0
Research Papers: Micro/Nanoscale Heat Transfer

Probing the Local Heat Transfer Coefficient of Water-Cooled Microchannels Using Time-Domain Thermoreflectance

[+] Author and Article Information
Mehrdad Mehrvand, Shawn A. Putnam

Department of Mechanical and
Aerospace Engineering,
University of Central Florida,
Orlando, FL 32816

Contributed by the Heat Transfer Division of ASME for publication in the JOURNAL OF HEAT TRANSFER. Manuscript received October 14, 2016; final manuscript received March 21, 2017; published online June 21, 2017. Assoc. Editor: C. A. Dorao.

J. Heat Transfer 139(11), 112403 (Jun 21, 2017) (12 pages) Paper No: HT-16-1665; doi: 10.1115/1.4036691 History: Received October 14, 2016; Revised March 21, 2017

The demands for increasingly smaller, more capable, and higher power density technologies have heightened the need for new methods to manage and characterize extreme heat fluxes. This work presents the use of an anisotropic version of the time-domain thermoreflectance (TDTR) technique to characterize the local heat transfer coefficient (HTC) of a water-cooled rectangular microchannel in a combined hot-spot heating and subcooled channel-flow configuration. Studies focused on room temperature, single-phase, degassed water flowing at an average velocity of ≈3.5 m/s in a ≈480 μm hydraulic diameter microchannel (e.g., Re ≈ 1850), where the TDTR pump heating laser induces a local heat flux of ≈900 W/cm2 in the center of the microchannel with a hot-spot area of ≈250 μm2. By using a differential TDTR measurement approach, we show that thermal effusivity distribution of the water coolant over the hot-spot is correlated to the single-phase convective heat transfer coefficient, where both the stagnant fluid (i.e., conduction and natural convection) and flowing fluid (i.e., forced convection) contributions are decoupled from each other. Our measurements of the local enhancement in the HTC over the hot-spot are in good agreement with established Nusselt number correlations. For example, our flow cooling results using a Ti metal wall support a maximum HTC enhancement via forced convection of ≈1060 ± 190 kW/m2 K, where the Nusselt number correlations predict ≈900 ± 150 kW/m2 K.

FIGURES IN THIS ARTICLE
<>
Copyright © 2017 by ASME
Your Session has timed out. Please sign back in to continue.

References

Figures

Grahic Jump Location
Fig. 1

Yearly trends in microelectronics (Reproduced with permission from Waldrop [3]. Copyright 2016 by Nature Publishing Group). (a) Number of transistors per chip, viz., Moore's law (black line). (b) Microprocessor clock speeds, where the plateau region signifies the “speed limit” implemented in 2004. (c) Hot-spot heat fluxes calculated via the transistor and clock-speed trends (i.e., a and b), a processor die area of 500 mm2, DARPA's goal of 20 pJ per (fl)op, and the indicated (fl)op efficiencies (90%, 98%, respectively).

Grahic Jump Location
Fig. 2

Overview of the microchannel flow-loop apparatus and TDTR measurement methodology: (a) syringe pump-based microchannel flow loop, (b) optical bench schematic of our two-tint TDTR measurement setup, (c) expanded, construction view of the microchannel sample stage (also depicted in (a) and (b)), and (d) schematic illustrations of both hydrodynamic BL growth (δh(x)) in a microchannel of height (H≈ 400 μm) and thermal BL growth (δth(x)) from a hot-spot in the metal-coated glass wall by the TDTR pump-probe lasers

Grahic Jump Location
Fig. 3

TDTR ratio data (symbols) and model predictions (lines) as a function of pump-probe delay-time for a Ti-coated FS glass window in thermal contact with nonflowing (stagnant) water or air in the microchannel (fmod  = 962 kHz)

Grahic Jump Location
Fig. 4

Predicted dependence of the TDTR ratio on (a) the thermal effusivity and (b) thermal diffusivity of the sample/fluid in thermal contact with a Ti-coated FS substrate (see Fig. 3). Predictions are provided for different materials (symbols) at two different pump-probe delay times, τd=100 ps and τd=3 ns (fmod = 962 kHz). The magnitude of the difference between the open (100 ps) and closed (3 ns) symbol data is indicative of the cooling rate of the Ti metal thin film.

Grahic Jump Location
Fig. 5

(a) Schematic illustration of the anisotropic TDTR method with a flowing fluid (not-to-scale), where Δx is the pump-probe offset, w is the pump beam waist, vavg is the average flow field velocity, and ℓth is the thermal penetration depth. (b) Probing upstream the pump-induced thermal BL. (c) Probing downstream (or within) the pump-induced thermal BL.

Grahic Jump Location
Fig. 6

Anisotropic TDTR measurements corresponding with heat conduction and natural convection of water and air in the microchannel (τd= 100 ps, fmod  = 962 kHz)

Grahic Jump Location
Fig. 7

(a) Schematic depiction of probing upstream (Δx/w<0) or downstream (Δx/w>0) the pump induced hot-spot in the microchannel. (b) Anisotropic TDTR measurements for Ti-coated glass with flowing or stagnant water in the microchannel. (c) Corresponding thermal effusivity of water (left axis) and HTC (right axis) based on our differential TDTR analysis scheme (τd= 100 ps, fmod= 962 kHz, w= 9.5 μm).

Grahic Jump Location
Fig. 8

(a) TDTR ratio data and (b) corresponding HTC data atzero pump-probe offset (Δx/w≅ 0) as a function of the waterflow rate in the microchannel (Ti heater/thermometer, fmod= 962 kHz, w=9.5 μm)

Grahic Jump Location
Fig. 9

(a) Anisotropic TDTR measurements for Hf80-coated glass with flowing or stagnant water in the microchannel and (b) corresponding thermal effusivity of water (left axis) and HTC (right axis) based on our differential TDTR analysis scheme (τd= 100 ps, fmod= 976 kHz, w= 8.7 μm)

Grahic Jump Location
Fig. 10

(a) Schematic of probing upstream (Δx/w<0) or downstream (Δx/w>0) the pump induced hot-spot in the microchannel, where the dotted lines represent the flow-induced anisotropic metal wall temperature. (b) Comparison between the measured (symbols) and predicted (lines) enhancement in the local HTC due to forced convection over the hot-spot in the microchannel. The filled circles are for Ti/FS, whereas the open circles are for Hf80/FS (see Figs. 7 and 9, respectively).

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In