Presentation is loading. Please wait.

Presentation is loading. Please wait.

Report on testing Snake2 u-channel. P. Jalocha & J. Buytaert. 8 June 2015.

Similar presentations


Presentation on theme: "Report on testing Snake2 u-channel. P. Jalocha & J. Buytaert. 8 June 2015."— Presentation transcript:

1 Report on testing Snake2 u-channel. P. Jalocha & J. Buytaert. 8 June 2015

2 Outline Description of SNAKE2 and thermal loads. New SNAKE2 measurements.  Fluidic characterisation  Cooling performance. 8 June 2015 CO2 cooling meeting 2

3 SNAKE2. 8 June 2015 CO2 cooling meeting 3 All 19 channels have nearly the same length and fluidic resistance. Si–Pyrex prototype 2 new samples were produced at EPFL by PH/DT. Adrien Toros & Alessandro Mapelli. 1 sample has a small perforation in a channel, but can be repaired. See first CO2 circulation in : https://www.youtube.com/watch?v=8tzuo8y78y4

4 SNAKE2 main features. 8 June 2015 CO2 cooling meeting 4 Restrictions Alignment holes No manifolds in Silicon Each channel has its own inlet and outlet Manifolds are moved inside metallic connector.

5 SNAKE2 vs SNAKE1 channels dimensions 8 June 2015 CO2 cooling meeting 5 Optimizations to reduce fluidic resistance (by a factor ~4) while maintaining resistance to pressure: Main channels have the same width (200 µm) but increased depth (120 µm) Restrictions made squared (60 µm x 60 µm) to avoid clogging Considering the CO 2 is at -20˚C and assuming 30% vapor quality, we need a flow of 0.52g/s to dissipate 43 W/module. The pressure drop will be ~ 3 bar. Old layoutNew layout 70 µm 30 µm 200 µm Restriction Main Channel 60 µm200 µm 120 µm 60 µm 400 µm Restriction Main Channel

6 Provisional fluidic connector. 8 June 2015 CO2 cooling meeting 6 Metallic clamp with O-rings. The sample is in contact with the metal only on the pyrex side (good isolator). Not on the Si side. But the cooling tubes run through one half of the clamp.(->some thermal inertia)

7 Cooling test using dummy heaters. 8 June 2015 CO2 cooling meeting 7 Sensor heaters ASIC heaters Thermal mockups are made of 300 µm Si metallized to simulate the heat density ASIC Sensor ASIC Sensor Pyrex 2 mm Temperature probes Stay stick 672 75µm, 1W/m.K 3M acrylic tape 9461P: 20um, 0.17W/m.K Silicon 260µm

8 Fully wired up. 8 June 2015 CO2 cooling meeting 8 T_uch1 T_uch2 T_uch3 T_clamp 4 temperature probes: 2 probes on heaters « uch1,uch2 » 1 probe directly on silicon: « uch3 » 1 probe on metallic clamp « T_clamp » From cooling inlet to outlet: uch1 -> uch2 ->uch3 Heaters : 12 ASIC heaters: two groups of 6 ASIC in parallel No power on sensor heaters Preheater on inlet tube. All heaters are mounted on Si side Preheater on inlet tube CO2 flow

9 Installed in setup: 8 June 2015 CO2 cooling meeting 9 Installed with Pyrex upwards: (heaters are below) -> flow is visible through window in vacuum tank.

10 Fluidic system setup 8 June 2015 CO2 cooling meeting 10 Mass Flow meter. Manual valve Needle Valve. ‘bypass’ pump ? Micro channels Pressure Gauge P1 Pressure Gauge P2 Vacuum tank TRACI_V3 F m2 = f(dP, T, Heat) ? dP=P1-P2 FmFm F m1 F m2 Heat F m1 = ~5 x F m2 Manual valve is used to exclude micro channels from the flow. -> Only bypass needle valve

11 Labview Measuring & logging. 8 June 2015 CO2 cooling meeting 11 dP = Differential pressure across the microchannel [bar]: Red = instantaneous White = averaged Total flow through [g/s]: Red = instantaneous White = averaged Cv = flow coefficient of microchannels [g/s /sqrt(bar)] Temperatures measured on sample 1.Uch1 = White = heater close to inlet 2.Uch2 = Red = heater half-way 3.Uch3 = Cyan = silicon near the outlet 4.Clamp = Orange = clamping block

12 Method for obtaining fluidic characteristics of u-channel. F m2 = f(dP,T,Heat) = F m – C v (T)/sqrt(dP). F m,dP, T and Heat are measured C v (T) of needle valve must be calibrated precisely.  It is Independent of pressure and flow.  Only depends on temperature and type of fluid.  temperature calibration has been done in our setup 8 June 2015 CO2 cooling meeting 12 flow in uchannels total flow flow in bypass

13 Cv of bypass temperature calibration. 8 June 2015 CO2 cooling meeting 13 This calibration allows to deduce the flow in the uchannels by subtracting the flow in the bypass from the total measured flow.

