Document GmjvdoeZ8pbKQoYVZ0GbDyYwN

STS COOLING CONCEPT AND DESIGN K. Agarwal Eberhard-Karls-Universitt Tbingen - Tbingen (DE) - for the CBM-STS Team - 37th CBM Collaboration Meeting Virtual Meeting 05/03/2021 HONEST CONFESSION "Remember that all models are wrong; the practical question is how wrong do they have to be to not be useful." G.E.P. Box; N.R. Draper, Empirical Model-Building and Response Surfaces, John Wiley & Sons, pg. 74 (1987) George Edward Pelham Box 1919 - 2013 FRS; Statistician The contents shown in this presentation are just simulations/approximations/expectations. Reality could be, and probably would be different. 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design A. F. M. Smith, George Edward Pelham Box, Biogr. Mems Fell. R. Soc. 61, 23-37 (2015) 2 MOTIVATION TO COOL SILICON SENSORS Non-Ionizing Radiation Damage Defects in Si-Lattice Intermediate Energy States (Higher) Leakage Current Frank Hartmann, Evolution of Silicon Sensor Technology in Particle Physics (2nd Edition) - Springer Tracts in Modern Physics, Volume 275 (2017) Conduction Band; EC e- e- e- Intermediate States "Steppingstones" Valence Band; EV h+ h+ h+ 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 3 MOTIVATION TO COOL SILICON SENSORS Leakage Current varies linearly with fluence: = This also increases the detector shot noise and degrades the S/N ratio: Non-Ionizing Radiation Damage Defects in Si-Lattice Intermediate Energy States (Higher) Leakage Current Higher Radiation Higher Current Higher Noise Frank Hartmann, Evolution of Silicon Sensor Technology in Particle Physics (2nd Edition) - Springer Tracts in Modern Physics, Volume 275 (2017) = Power Dissipation in a silicon sensor is described as: = ~ Higher Temp. Cooling Power Higher Power Dissipation Power Dissipation (or leakage current) doubles whenever the silicon temperature is increased by 10 C If the Cooling Power can't neutralise the Heating Power caused due to high STS irradiation environment, then this self feeding system goes into `Thermal Runaway' 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 4 MOTIVATION TO COOL SILICON SENSORS (& PLANET EARTH) Planet Earth is also in a `Thermal Runaway' Non-Ionizing Radiation Damage Defects in Si-Lattice Intermediate Energy States (Higher) Leakage Current Higher Radiation Higher Current Higher Noise Arne Navra, www.NaturBilder.no 05/03/2021 - 37th CBM Collaboration Meeting Higher Temp. Higher Power Dissipation Cooling Power If the Cooling Power can't neutralise the Heating Power caused due to high STS irradiation environment, then this self feeding system goes into `Thermal Runaway' K. Agarwal - STS Cooling: Concept and Design 5 IDENTIFYING POWER INTENSIVE AREA STS Front View STS Top View Only the innermost sensors of the central ladders (x,y 10 cm) of all stations, where power dissipation is less than 3mW/cm (@-10C, EOL) requires active cooling The peripheral sensors of the central and peripheral ladders can be cooled by natural convection (-10C) Online Tool: https://fair-center.eu/fileadmin/fair/experiments/CBM/tmp/CBM_FLUKA.htm Detailed power dissipation numbers for inner ladders of Station #1 and #8 are tabulated in the backup slides (#26) 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 6 BOUNDARY CONDITIONS AND POTENTIAL SOLUTION Detailed Calculations: K. Agarwal, Thermal Runaway Analysis for STS Sensors, STS Coordination Meeting, 25.01.2021 | Link BOUNDARY CONDITIONS - Power dissipation for innermost sensors ~ 6 mW/cm (at -10C at EOL fluence 1014 neq(1 MeV)/cm) Target sensor temp. -10 ... -5C to achieve S/N 10 To be done by adding minimum additional material budget in active detector geometry Thermal Runaway COOLING IDEA Forced convective air cooling Cold Gas , Warm Sensor Surface B , L If Tstable < Tsensor < Tcritical If Tsensor Tcritical Stable operation at Tstable Thermal Runaway Theoretically, it's possible to stably operate the innermost STS sensors at -7C with sufficient margin from thermal runaway (by assuming that one can achieve an infinite heat sink around at -10C at 3 m/s) 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 7 PRACTICALITIES C-Frame Silicon Sensors Microcables CF Ladders Perforated Tube O.D. 2.4 mm Wall thickness 200 m Silicon Sensor Surface Cooling Element blowing cold gas Top View Beam Pipe Cooling Element: Perforated Tube running along the sensors 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 8 TOY MODEL - SETUP C-Frame Silicon Sensors Microcables CF Ladders Side View Front View Inlet Cooling Element blowing cold gas Top View Beam Pipe Cooling Element: Perforated Tube running along the sensors Toy Model developed in SolidWorks for CFD Simulations with 2 dummy stations resembling the STS stations no. 1 and 8 in an enclosure (only active volume of the detector) 20 lt/min of cold air flow at -10C in every tube to cool only the innermost sensors 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design Dummy Station 8 Dummy Station 1 Outlet Outlet Inlet 9 TOY MODEL - FLOW AND TEMP. DISTRIBUTION Dummy Station 1 Dummy Station 8 05/03/2021 - 37th CBM Collaboration Meeting Dummy Station 1 For an ambient temperature of -10C, with flow @-10C and 20 lt/min per tube, a stable operation with max. temp. at -4.9C is possible! K. Agarwal - STS Cooling: Concept and Design Dummy Station 8 10 Material Budget Estimates: J.M. Heuser, U. Frankenfeld, Production Readiness Review for the STS Carbon Fiber Ladders, CBM-TN-19006 (2019) | Link PERFORATED TUBE - MATERIAL & MANUFACTURABILITY Perforated tubes made from carbon fibre (CF); Outer Diameter = 2.4 mm, Thickness = 200 m Local addition of x/X0tube = 0.14% radiation length to its location of mounting. So, the average material budget is only increased very marginally and will soon be added in the STS GEANT geometry. Manufacturability tested by ICM-Composites (DE) Milling down of thicker CF-tube 05/03/2021 - 37th CBM Collaboration Meeting Perforated CF-tube K. Agarwal - STS Cooling: Concept and Design 10 samples delivered at GSI 11 STS W/O AND W/ PERIPHERAL INFRASTRUCTURE 40 kW Power Dissipation! y x z 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 12 REQUIREMENTS ON ELECTRONICS COOLING C-Frame (support structure) Cooling Plate carrying the coolant 05/03/2021 - 37th CBM Collaboration Meeting Front-End Electronics Board (FEB) Boxes PFEB = 10 W (approx.) PFEB Box = 100 W Silicon Sensor mounted on the CF ladder (microcables hidden) Since the electronics are 25...50 cm away from the innermost sensors, one must neutralise the electronics power dissipation Temp. exposed by the FEB Boxes Temp. on the sensors -10C K. Agarwal - STS Cooling: Concept and Design 13 TRANSITION FROM BI-PHASE CO2 TO MONO-PHASE 3M NOVEC 649 J.M. Heuser (eds.) et al., Technical Design Report for the CBM Silicon Tracking System (STS) - GSI Report 2013-4 (2013) Biphase CO2 (GWP = 1) TM TM Monophase 3M NOVEC 649 (GWP = 1) Operational in various trackers at (HL-)LHC, ISS Great performance less mass flow; = 150 kJ/kg (at -20C) low pressure drop; = 0.14 cSt (at -23C) smaller tubes; = 10 kW/m.K uniform temperature Higher system pressures i.e., safety regulations Potentially difficult for commercial manufacturing (2PACL-type system) CBM-STS FEE Cooling Conceptual Design Review, 10.12.2019 | Link To be used in LHCb Sci-Fi Tracker. Considered for more... Relatively lower performance higher mass flow; = 1.1 kJ/kg.K higher pressure drop; = 0.70 cSt (at -40C) larger tubes; = 2 kW/m.K non-uniform temperature Lower system pressures i.e., safe to use Easier commercial manufacturing Based on recommendations from CERN, CBM and industrial experts, the coolant for STS-FEE is NOVECT6M49 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 14 CFD SIMULATIONS - COOLING PLATE Finned cooling channels to increase heat transfer between the coolant and plate's inner surface Material: Inlet: Total power dissipation: Temperature outlet: Max. temp. of cooling plate: Pressure loss: Aluminium 3MTNMovecT6M49 -40C at 3 litre/min 800 W -33.8 C -28.2 C 1.32 bar 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 15 THERMAL FEA SIMULATIONS - FEB BOX TIM-4 TIM-3 TIM-3 TIM-4 LDOs ASICs T-Fin PCB Vias TIM-2 Shelf Plate TIM-1 Cooling Plate TIM: Thermal Interface Material 05/03/2021 - 37th CBM Collaboration Meeting PCB LDOs Vias ASICs Product TIM-1 TIM-2 TIM-3 TIM-4 DSN5040-10DC10DC DSN5040-10DC10DC Stycast 2850FT (+23LV) EPO-TEK E4110 K. Agarwal - STS Cooling: Concept and Design Thermal Conductivity (along z-) [W/m.K] 20 20 1.02 1.37 Thickness [m] 40 40 150 100 16 THERMAL FEA SIMULATIONS - FEB BOX 05/03/2021 - 37th CBM Collaboration Meeting Since the temp. exposed to the environment (i.e., temp. `seen' by the silicon sensors) is ~ -20C which is much lower than -10C, there is substantial headroom for increasing the coolant temp. from -40C K. Agarwal - STS Cooling: Concept and Design 17 BABY COOLING PLANT - SIMPLIFIED P&ID Heat Exchanger Expansion Valve Heat Exchanger Expansion Tank STS Water Cooling Compressor Biphase CO2 Vapor Compression Cycle (Primary Circuit) Pump TM TM 3M NOVEC 649 Monophase Cycle (Secondary Circuit) All coolants used in this concept are GWP = 1, which makes this cooling plant usable for coming decades `Simple' to manufacture commercially by using established technology and industrial practices To be used by STS for the Thermal Demonstrator and detector assembly procedure & testing 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 18 BABY COOLING PLANT - COMPLETE P&ID Biphase CO2 Vapor Compression Cycle TM TM 3M NOVEC 649 Monophase Cycle 05/03/2021 - 37th CBM Collaboration Meeting TM K. Agarwal - STS Cooling: Concept and Design Water Cooling STS 19 BABY COOLING PLANT - DELIVERED & UNDER COMMISSIONING (pilot refrigeration plant for the pilot refrigeration system of the detector) Baby Cooling Plant delivered at GSI on 18.12.2020 (Pilot refrigeration/cooling plant for the pilot refrigeration system of the detector) 7.5 kW cooling capacity at -40C Commissioning to be done within 2 months 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 20 SUMMARY AND OUTLOOK SENSOR COOLING Based on CFD Flow Simulations, a cooling concept to avoid thermal runaway of the STS sensors at their end-of-lifetime has been established Manufacturability of low material budget CF-perforated tubes, which are the active sensor cooling elements, has been demonstrated ELECTRONICS COOLING tm tm 3M NOVEC 649 has been chosen as the coolant to neutralise the electronics power dissipation of 40 kW Combination of CFD and Thermal FEA Simulations shows that sufficient margin is available if one uses the coolant at -40C A Baby Cooling Plant, with 7.5kW cooling capacity at -40C, was delivered in 12/2020 which will be used for STS assembly and any intermediate tests BUSY ROAD AHEAD (2021) To experimentally demonstrate the feasibility of STS concepts in realistic boundary conditions, the construction of a thermal demonstrator is ongoing! 