THERMOHYDRAULIC TRANSIENTS IN BOILING HELIUM NATURAL CIRCULATION LOOPS

Transcription

1 THERMOHYDRAULIC TRANSIENTS IN BOILING HELIUM NATURAL CIRCULATION LOOPS Hernán FURCI Director: Chantal MEURIS Supervisor: Bertrand BAUDOUY Laboratoire de Cryogénie et Station d Essais SACM/IRFU/CEA de Saclay D-DAYS 2014 IRFU JULY 2 ND AND 3 RD JUILLET 2014 PAGE 1

2 JOURNÉES DES THÉSARDS- 2 ET 3 JUILLET 2014 FURCI Hernán Nuclear Engineer, Instituto Balseiro, Argentine. Master 2 Sciences de la Fusion, Université d Aix-Marseille. Second Year PhD Candidate at MIPEGE, Paris Sud. I have been recommended to Bertrand BAUDOUY by Jean-Luc DUCHATEAU, a senior researcher at CEA and Master s professor. When I visited the lab, I was highly seduced by the experimental facility I would use and by the challenges involved in the subject, the contribution of this research project to bigger ones, the diversity and nature of the R&D at LCSE, the professional and human qualities of my prospective colleagues. July 2 nd 2014 PAGE 2

3 JOURNÉES DES THÉSARDS- 2 ET 3 JUILLET 2014 Thermohydraulic transients in boiling helium natural circulation loops Helium natural circulation is a cooling scheme in large superconducting magnets CMS at the LHC for CERN R3B-GLAD for GSI Passive safety reasons Already studied in steady state Not thoroughly studied in transient CMS thermosiphon cooling pipes R3B-GLAD cooling system We perform experiments on a big size helium natural circulation facility To explore the existing boiling regimes during transients at different powers and positions of a heated section; To identify heat transfer deterioration phenomena; To determine ways of mitigating its effects. We do numerical simulations To understand more deeply the phenomena To extrapolate results to other systems July 2 nd 2014 PAGE 3

4 Electric heating (spiral wire) EXPERIMENTAL SET-UP Liquid inlet PA Gas outlet Liquid level PV PT T5 Concentrated heating T4 T3 T2 T1 July 2 nd 2014 PAGE 4

5 Heat flux q Mass flow rate (kg/s) T-T(q=0) (K) STEADY-STATE REGIME Reference to which compare transient results. Power is increased gradually, at steps (quasi-steady evolution) Mass flow takes place as a result of buoyant forces. Two boiling regimes take place depending on power (and position). q" t Hydraulic regimes Heat transfer regimes Critical heat flux (CHF) Friction and acceleration pressure gradients limit flow growth and eventually reduce it. T5 Column weight decrease produces flow increase T4 T3 T2 Nucleate Boiling Film Boiling T1 Heat flux (W/m 2 ) Heat flux (W/m 2 ) April 10 th 2014 PAGE 5

6 TRANSIENT CRISIS q" q i "=0 PA q f " t Zero initial power transients, increasing final power We focus on high power transients only (the most critical) Temperature evolution at T5 CHF (steady) 1450 W/m 2 T5 In transient, crisis can appear (temporarily) at q < CHF. T4 T3 Observed only at T3, T4 and T5 (high void and transit time from inlet) T2 T1 July 2 nd 2014 PAGE 6

7 TRANSIENT BOILING CRISIS PARAMETERS We analyze the transients by calculating three main parameters: t c, the critical time: duration of nucleate boiling before crisis; t, the crisis duration: time between the initiation of the crisis and its end; T max, the maximum temperature attained. T max t c t July 2 nd 2014 PAGE 7

8 TRANSIENT CRISIS Critical time, t c t c T max Low vapor quality regime (near entrance) Wall-dominated? t PA Transition High vapor quality regime (t ctt./q) Bulk-dominated? High void flow patterns increase crisis onset velocity T5 T4 T3 T2 T1 July 2 nd 2014 PAGE 8

9 TRANSIENT PREMATURE CRISIS ATTENUATION q" q i " 0 q f " t Non-zero initial power transients Fixed final power (e. g W/m 2 ) Increasing initial power q i reduces duration and intensity of T excursion PA T4 q f = 1378 W/m 2 T5 q f = 1378 W/m 2 T5 T4 T3 T2 T1 July 2 nd 2014 PAGE 9

