|
Research Objectives |
|
Microchannels are increasingly being used to yield
compact geometries for heat transfer in wide variety of applications.
Two-phase flow in microchannels have received considerable attention due to
growing interest in high heat fluxes made possible by these micro-channel
geometries. Limited research has been done on flow regimes, and the
measurement of pressure drop and heat transfer during condensation. Heat
removal and rejection applications can also benefit from high heat flux
condensation. The ultimate rejection of large heat duties, in the
electronics cooling industry, through compact condensers has not been
addressed adequately.
Research
Objectives:
 |
To understand condensation flow phenomena and transitions for R134a in
mini-channel geometries |
|
To study heat transfer and pressure drop for R134a in micro-channel
geometries with 100 mm < Dh < 5 mm and understand the effect of
Dh and shape |
To develop experimentally validated,
flow-mechanism-based pressure drop and heat transfer models for each flow
regime for a wide range of operating conditions and mini-channel geometries. |
|
Minichannel Geometries |
|
Flow visualization
studies were conducted on the circular, rectangular and square tube
geometries of varying hydraulic diameter and aspect ratios as shown on the
left side. The flow pattern maps based on these studies were further used
to do determine the applicable flow regime for the pressure drop and heat
transfer data for a variety of circular and noncircular channels with 0.4 <
Dh < 5 mm as shown in center and right side. Experiments
were conducted for 0 < x < 1 and 150 < G < 750 kg/m2. |
|
 |
|
For pressure drop and heat
transfer experiments the test sections were fabricated as flat tubes with
multiple extruded parallel channels. Three such tubes were brazed together
with refrigerant flowing through the center tube, and coolant flowing in
counterflow through the top and bottom tubes. |
 |
|
Condensation Flow Mechanisms |
|
A novel
test facility that enabled visualization of the actual condensation process
(and not simulations using air-water mixtures) was developed. Essentially,
two-phase refrigerant of a desired inlet quality was supplied to a glass
test section with the cross-section under consideration. The glass section
was enclosed in an outer Plexiglas annulus through which compressed air
flowed. The compressed air decreased the differential pressure to be
withstood by the glass channel while also serving as a coolant that enabled
condensation. Air temperatures and flow rates were adjusted to allow a
small amount of condensation (0.05 < ∆x < 0.10) in the test section. With
this arrangement, condensation flow mechanisms were recorded using video
photography with fast shutter speeds across the entire vapor-liquid dome in
small increments (high resolution) of vapor quality.
|
|
 |
|
Effect of Geometry on Flow Regimes |
|
Extent of the intermittent flow regime increases
as the hydraulic diameter decreases, signifying an increasing influence of
surface tension at the small diameters. Also, the wavy flow regime
progressively decreases and disappears as the diameter decreases, giving way
to the annular flow regime, signifying a diminishing influence of
gravitational forces at the small diameters. The effect of changing tube
shapes (round, square, rectangular with different aspect ratios) was also
documented. Tube shape, however, was found to be less significant than
hydraulic diameter in determining the applicable condensation flow pattern. |
|
|
 |
|
Pressure Drop Modeling |
|
Pressure drop measurements on a multitude of
circular and noncircular tubes across the vapor-liquid dome were used, in
conjunction with the insights from the flow visualization studies, to
develop experimentally validated models for condensation pressure drop.
The intermittent flow pressure drop model
(also shown to apply to discrete-wave flow) treats the overall pressure drop
as a combination of the contributions due to the liquid slug, the
film-bubble interface region, and the transitions between the slug and the
bubble. A slug frequency model was used to provide closure to the
intermittent flow model. |
 |
|
In the annular flow pressure drop model
(also shown to apply to disperse-wave and mist flows), the interfacial
friction factor derived from the measured pressure drops was correlated in
terms of the corresponding liquid-phase Reynolds number and friction factor,
the Martinelli parameter, and a surface tension-related parameter.
The resulting model predicted 82% of the annular
flow pressure drop data within ±20%.
It was also shown that at the same mass flux, quality and L/D, the two-phase
pressure drop increases as the tube diameter decreases. |
|
|
 |
 |
|
Heat Transfer Coefficents |
|
Condensation heat transfer coefficients were
measured, for a variety of circular and noncircular channels with 0.4 < Dh
< 5 mm, using an innovative thermal amplification technique developed to
measure heat transfer rates at small scales accurately.
Thermal amplification technique decouples
the issue of measuring low heat duties and maintaining high resistance
ratios by using a combination of a high flow rate closed loop primary
coolant to establish high resistance ratio (Rrefg/Rcoolant)
and a low flow rate open loop secondary coolant to ensures the accurate
measurement of the small heat duties in these microchannels and the
deduction of condensation heat transfer coefficients from measured UA
values. Low secondary coolant flow rate ensures high
DT, low uncertainty in Q.