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 21 STS THERMAL DEMONSTRATOR - A `COOL' MOCK-UP https://atlas.cern/updates/blog/exploring-coolest-mockup A lot to look forward with the STS Thermal Demo 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 22 THANKS A LOT... QUESTIONS? COMMENTS? 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 23 (ROUGH) ESTIMATION OF COOLING & HEATING POWER Cold Air , Warm Surface , B L Bias Voltage = 300 V Temperature = 20C < 15 cm Q = h A (T - T ) Q T =T +h A k h = L Nu Nu = 0.664 Re . Pr vL Re = (Laminar Flow) Formulas: Yunus, Cengel, Turner, Fundamentals Of Thermal-fluid Sciences (2004), McGraw-Hill Assuming that the sensor thickness is 320m and , Leakage Current = 206.56 A/cm (+20C, 500V, 1014 neq/cm) Power Density = 103.28 mW/cm (+20C, 500V, 1014 neq/cm) Pavel Larionov (Uni. Frankfurt), PhD Thesis (2016) 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 24 LEAKAGE CURRENT IN HAMBURG MODEL 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 25 ESTIMATES FOR STS SENSOR HEATING FROM THE HAMBURG MODEL Olga Bertini, Results from mass QA inspection from HPK sensors Michael Moll - DESY THESIS-1999-040 Theoretical estimate of Leakage Current after EOL Neutron Irradiation at 20C at 500V: () = = + ____________________________ Assuming the following values: () < . = < = = = [] ____________________________ Theoretical estimate of Power Dissipation after EOL Neutron Irradiation at 20C at 500V: () = . = . () = . + = . Power Dissipation from the Hamburg Model is < 7.06 mW/cm (-10C, 500V, 1014 neq/cm) 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 26 POWER DISSIPATION CHART FOR THIS ANAYLSIS With Active Cooling Without Active Cooling Power Dissipation for Station 1 (@-10C, EOL) Inner Ladders (Left), Peripheral Ladders (Right) Sensor # () 1 () 2 () 3 () 4 () 5 y-Length [cm] EOL Power [mW/cm] Total Power [W] 2,2 6,00 0,08 2,2 6,00 0,08 4,2 6,00 0,16 6,2 3,16 0,12 6,2 0,49 0,02 Total Power Dissipation per Ladder = 0.92 W Sensor # () 1 () 2 () 3 () 4 () 5 y-Length [cm] EOL Power [mW/cm] Total Power [W] 2,2 0,70 0,01 2,2 0,70 0,01 4,2 0,70 0,02 6,2 0,51 0,02 6,2 0,37 0,01 Total Power Dissipation per Ladder = 0.14 W Power Dissipation for Station 8 (@-10C, EOL) Inner Ladders (Left), Peripheral Ladders (Right) Sensor # () 1 () 2 () 3 () 4 () 5 y-Length [cm] EOL Power [mW/cm] Total Power [W] 4.2 6.00 0.16 6.2 6.00 0.23 12.4 1.46 0.11 12.4 0.73 0.06 12.4 0.45 0.03 Total Power Dissipation per Ladder = 1.18 W Sensor # () 1 () 2 () 3 () 4 () 5 y-Length [cm] EOL Power [mW/cm] Total Power [W] 4,2 1,37 0,04 6,2 1,37 0,05 12,4 1,00 0,08 12,4 0,64 0,05 12,4 0,42 0,03 Total Power Dissipation per Ladder = 0.50 W 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 27 WHAT'S A REASONABLE FLOW VELOCITY ? Fluid NOVEC 649 Nitrogen Density [lbs/ft] [kg/m] 100 1600 0.0725 1.1606 Erosional Velocity [ft/sec] [m/sec] 10 2.5 371.4 92.8 Result for Novec 649 is in agreement with CERN recommendations This means that 100 m/s flow velocities could be O.K. for gases I.D. 4 mm 80 litre/min 