cfd analysis of lar flow in 35 ton prototype protodune
play

CFD Analysis of LAr Flow in 35 ton prototype, ProtoDUNE, & LBNF - PowerPoint PPT Presentation

CFD Analysis of LAr Flow in 35 ton prototype, ProtoDUNE, & LBNF cryostats Gregory Michna Stephen Gent Aaron Propst Department of Mechanical Engineering South Dakota State University November 10, 2017 Project Goals Study impurity levels


  1. CFD Analysis of LAr Flow in 35 ton prototype, ProtoDUNE, & LBNF cryostats Gregory Michna Stephen Gent Aaron Propst Department of Mechanical Engineering South Dakota State University November 10, 2017

  2. Project Goals • Study impurity levels within a LAr cryostat using Computational Fluid Dynamics (CFD) simulation methods. • Explore effect on impurity levels by changing: – LAr circulation flow rate and inlet temperature – LAr inlet and outlet locations – Internal electronics heat load • Desire a uniform and stable distribution of impurities 11/10/2017 Liquid Argon Flow CFD Simulations 2

  3. Methods • Simulate LAr motion due to natural convection (buoyancy) with Boussinesq model. – Fluid body force equation: � � � � � � � ��� � � � � thermal expansion coefficient, � ��� � average temperature • Simulate impurity levels with a passive scalar. – Passive scalar is carried (convected and diffused) by LAr similar to colored dye in water. – One ‐ way coupling: “passive” scalar does not affect the LAr motion. • Simplify cryostat FC and APA geometry using porous regions. 11/10/2017 Liquid Argon Flow CFD Simulations 3

  4. Impurity Level Quantification • Method 1: Electron Lifetime ����� ��∗�� ��� ���� – τ �������� �� � ����� ���� ∗���������� ����� �� � ∗�� ��� ����� – Useful when exact value of the impurity surface flux is known. – Can compare to experimental electron lifetime measurements from 35 Ton cryostat. 11/10/2017 Liquid Argon Flow CFD Simulations 4

  5. Impurity Level Quantification • Method 2: Normalized Percent Difference – Impurity level scaled such that the average level within the field cage is 1 (or 100%). – Levels expressed as +/ ‐ % above or below 100%. – Useful when exact value of the impurity surface flux is unknown. – Easier to compare impurity levels between simulations. 11/10/2017 Liquid Argon Flow CFD Simulations 5

  6. Simulations to Date ProtoDUNE 35 Ton • 35 Ton • V1 Design – Full and symmetric models • V2 Design – Symmetric only • Latest Design Top View of Latest Design – Full and symmetric – Various operating conditions � � �. � � • ProtoDUNE � � �. � � Removed in Symmetric Models 11/10/2017 Liquid Argon Flow CFD Simulations 6

  7. 35 Ton Simulation • Red: heat enters through wall • Blue: constant temperature, constant impurity flux • Yellow: field cage is a 23% porous wall • 9.5 GPM LAr flow rate • Constant inlet temperature: 87.808 K 11/10/2017 Liquid Argon Flow CFD Simulations 7

  8. 35 Ton Simulation Impurity Distribution Fermilab Results (Erik Voirin, Fermilab) 11/10/2017 Liquid Argon Flow CFD Simulations 8

  9. 35 Ton Simulation 4 purity monitors in this corner (Geometry Not accounted for in CFD model) (Erik Voirin, Fermilab) 11/10/2017 Liquid Argon Flow CFD Simulations 9

  10. 35 Ton Simulation 6000 5500 5000 SDSU Simulation Electron LIfetime [ μ s] Fermilab Simulation ‐ Probe 1 4500 Fermilab Simulation ‐ Probe 2 Fermilab Simulation ‐ Probe 3 4000 Experimental 3500 3000 2500 0 0.5 1 1.5 2 2.5 Elevation from Cryostat Floor [m] 11/10/2017 Liquid Argon Flow CFD Simulations 10

  11. 35 Ton Simulation • Simulation agrees with experimental data. • Can apply same CFD methods to other designs. 11/10/2017 Liquid Argon Flow CFD Simulations 11

  12. LBNF Cryostat ‐ Geometry APA CPA APA CPA APA • APA – approx. 80% open Field Cage • CPA – impermeable • Field Cage – 23% open Field Cage 11/10/2017 Liquid Argon Flow CFD Simulations 12

  13. LBNF Cryostat ‐ Geometry Cross Section from Side View Field Cage Field Cage Field Cage Field Cage 11/10/2017 Liquid Argon Flow CFD Simulations 13

  14. LBNF Cryostat – Boundary Conditions • Top Wall (LAr surface): V1 Full V1 Sym. Latest Sym. – LAr Saturation Temperature: 88.348 K Inlet Flow Rate 4 Pumps 4 (2) Pumps 1 (0.5) pump – Passive Scalar Flux: 1 # of Inlets 1 1 (0.5) 12 (6) • Remaining Exterior Walls: # of Outlets 4 4 (2) 7 (7) – Heat Flux: 7.2 W/m^2 • Electronics Surfaces: Single Pump = 103 GPM – Total Heat Source: 23,700 W • Inlet Temperature: – Maintained at 0.4418 K above outlet temperature to account for pump work – Flow rate in table on the right • APA and FC Planes: – Treated as Porous Region, see next slide Electronics Surfaces in pink 11/10/2017 Liquid Argon Flow CFD Simulations 14

