Slide Show, Microscale Phase Change Heat Transfer to Water

This transformative presentation takes 10 minutes -- or less.  It was ahead of its time over 20 years ago and it still is.

This paper is Nukiyama in microscale over the pressure range from 200 to 6000 PSI.  Nukiyama used a platinum wire simultaneously as a heat transfer element and a resistance thermometer. 

Here is my Nukiyama element: diameter 7.5 microns, length 1.15 millimeters.

It is installed in the lower end of an upright tube that is filled with pressurized water, pure and degasified at about 20 Centigrade.  The steel tube is used for these pool boilng runs at higher pressures.

The benchtop apparatus has the pool boiling tube, the pressurizing pump, the junction box and programmed power supply, a PC with Excel for data logging, and back here, a Rosemount recording pressure transducer.

A run takes about 11 seconds and data is recorded every tenth of a second.  Heat flux is smoothly increased and then decreased.  In this run at 1000 PSI with the water at about 20 Centigrade, the peak was over 4000 Watts per square centimeter.  The average power is about one half watt, so the energy is about 6 watt-seconds, or less than two calories for the 11 second run.  So, the temperature of a local gram of water increases by less than two Centigrade.

This is the reduced data for the 1000 PSI run.  It doesn’t get any better than this although it gets more exciting.  The transition to phase change, I call it the Nukiyama point, is at about 2100 watts per square centimeter.  This transition is extremely sharp, there is no sluggishness.  That is also the case at the turnaround at 4100.  I want to increase the recording rate to every millisecond in further runs.  Note that the return is on the same path as the rise.  Clearly, the local temperature stays very close to 20 Centigrade, otherwise the return would not overlay. 

Here are several runs, three subcritical, one transcritical, and four supercritical. You have seen the dissection of the 1000 psi run; the 200 and 2500 of this plot are similar.  The dotted line is the transcritical run at 3000.  The four supercritical runs range from 3600 to 6000.

Here are 3600 and 6000 runs.  The 6000 has a narrower spread between the power up and the power down.

The 3,000 transcritical run is superimposed on the runs you have just seen.  I want to expand this territory with runs having millisecond data collection.

The supercritical runs are all well-behaved in this plot.  Even the transcritical falls in its proper place.

The point of this plot of the subcritical runs is that at 200 psi the Nukiyama point is 65 Centigrade beyond saturation while at 2500 and above, that delta T  is very narrow.

With the permission of Professor Victor Yagov of Moscow, I am presenting his recent analysis.At 200 psi he reports substantial Delta T’s, at 3000, the Delta T’s are vanishing.  Multiply the heat fluxes in this table by 100 to get to my units.

This is the transcritical run.  Again, I want millisecond data recording next time in the search for the aspects of the jump.  Note that saturation temperature is very close to critical.

Here is limited data.  With a somewhat larger microscale element the heat fluxes at the Nukiyama points are reduced, however the temperature of the points is unchanged.  The transcritical characteristic is the same even though the pressure is 100 psi less; the saturation temperature is still very near critical.

I think this is sensational.  The procedure is to pressurize the apparatus to about 6000 psi, apply a substantial heat flux in one step, and maintain that heat flux as pressure is smoothly reduced over about 20 seconds, turn off power at about 200 psi.  Note the gradual increase in temperature as pressure is reduced, then the upward jump of over 200 Centigrade at around 3700 psi, followed by a smooth increase of another 100 Centigrade to the critical pressure, then a smooth decrease of nearly 200 Centigrade, then a downward jump of about 150 centigrade to the critical temperature. At that point the subcritical boiling begins and continues until the power is turned off. At the lowest heat flux there is a smooth continuous plot over the entire pressure range with no intervening steps. The four plots at increasing heat fluxes are neatly nested.

Here are the data points.  I captured one point during the upward jump in two runs. I now propose runs with millisecond recording, an expanded range of heat fluxes, precise control of assorted rates of pressure changes, etc.

Here are the plots of subcritical phase change only.  The plots are very close; however the heat fluxes are distinct.  As was measured in the constant pressure runs at subcritical pressures and also calculated by Professor Yagov, the delta T from phase change to saturation increases as the pressure is reduced.

This shows the impact of nitrogen saturation at 1000 psi.  I never reduced these plots to heat flux and temperature; I did not need to in order to get the patent. Nitrogen reduces the heat transfer coefficient during natural circulation; however phase change heat transfer begins at a somewhat lower temperature and heat flux.  I filed this patent during 1995 and it issued during 1997, the same year that Berkeley started its microscale journal. 

If you GOOGLE “Lienhard Boiling” you will find his 1971 paper “Boiling from Small Cylinders”.  There are experts who still believe that stuff.

. 

The abstract on page 2011 includes the assertion, “Nucleate boiling does not occur on the small wires.”  It should add, or wires of any size at 3” Hg.

Of course, it would be a relatively easy experiment to deploy those wires at higher pressures in order to reveal a transition to nucleate boiling.

I’ll keep this brief.  I’ve already emphasized the need for millisecond recording and better control of the constant power runs at varying pressure.  The determination of circulation patterns will be tough, and I am interested in that for reasons of my own.

I drag this stuff out about every 10 years.  Millie, Gang and Arun were there.