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 28 QUESTION ABOUT THE ADDED MATERIAL BUDGET CF pipes aren't the dominant contributor to the material budget Material budget worth 1 additional pipe shouldn't be a big problem Must be proven with updated material budget simulations 05/03/2021 - 37th CBM Collaboration Meeting I.D. O.D. These scenarios are same from material budget point-of-view O.D. [mm] I.D. [mm] Thickness [mm] Shell Area [mm] 1.50 0.70 0.40 1.38 1.60 0.90 0.35 1.37 1.75 1.15 0.30 1.37 2.00 1.50 0.25 Currently 1.37 2.40 2.00 0.20 1.38 3.10 2.80 0.15 1.39 K. Agarwal - STS Cooling: Concept and Design 29 STS SENSOR MODULE To obtain high momentum resolution (which is imperative for tracking, especially at low momentum), its important to reduce any unnecessary material inside the detector acceptance: . .. Therefore, design of STS sensors are a `bit' different than conventional silicon sensors (e.g., at LHC exp.) Ideally, electronics should be immediately behind the sensors to reduce any in-line capacitance/noise But given that electronics would have added too much to the material budget, ultrathin microcables are used in-between 6 mW/cm @ -10C Non-ionising damage 1014 neq/cm Nominal FEE Power Dissipation per STS Module 27 W For whole STS, ~ 40 kW Silicon Sensor (inside detector acceptance) 05/03/2021 - 37th CBM Collaboration Meeting Microcables; 55 cm in length (inside detector acceptance) K. Agarwal - STS Cooling: Concept and Design Front-End Electronics (outside detector acceptance) 30 CARBON FIBRE LADDER STS Module: Silicon Sensors + Microcables + FEE Boards Light-weight Carbon-Fibre Ladders 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 31 FEB BOX STS Module: Silicon Sensors + Microcables + FEE Boards Front-End Electronic Board (FEB) Box 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 32 WHY 3M NOVEC 649 ? Novec and 3M are trademark of 3M (TM) Corporation Radiation hard (resistant to gamma radiation > 10 kGy). So minimal production of radiation byproducts, which otherwise could block cooling lines or cause corrosion! Usable down to -108C Unfortunately higher pressure drops, but could be compensated by higher tube diameters (for STS, ID = 6 mm) Green and possible long-term availability 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 33 3M NOVEC 649 COMMERCIAL AVAILABILITY Novec and 3M are trademark of 3M (TM) Corporation TM Higher quantities could be purchased directly from 3M Lower quantities are readily available from Ionic Liquid Technologies - IoLiTec (Heilbronn DE) 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 34 3M NOVEC 649 - WHERE IN USE SO FAR? Novec and 3M are trademark of 3M (TM) Corporation 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 35 CFD SIMULATIONS - COOLING PLATE Finned cooling channels to increase heat transfer between the coolant and plate's inner surface 05/03/2021 - 37th CBM Collaboration Meeting Material: Inlet: Total power dissipation: Temperature outlet: Max. temp. of cooling plate: Pressure loss: Aluminium TM TM 3M Novec 649 -40C at 3 litre/min 800 W -33.8 C -28.2 C 1.32 bar K. Agarwal - STS Cooling: Concept and Design 36 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 37 THERMAL FEA SIMULATIONS FOR FEB BOX PCB Vias LDOs T-Fin TIM-3 TIM-4 Shelf Plate Cooling Plate TIM-1 TIM-2 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 38 BABY COOLING PLANT - CAD SNAPSHOTS 05/03/2021 - 37th CBM Collaboration Meeting K. Agarwal - STS Cooling: Concept and Design 39