10 TRANSIENT PREMATURE CRISIS ATTENUATION q" q i " 0 q f " t We could draw the limits for the three types of behavior: Nucleate Boiling, Stable film boiling and Transient Crisis (final NB) PA SFB SFB NB SFB SFB NB SFB SFB NB T5 T4 T3 T2 Dotted lines give the inhibition limits determined increasing initial heat flux T1 July 2 nd 2014 PAGE 10

11 EXPERIMENTAL CAMPAIGNS The heated section of the loop can be changed. We used three different test sections so far: Vertical, D = 10 mm, L = 95 cm Vertical, D = 5 mm, L = 102 cm Horizontal spiral, D = 10 mm, L = 427 cm The results discussed in this talk concern only the vertical, D = 10 mm, L = 95 cm heated section experiments. Loop with horizontal heated section has very distinct behaviour from vertical. Thermohydraulic instabilities have been observed in wide power ranges. July 2 nd 2014 PAGE 11

12 MODELLING HOMOGENEOUS TWO-PHASE FLOW ρu t ρ t ρu2 + x + ρu x = 0 = p x p fr. x + ρg e x ρ = f ρ (h, p) ρ e + u2 ρu h + u2 Re h = ρud 2 2 μ h + = q + ρg u f Re = 0,316 Re t x 1/4 First approach: The systems was modeled in COMSOL, with some additionnal assumptions p fr. x = f Re h D ρu 2 2 sign(u) T5 T4 T3 T2 T1 a Energy Conservation Clossure Equation Momentum Cons. (Integral form) Mass Conservation g = ge x ρh t + ρuh x = q ρg u ρ = f ρ h = h h f h fg ρ g + h g h h fg ρ f G = P t a P e + ρu 2 a ρu 2 e p fr. + ρg e x dx a u x = q v fg h fg A D, u D R g = 0 A i, u i 1 u D = A i A D u i g = ge x q = 0 q 0 q = 0 e e G = ρu dx a e + ρu2 2 R July 2 nd 2014 PAGE 12

13 Quality at T5 Inlet velocity u i (m/s) Enthalpy (J/kg) MODELING SOME RESULTS Time delay 0.3 sec u i < 0 No cold mass entry, uniformity u i > 0 Cold front transport Incompressibility Sudden jump q f = 1300 W/m 2, q i = 0 Gibbs (numerical) q f = 1300 W/m 2, q i = 0 Time (s) Time (s) qf = 1300 W/m 2 Increasing q i Decreasing quality peak q i (W/m 2 ) Position (m) Essential overall behaviour captured. It lets us estimate the local variations of mass quality. Local values of quality can exhibit an overshoot, whose amplitude is higher as the initial power is closer to 0, due to finite transit time from the entrance of the test section This could explain transient crisis at the downstream positions of the test section, and not upstream. Refinement of simulations shall be done by including pressure effects and improving numerical aspects. July 2 nd 2014 PAGE 13

14 CONCLUSION Findings of this work Transient behavior of a boiling helium thermosiphon with a vertical heated section has been studied experimentally. Power-premature boiling crisis has been observed after power step pulses. Boiling crisis is more likely to happen at higher positions (higher void and transit time from entrance). Initially established flow can inhibit this transient feature, which gives us hints of how to protect devices from this effect. A first simple model predicts quite precisely the mass flow rate measurements, which would allow to evaluate correctly bulk vapor concentration evolutions. Other work-lines Horizontal heated section (experiments finished, data being processed) Thinner vertical heated section (experiments finished, data being processed) Modeling of the hydraulics: consider pressure effects (coming soon) Evaluation of macroscopic conditions that may lead to crisis during transients in horizontal and vertical sections. July 2 nd 2014 PAGE 14

15 Thank you for your attention!??? COMMISSARIAT À L ÉNERGIE ATOMIQUE ET AUX ÉNERGIES ALTERNATIVES DSM/IRFU/SACM CENTRE DE SACLAY GIF-SUR-YVETTE CEDEX 2 JUILLET Etablissement 2014 public à caractère industriel et commercial RCS paris B PAGE 15