Singlephase ambient and cryogenic temperature heat transfer coefficients
Single-phase ambient and cryogenic temperature heat transfer coefficients in microchannel Seungwhan Baek and Peter E Bradley NIST Boulder, CO 80305 USA Cryogenic Engineering Conference 2015 06 -29 -2015 11: 45 AM Material Measurement Laboratory
Contents • Introduction • Experiments • Results • Discussion • Summary Material Measurement Laboratory 2
Introduction • Microscale J-T cryocooler development – Requires microchannel heat exchanger – Requires microchannel heat transfer characteristics Revised MCC Design Early prototype Compressor J-T valve Evaporator Microchannel Heat exchanger Material Measurement Laboratory Microchannel Heat exchanger 3
Thermal design of MCC heat exchanger • Operating condition of MCC – Fluid flow in microchannel ( Dh < 100 μm ) – Extremely low flowrate ( Re < 100 ) – Low pressure ratio (Pr<4: 1) • Cooling Performance of MCC depends on – Heat exchanger performance – Need heat exchanger/heat transfer characteristics for MCC operating condition – No previous research at these condition) Recuperative HX • Two different heat exchanger in MCC – Recuperative HX: Single phase (gas or liquid) • Isothermal HX Phase I research – Isothermal HX: two-phase (gas+liquid) • Phase 2 research Dh: hydraulic diameter, Re: Reynolds number Material Measurement Laboratory 4
Thermal design of heat exchanger • W L Flo el ann h w c ion th ect ir w d Flo H ll Wa Material Measurement Laboratory h: heat transfer coefficient (W/m 2 K) kf: thermal conductivity of fluid (W/m. K Dh: hydraulic diameter (m) Re: Reynolds number Pr: Prandtl number 5
Nusselt number: Previous research • Theoretical Nusselt number: Nu=4. 36 const. (circular channel, Re < 2000) • From previous research: single-phase Nu # correlation for Re < 2000 • • No experiment at low temperature (T < 200 K), low flow rate (Re = 10 - 50) Experiments do not follow theory, Differ from each other, No recent research • Need to verify heat transfer characteristic for better design of MCC From previous research, Nu=f(Re, Pr) for Re < 2000 Researcher Diameter Kays (1970) Dh > 1 mm Sieder & Tate (1996) Dh > 1 mm Shah and London (1978) Dh > 1 mm Wu & Little (1983) Dh = 150 μm Choi et al. (1991) Dh = 81 μm Grigull and Tratz (1965) Dh > 1 mm Correlation Material Measurement Laboratory Previous Nu# for single phase flow (gas) MCC Operating regime Laminar Nu=4. 36 Review of heat transfer and pressure drop characteristics of 6 single and two-phase microchannels, Asadi, 2014, (Review paper) Turbulent
Experiments Material Measurement Laboratory 7
Experiments • Measure the single phase heat transfer coefficients • Nitrogen • microchannels (180 μm, 110 μm, 65 μm) • Operating condition • cryogenic liquid flow (~70 K) • ambient gas flow(~300 K) 180 μm 110 μm Friction factor 300 K gas N 2 Heat transfer Coefficient 70 K Liquid N 2 300 K Gas N 2 Material Measurement Laboratory 65 μm 8
Experimental setup (schematic) Vacuum Chamber ‘Not to scale’ Pressure sensor Vacuum chamber Constant heat flux 2 nd stage Microchannel Tinlet 1 st stage GM cryocooler Twall_1 Twall_2 Twall_3 Toutlet Feedthrough collar + Test section Radiation shield T=64 K Recuperator Flow conditioner Mass flow meter GM-cryocooler 2 nd stage Compressor Microchannel assembly GM-cryocooler 1 st stage Material Measurement Laboratory Compressor 9
Microchannel test section L/4 microchannel L L/2 Heating element Thermal grease Flow ‘Not to scale’ Solder Epoxy Tin Setup for Dh=180 μm Tw 1 Tw 2 Tw 3 Tout thermocouple Unit: mm 180 μm 36 AWG Do=130 μm E-type thermocouples Heating length=3 cm Material Measurement Laboratory 380 μm Heating wire Do=160 μm 380 μm 10
SEM pictures of microchannels Identical magnification Dr. Baek’s hair cross section Not typical human 100 μm 139 μm Dh=180 μm Dh=110 μm Dh=65 μm Din (μm) 180 110 65 Dout (μm) 380 310 160 Din/Dout 0. 47 0. 35 0. 4 Thinnest wall Material Measurement Laboratory Thickest wall 11