14 C v of uchannels. The flow/pressure relation is ‘orifice’-like: F m = C v /sqrt(dP) 8 June 2015 CO2 cooling meeting 14 Cv ~ 0.2 gram/s/sqrt(bar) dP : 2.5 to 6.5 bar

15 Cv of uchannel depends on flow type. 8 June 2015 CO2 cooling meeting 15 Temperatures drop, when CO2 becomes better coolant in two phase. Pressure rises when the CO2 starts to boil triggered by the pre-heater Flow drops when boiling starts Flow coeff. drops by ~0.06 (=~30%) when boiling starts. (This is averaged data thus it is delayed) Preheater is switched ON causing CO2 boiling in inlet tube. (No heating power on Si is applied )

16 Cv of uchannel depends on heat input ( vapor quality). 8 June 2015 CO2 cooling meeting 16 Cv drop becomes larger at lower temperatures. To be studied further.

17 10W applied but all temperatures go up and follow each other Pre-heater applied to trigger the boiling Pre-heater applied again and sustained, as some channels stop to boil At 60W white and red start to separate At 70W the outlet temperature jumps up. Stop heating. Applied power (white) and max. theoretical power (red) derived from CO2 flow and latent heat. Note that the red curve is delayed as it is based on averaged CO2 flow coeff. Cooling test at -30 C and 6 bar pressure 8 June 2015 CO2 cooling meeting 17

18 Maximum cooling power at 0.4 gr/s flow. Obtained at flow ~ 0.4gr/s (dp=4bar) Required maximum module cooling power = 40W Required expected module cooling power = 30W What criteria to define maximum power ?  Sofar we used run-away of uch3.  We will try to calculate Si surface temperature underneath the probes by subtracting drop in thermal resistance. 8 June 2015 CO2 cooling meeting 18 T_traciTin1Tuch3 Max power Vapor quality Cv uch @no load Cv uch @ max load dCv uchdP -25-22.8-23.6600.760.2060.130.0764.07 -15-12.8-13.6660.770.210.150.064 -5-2.8-3.6660.80.2160.160.0564 56.16.6580.84 0.2120.170.0423.95 1516.616.157? 0.1890.160.0294

19 Full VELO cooling plant requirement. Need 0.4gr/s flow per module for 60W cooling power at 80% vapour quality.  This gives almost a factor 2 margin compared to expected heat load (30W). Need 20.8gr/s flow for 52 modules :cooling capacity is 3kW at 80% vapour quality. Obtained at 4 bar pressure on uchannels. 8 June 2015 CO2 cooling meeting 19

20 Superheated liquid behaviour. 8 June 2015 CO2 cooling meeting 20 Pre-heater power (6W) applied triggers boiling of CO2. Single phase two phase Superheated state is very frequent ! Almost always. Preheater is very effective. Observed also that some channels are not boiling when others are. In a reproducible pattern.

21 Superheated liquid behaviour. Patterns ? 8 June 2015 CO2 cooling meeting 21 Channels boil at different points ? Also observed that some channels are not boiling when others are. In a very reproducible pattern. Need fast camera to study.

22 Plans Snake2 cooling performance is very good:  Can cool up to ~60W with vapour quality ~80% at flow rate ~0.4gr/s. (dp=4bar) Uchannel design has good hydraulic characteristic. :  Required flow (04.gr/s) can be reached with low dP (~4bar).  Cv lowers by ~30% at max cooling power and at -30C. Cooling plant specification: 20g/s flow rate for and 4bar differential pressure There are issues with super-heated or subcooled state of the liquid CO2 in the channels.  Preheater on common inlet tube is necessary.  Need study with fast camera. 8 June 2015 CO2 cooling meeting 22


Download ppt "Report on testing Snake2 u-channel. P. Jalocha & J. Buytaert. 8 June 2015."

Similar presentations


Ads by Google