TFAWS Active Thermal Paper Session Design of High

  • Slides: 26
Download presentation
TFAWS Active Thermal Paper Session Design of High Performance Gas-to-Fluid Heat Exchangers using Shape

TFAWS Active Thermal Paper Session Design of High Performance Gas-to-Fluid Heat Exchangers using Shape and Topology Optimization Daniel Bacellar, Vikrant Aute, Zhiwei Huang , Reinhard Radermacher Center for Environmental Energy Engineering Department of Mechanical Engineering, University of Maryland Presented By Daniel Bacellar Thermal & Fluids Analysis Workshop TFAWS 2016 August 1 -5, 2016 NASA Ames Research Center Mountain View, CA

Contents • • • 2 Introduction Objectives Background Problem Formulation Airside Modeling Proof-of-concept Design

Contents • • • 2 Introduction Objectives Background Problem Formulation Airside Modeling Proof-of-concept Design & Validation Numerical Optimization Results Conclusions

Introduction • Gas-to-Fluid air-to-water heat exchanger (HX) • Low pressure (~1. 0 atm), low

Introduction • Gas-to-Fluid air-to-water heat exchanger (HX) • Low pressure (~1. 0 atm), low Mach number (<<0. 3) – Airside thermal resistance: 75%~95% overall thermal resistance • Conventional HX require fins • High-performance, compact HX: (UA)↑, (A/V)↑ – (UA)↑ A↑ – Enhancing fin – (A/V)↑ A↑ + V ↑ • Fins drawbacks – – Temperature gradient Friction resistance Fouling/frosting sensitive Material (consumption, cost, weight) TFAWS 2016 – August 1 -5, 2016 3

Objectives • Present a multi-scale analysis and shape optimization method for novel HX’s –

Objectives • Present a multi-scale analysis and shape optimization method for novel HX’s – Enhance heat transfer on primary surfaces by reducing the characteristic length – Optimize shape (of reduced characteristic length surface) for better overall thermal-hydraulic performance • Minimize, or eliminate, the need of secondary surfaces (fins) • Demonstrate the validity of the method with a proof-of-concept design that was prototyped, tested and validated TFAWS 2016 – August 1 -5, 2016 4

BACKGROUND 5

BACKGROUND 5

Same refrigerant cross section area Next Generation of HX’s Airside surface to volume ratio

Same refrigerant cross section area Next Generation of HX’s Airside surface to volume ratio (cm²/cm³) First TIME Order Analysis Fluid volume (cm³) 2660 345 This Research Current Interest State-of-the-art Channel hydraulic diameter (mm) 6

Heat Transfer Enhancement Secondary Heat Transfer Surfaces Primary Heat Transfer Surfaces Why fins? Significant

Heat Transfer Enhancement Secondary Heat Transfer Surfaces Primary Heat Transfer Surfaces Why fins? Significant contribution to Compactness & Thermal Conductance for conventional sized tubes Boundary layer disruption and attachment regions on different fin types: a) Plain; b) Louver; c) Slit; d) Perforated; e) Vortex generators. 7

Small Hydraulic Diameter For constant velocity and air properties: Ellipse Flat ? ? ?

Small Hydraulic Diameter For constant velocity and air properties: Ellipse Flat ? ? ? 8

Tiwari et al. (2003), Min & Webb (2004) Jang & Yang (1998), Study type

Tiwari et al. (2003), Min & Webb (2004) Jang & Yang (1998), Study type Full-Scale HX design and optimization Fin minimization /elimination x Numerical x Experimental Matos et al. (2001) x Onishi et al. (2010 -14) x x Paitoonsurikarn et al. (2000), Kasagi et al. (2003), Abdelaziz et al. (2010), Bacellar et al (2014 -15) Hilbert et al (2006), Ranut et al (2014) Bacellar et al. (2014 -2016) Reduced Characteristic length Shape optimization Selected Studies Alternate shapes (reference: round) Literature Survey N/A x x Numerical x x x Numerical / Experimental x x x Numerical x x TFAWS 2016 – August 1 -5, 2016 x Numerical / Experimental 9

Problem Formulation • 1. 0 k. W Air-to-Water HX – Cross-flow; water (once through)

Problem Formulation • 1. 0 k. W Air-to-Water HX – Cross-flow; water (once through) – Fixed flow rates • Baseline: Microchannel HX – – V = 230 cm³ Af = 0. 0102 m² ΔPair = 78 Pa hair = 144 W/m². K TFAWS 2016 – August 1 -5, 2016 10

AIRSIDE MODELING TFAWS 2016 – August 1 -5, 2016 11

AIRSIDE MODELING TFAWS 2016 – August 1 -5, 2016 11

Non-Uniform Rational B-Splines Piegl and Tiller (1997) TFAWS 2016 – August 1 -5, 2016

Non-Uniform Rational B-Splines Piegl and Tiller (1997) TFAWS 2016 – August 1 -5, 2016 12

CFD Model • Platform – Geometry and Mesh: Gambit 2. 4. 6 – Simulations:

CFD Model • Platform – Geometry and Mesh: Gambit 2. 4. 6 – Simulations: ANSYS Fluent 14. 5 • Assumptions – Airside only constant wall temperature – Dry air – Steady-state flow – Non-existent energy and mass sources nor external forces – Negligible gravitational effects – Density (ideal gas) – Polynomial curve fit for thermophysical properties – Negligible pressure work and kinetic energy • Settings – Turbulence model: k-ε Realizable – Solver: coupled pressure-velocity • Mesh – Mapped growing elements (1. 2) near wall – Pave (core) TFAWS 2016 – August 1 -5, 2016 13