Scale comparison air H ek m a B μ μm Dr. ~100 OD=160 OD Dr. Baek hair 110 μm 310 μm Stainless steel tube Dh= 110 μm 10 3 = D O 65 μm 160 μm Stainless steel tube Dh= 65 μm 36 AW 106 μm G 300 μm Thermocouple tip Not able to measure fluid temperature and wall temperature separately. Material Measurement Laboratory 12
‘Classic’ Nu# estimation method heater L Constant heat flux Microchannel x Tinlet Twall, x 1. Find energy input to fluid 2. Estimate fluid temperature inside microchannel (based on linear temperature profile) 3. Measure the wall temperature 4. Determine the heat transfer coefficient 5. Calculate the Nusselt number Material Measurement Laboratory Toutlet 13
Result & Discussion Material Measurement Laboratory 14
Friction factor measurements (Gas, 300 K) • Hydraulic characteristic of fluid in microchannels – Friction factor follows conventional theory Experimental friction factor Laminar theory friction factor Turbulent theory friction factor (Blasius equation) Material Measurement Laboratory Turbulent Laminar 15
Nu # measurement Gas N 2 (300 K) • • • Liquid N 2 (70 K ) Nu # degrades from Re < 1000 Nu #: 180 μm > 65 μm > 110 μm Similar trend with previous research (Morini, Choi, Little) • • • Nu # degrades from Re > 200 Nu # : 180 μm ~= 65 μm ~= 110 μm No other research to compare Nu=4. 36 Laminar Material Measurement Laboratory Turbulent 16 Turbulent
Scaling effect • Scaling effect can influence thermal behavior of fluid flow in microchannels* Non-D number Effect Ignored when Kn Knudsen Ma Mach Brinkman λ Lambda axial conduction of wall λ < 0. 01 Pe Peclet axial conduction of fluid Pe > 50 Microchannel Phase Re Dh=180 μm Dh=110 μm Dh=65 μm Kn (<0. 001) gas rarefaction Kn < 0. 001 flow compressibility Ma < 0. 3 viscous heating Br < 0. 005 Ma (<0. 3) Re=1 Re=3000 Br (<0. 005) λ (<0. 01) Pe (>50) Re=1 Re=3000 gas 0. 00007 0. 0006 0. 10 2. 18 0. 005 10 2340 liquid 0. 00001 0. 0003 0. 04 0. 10 0. 0002 3 200 gas 0. 00012 0. 0018 0. 27 1. 63 0. 006 10 2400 liquid 0. 00002 0. 0007 0. 11 0. 80 0. 0002 5 200 gas 0. 00021 0. 0031 0. 46 1. 10 0. 002 8 2430 liquid 0. 00005 0. 0010 0. 20 0. 0004 5 2700 *Guo Z-Y and Li Z-X, 2003 International Journal of Heat and Mass Transfer 46 (1) 149 -159 Material Measurement Laboratory 17
Nu # degradation • • Nu # degradation is related to axial conduction effect through the wall. Axial conduction changes temperature profile to ‘non-linear’. (Baek et al, 2014) Non-linear temperature profile violates the assumption in classic Nu # measurement. Classic Nu # measurement including axial conduction effect leads to estimate ‘apparent Nu #’. • Apparent Nu # is neither actual nor theoretical Nu #. • Apparent Nu # (Lin & Kandlikar, 2012) – Nu # degrades due to axial conduction effect with classic Nu# measurement method (1) Baek et al. , 2014, Cryogenics 60 49 -61 Lin T-Y and Kandlikar S G, 2012 Journal of Heat Transfer 134 (2) 020902 Material Measurement Laboratory 18
Comparison of Experiment & Nuapp • Comparison shows identical trend with experiment & equation (1). • Comparison implies actual Nu=4. 36 holds in low Re # flow. Material Measurement Laboratory 19
Summary Material Measurement Laboratory 20
Summary • Design of heat exchangers influence development of MCC – High uncertainty in operation Due to very small Dh, Low Re# , Low temperature • The hydraulic and thermal characteristics of fluid in the microchannel are investigated by the experiments. – Friction factors : comparable to macro-scale tubes – Nu # : decreased value @ low Re# , which are affected by axial conduction • Axial conduction effect influence the fluid & wall temperature profile to become non-linear. • Comparison of experimental result and theoretically derived Nuapp imply validation of Nu = 4. 36 In laminar flow for single-phase fluid. Material Measurement Laboratory 21
Experiment vs Previous work Material Measurement Laboratory 22
Thank you! Test Specimen is available! Ask Peter & Seungwhan for observation! Material Measurement Laboratory 23
- Slides: 23