A model for predicting heat transfer during
condensation of refrigerant R134a in three horizontal circular microchannels
(0.5 < Dh < 1.5 mm) over the mass flux range 150 < G < 750 kg/m2-s
was developed. Results from previous work on condensation flow mechanisms in
microchannel geometries were used to interpret the results based on the
applicable flow regimes. The utilization of the appropriate existing shear
stress models developed specifically for microchannels in the shear driven
heat transfer models is able to account for the related heat transfer
phenomena accurately. The heat transfer coefficient increases with an
increase in mass flux and quality, and with a decrease in the tube diameter.
|
 |
|
 |
|
 |
|
Fabrication of Channels With 100<Dh<200 µm |
|
The fabrication process starts with the
fabrication of an X-ray mask, which is subsequently used to pattern a PMMA
mold over a copper wafer. Copper is then deposited into the mold using the
process of electroforming followed by planarization. Holes for refrigerant
inlet and exit are then drilled into each of the refrigerant channel sets.
Figure on left shows the layout of the mask used for X-ray lithography. As
shown in the layout, 7 different devices (each set of channels, for example
twenty 0.1 ´ 0.1 mm channels, is
referred to as device) are obtained. All devices have constant depth of 100
mm, but varying number of channels
and widths to test the effect of aspect ratio. Figure on right shows picture
of the developed refrigerant channels after the removal of the PMMA mask.
Open header area and the location of inlet and exit ports allow for the
distribution of refrigerant into various parallel channels. |
|
|
 |
|
These channels are
then closed using the process of diffusion bonding with another copper wafer
on the top of these developed refrigerant channels. No adhesive is used in
this process, and bonding is performed in a vacuum oven
at 450oC for 3 hours. In this
process the bonded joint has almost bulk material like properties, allowing
the channels to withhold pressures as high as 10 MPa. |
 |
|
Experimental Setup |
|
The coolant, water flows through the water channel
blocks made out of copper shown in figure. Each water channels block has
five 1.5 cm long holes of 0.794 mm (1/32”) diameter drilled in them. These
water channel blocks are soldered to the refrigerant channel blocks as shown
in figure. Water flows in the counter flow direction.
A novel measurement technique is
developed to measure the very high heat transfer coefficients at these
microscales, especially due to the low heat transfer rates at extremely
small flow rates under consideration. Energy balances on pre- and
post-heaters are used to establish the refrigerant inlet and outlet states
at the test section. Cooling of the refrigerant in the test section is
accomplished using water at a high flow rate to ensure that the condensation
side presents the governing thermal resistance.
This method yields accurate measurements of heat
transfer coefficients for refrigerant R134a for 300 < G < 800 kg/m2-s
and 0 < x < 1 over a wide range of saturation temperatures (30 < T
< 60oC). The measured heat transfer coefficients and pressure
drops are being analyzed to develop models for pressure drop and heat
transfer during condensation at the microscales. The results from this work
will directly benefit the design of a variety of high heat flux heat
transfer applications at the microscales. |
|
|
|
Potential Applications |
|
A fundamental understanding of condensation at the
micro-scales will yield far reaching benefits not only for electronics
industries, but also for other as-yet untapped applications such as:
–
Portable personal cooling devices
–
Hazardous duty and high ambient air-conditioning
–
Medical/surgical devices
–
Fuel cells
–
Miniature heat pumps
–
Microchannel heat exchangers |
|
Sponsors |
|
This study was
supported by a research grant from Modine Manufacturing Company, Racine, WI,
and National Science Foundation. |
|
Papers |
|
1.
Garimella, S., A. Agarwal and J. D. Killion (2005), "Condensation
Pressure Drop in Circular Microchannels," Heat Transfer Engineering
Vol. 26(3) pp. 1-8.
2.
Garimella, S. (2004), "Condensation Flow Mechanisms in
Microchannels: Basis for Pressure Drop and Heat Transfer Models,"
Heat Transfer Engineering Vol. 25(3) pp. 104-116.
3.
Coleman, J. W. and S. Garimella (2003), "Two-Phase Flow Regimes
in Round, Square and Rectangular Tubes During Condensation of
Refrigerant R134a," International Journal of Refrigeration Vol.
26(1) pp. 117-128.
4.
Garimella, S., J. D. Killion and J. W. Coleman (2003), "An
Experimentally Validated Model for Two-Phase Pressure Drop in the
Intermittent Flow Regime for Noncircular Microchannels," Journal of
Fluids Engineering Vol. 125(5) pp. 887-894.
5.
Garimella, S., J. D. Killion and J. W. Coleman (2002), "An
Experimentally Validated Model for Two-Phase Pressure Drop in the
Intermittent Flow Regime for Circular Microchannels," Journal of
Fluids Engineering Vol. 124(1) pp. 205-214. |
| |
|