PROOF-OF-CONCEPT DESIGN & VALIDATION TFAWS 2016 – August 1 -5, 2016 14

PROOF-OF-CONCEPT DESIGN & VALIDATION TFAWS 2016 – August 1 -5, 2016 14

NTHX-001 Design Coil. Designer® (in-house software for air-to-refrigerant HX simulation in crossflow) Metric Face

NTHX-001 Design Coil. Designer® (in-house software for air-to-refrigerant HX simulation in crossflow) Metric Face Area (Af) Water Cross Section Area (Acs) Airside Heat Transfer Area (Ao) Envelope Volume (VHX) Material Volume Internal Volume (Fluid) Compactness (Ao/VHX) Air Frontal Velocity (u) Airside Heat Transfer Coefficient (hair) Airside Pressure Drop (ΔPair) Airside Conductance (UAair) Heat Load (Q) Unit m² m² cm³ cm³ cm²/cm³ m/s W/m². K Pa W/K W MCHX 0. 0102 173 0. 312 230 77 32. 9 44. 3 2. 94 144 78 44. 9 1109 TFAWS 2016 – August 1 -5, 2016 NTHX-001 0. 0100 186 0. 211 174 46. 8 18. 6 57. 5 3. 00 200 64 42. 2 1072 Relative Difference -2. 0% 7. 4% -32. 3% -24. 3% -39% -43. 5 29. 8% 2. 0% 38. 9% -17. 9% -5. 9% -3. 3% 15

NTHX-001 Experimental Setup TFAWS 2016 – August 1 -5, 2016 16

NTHX-001 Experimental Setup TFAWS 2016 – August 1 -5, 2016 16

NTHX-001 Experimental Validation 250 225 0. 9 200 +5% 0. 8 -5% 0. 7

NTHX-001 Experimental Validation 250 225 0. 9 200 +5% 0. 8 -5% 0. 7 ΔPair - CFD (Pa) Average Capacity - Simulation (k. W) 1. 0 +10% 175 -10% 150 125 100 75 0. 6 50 0. 5 25 0. 6 0. 7 0. 8 0. 9 1. 0 Average Capacity - Experimental (k. W) 25 50 75 100 125 150 175 200 225 250 ΔPair - Experimental (Pa) Average experimental uncertainty: ~5% for capacity, 3% for pressure drop Computational uncertainty: less than 1. 0% for both h and pressure drop TFAWS 2016 – August 1 -5, 2016 17

NUMERICAL OPTIMIZATION 18

NUMERICAL OPTIMIZATION 18

Framework Start Design of Experiments Random samples Run PPCFD Airside h, ΔP Define Design

Framework Start Design of Experiments Random samples Run PPCFD Airside h, ΔP Define Design Space Perform CFD Uncertainty Analysis Refine mesh, revise CFD settings Airside h, ΔP Evaluate metamodel yes no Acceptable? Create metamodel yes Run metamodel Evaluate metamodel Metamodel yes Create New HX Design Evaluate HX in Coil. Designer® Run MOGA Approximation Assisted Optimization Acceptable? no no Finish? yes Acceptable? Run PPCFD Optimum HX’s no End TFAWS 2016 – August 1 -5, 2016 19

Parallel Parameterized CFD (PPCFD) • Methodology to • Advantages – Generate geometries – Generate

Parallel Parameterized CFD (PPCFD) • Methodology to • Advantages – Generate geometries – Generate mesh files – Generate & execute CFD runs file – Post process output – Fast evaluation of parameterized geometries • allows topology/shape change – Applicable to most domains – Significant reduction in engineering time (~90%) TFAWS 2016 – August 1 -5, 2016 20

RESULTS 21

RESULTS 21

Normalized Pf (-) Optimization Problem 1 0. 9 NTHX-001 NTHX-O 021 0. 8 0.

Normalized Pf (-) Optimization Problem 1 0. 9 NTHX-001 NTHX-O 021 0. 8 0. 7 0. 6 0. 3 22 MCHX 0. 4 0. 5 0. 6 0. 7 0. 8 Normalized VHX (-) 0. 9 1

CFD Analysis (Re. Dh = 585) TFAWS 2016 – August 1 -5, 2016 23

CFD Analysis (Re. Dh = 585) TFAWS 2016 – August 1 -5, 2016 23

Conclusions • Presented a comprehensive multi-scale and shape optimization methodology • Presented novel designs

Conclusions • Presented a comprehensive multi-scale and shape optimization methodology • Presented novel designs that outperform MCHX without the aid of fins – Airside thermal-hydraulic performance (UA/ΔP): 14%↑ – Envelope volume: 20%↓; Material volume: 40%↓, Internal volume: 43%↓; Face area: 2%↓ • Successfully prototyped a novel HX and validated the methodology (less than 10%) • Great potential for weight reduction due to smaller fluid charge and material reduction • Application not limited to air-to-water, but any gas-to-fluid crossflow HX’s (evaporators, condensers) 24

Acknowledgements This work had support from the United States Department of Energy (US DOE)

Acknowledgements This work had support from the United States Department of Energy (US DOE) Grant Number DE-EE 0006114, Oak Ridge National Laboratory (ORNL) and the Modeling and Optimization Consortium (MOC) of the Center for Environmental Energy Engineering (CEEE) at the University of Maryland. 25

THANK YOU! 26

THANK YOU! 26