acus...iation-eaa.html could also help.

Edited by kkala, 10 October 2011 - 03:19 AM.

[carbon60]

Kkala,

Thank you very much for your response. I think that you are right that by combining the energy transfer of the U/S wave into the water with the relation of mass transfer of water into air it should prove a more accurate approximation. When I do this I get what I think seems like a reasonable number.

Thank you again for your reply, it was very helpful