  15. Representing APA Plane with Porous Region The APA planes consisted of 10 layers: • Plane 1 : Vertical wires (150 micron diameter at a 5 ‐ mm pitch) • Plane 2 : +60° wires • Plane 3 : ‐ 60° wires • Plane 4 : Vertical wires • Plane 5 : Mesh 80% open (90° wires of 0.528 ‐ mm dia. and 5 ‐ mm pitch) 1 • Planes 6 ‐ 10 : Symmetry of planes 1 ‐ 2 3 4 5 5, with a 75 mm space between planes 5 and 6. 11/10/2017 Liquid Argon Flow CFD Simulations 15

  16. Representing APA Plane with Porous Region • Motivation: Cells required to Outlet represent real APA geometry for APA Mesh Layers entire cryostat is vastly beyond computational resources. Symmetry on 3 sides • Mimic the flow resistance on the macro ‐ scale using porous regions. – Simulate only a small section of real APA plane geometry. Inlet – Find pressure drop across planes at several velocities in expected range. 11/10/2017 Liquid Argon Flow CFD Simulations 16

  17. Representing APA Plane with Porous Region • Plot pressure vs. velocity. 1.2 • Determine quadratic trend line. 1 Pressure Drop [Pa] • Use coefficients as inertial ( � � ) and 0.8 viscous ( � ) flow resistance 0.6 coefficients. 0.4 • Divide coefficient by porous region y = 563.21x 2 + 5.9315x thickness. 0.2 R² = 0.9999 • Final APA resistance coefficients: 0 0 0.01 0.02 0.03 0.04 – Inertial: 11,300 kg/m^4 Velocity [m/s] Fermilab Simulation SDSU Simulation – Viscous: 119 kg/m^3 ‐ s Poly. (SDSU Simulation) 11/10/2017 Liquid Argon Flow CFD Simulations 17

  18. Representing FC Plane with Porous Region • FC plane consist of 23% open, slot geometry, assumed 23 mm slot at 100 mm pitch. Symmetry on all • Used same method as APA plane four sides Outlet to find resistance coefficients. • Final resistance coefficients: 2.3 cm slot – Inertial: 411,000 kg/m^4 – Viscous: 247 kg/m^3 ‐ s Inlet 11/10/2017 Liquid Argon Flow CFD Simulations 18

  19. LBNF V1: Impurity and Temperature at z = 30.5 m plane (pump discharge) 11/10/2017 Liquid Argon Flow CFD Simulations 19

  20. LBNF V1: Impurity and Temperature at z = 0 m plane (center of cryostat) z = 0 m plane (center of cryostat) 11/10/2017 Liquid Argon Flow CFD Simulations 20

  21. Simulations • Latest Configuration: – Symmetric: standard operating conditions, electronics turned off, and half LAr flow rate. – Running full model: Erik Voirin’s results showed significant asymmetry. Top View of Latest Configuration � � �. � � � ����� � � �. � � Removed in Symmetric � ������ Models 11/10/2017 Liquid Argon Flow CFD Simulations 21

  22. Mesh Validation • Used two mesh types with varying levels of refinement. • Solutions have been in agreement. • Polyhedral mesh requires more iterations and time (about 30%) to solve the passive scalar for impurity distribution. • Currently using trimmed cell mesh (hexahedral, cubes of varying sizes). Trimmed Polyhedral 11/10/2017 Liquid Argon Flow CFD Simulations 22

  23. Latest Design: Symmetric vs. Full Model • Simulating half the cryostat will cut calculation time in half. • Must determine if both full and symmetric models yield similar results. 11/10/2017 Liquid Argon Flow CFD Simulations 23

  24. Sym. vs. Full: Temperature at Z = 5.17 m In Line with Inlet Full Symmetric 11/10/2017 Liquid Argon Flow CFD Simulations 24

  25. Sym. vs. Full: Impurity at Z = 5.17 m In Line with Inlet Full Symmetric 11/10/2017 Liquid Argon Flow CFD Simulations 25

  26. Sym. vs. Full: Temperature at X = 3 m Full Symmetric 11/10/2017 Liquid Argon Flow CFD Simulations 26

  27. Sym. vs. Full: Impurity at X = 3 m Full Symmetric 11/10/2017 Liquid Argon Flow CFD Simulations 27

  28. Electronics Turned Off • Heat flux on electronics changed from 23,700.0 W to 0.0 W. • No other changes. • Will compare impurity level minimum, maximum, and standard deviation after slides of images. 11/10/2017 Liquid Argon Flow CFD Simulations 28

  29. Electronics Off: Temperature at Z = 5.17 m In Line with Inlet Electronics On Electronics Off 11/10/2017 Liquid Argon Flow CFD Simulations 29

  30. Electronics Off: Impurity at Z = 5.17 m In Line with Inlet Electronics On Electronics Off 11/10/2017 Liquid Argon Flow CFD Simulations 30

  31. Electronics Off: Temperature at X = 3 m Electronics On Electronics Off 11/10/2017 Liquid Argon Flow CFD Simulations 31

  32. Electronics Off: Impurity at X = 3 m Electronics On Electronics Off 11/10/2017 Liquid Argon Flow CFD Simulations 32

  33. Half Flow Rate • LAr inlet flow rate changed from 103 GPM to 51.5 GPM 11/10/2017 Liquid Argon Flow CFD Simulations 33

  34. Half Flow Rate: Temperature at Z = 5.17 m In Line with Inlet Regular Flow Rate Half Flow Rate 11/10/2017 Liquid Argon Flow CFD Simulations 34

  35. Half Flow Rate: Impurity at Z = 5.17 m In Line with Inlet Regular Flow Rate Half Flow Rate 11/10/2017 Liquid Argon Flow CFD Simulations 35

Recommend


More recommend