Skip to content
SEINA-3D

What the module models

As a high-temperature melt jet penetrates a water pool, shear along the jet column strips off coarse droplets, hydrodynamic forces break liquid droplets into smaller sizes until surface solidification arrests deformation, and film boiling together with thermal radiation generates a rising vapor envelope. In SEINA-3D v1, this module supplies the interfield mass, interfacial area, and thermal exchange rates connecting the melt fields with liquid water and steam during premixing.

Governing equations

Connected-jet stripping and leading-edge breakup criterion

Along the lateral surface AjA_{\mathrm{j}} of the connected melt jet inside a control volume VV, shear stripping transfers melt mass and creates coarse-droplet interfacial area concentration ada_{\mathrm{d}} at rates:

Γjd=ρjv˙jd,v˙jd=UeAjV,a˙jd=6v˙jdDd,0\begin{aligned} \Gamma_{\mathrm{jd}} &= \rho_{\mathrm{j}} \dot{v}_{\mathrm{jd}}, \quad \dot{v}_{\mathrm{jd}} = \frac{U_{\mathrm{e}} A_{\mathrm{j}}}{V}, \\ \dot{a}_{\mathrm{jd}} &= \frac{6 \dot{v}_{\mathrm{jd}}}{D_{\mathrm{d},0}} \end{aligned}

where Dd,0D_{\mathrm{d},0} is the initial stripped droplet diameter and the entrainment velocity takes Ue=0.03 m s−1U_{\mathrm{e}} = 0.03\ \mathrm{m\,s^{-1}} in bubbly and transition regimes and Ue=0.01 m s−1U_{\mathrm{e}} = 0.01\ \mathrm{m\,s^{-1}} in vapor-dominated droplet and jet regimes.

To determine the jet breakup length LbL_{\mathrm{b}} consistently in three-dimensional transient simulations, SEINA-3D v1 identifies the melt region connected to the inlet across six-neighbour cell faces together with piecewise linear interface calculation (PLIC) front reconstruction. The normalized leading-edge speed uf/u0u_{\mathrm{f}}/u_0 is sampled at dimensionless time intervals of 0.100.10 and fitted over a five-point window of width 0.400.40; breakup is identified at the onset of sustained tip deceleration where the slope magnitude stays below 0.400.40 across three consecutive intervals.

Hydrodynamic droplet fragmentation and solid-crust arrest

Coarse melt droplets subject to relative velocity ur=ud−uc\mathbf{u}_{\mathrm{r}} = \mathbf{u}_{\mathrm{d}} - \mathbf{u}_c undergo Pilch–Erdman hydrodynamic fragmentation once the Weber number exceeds Wecri=4π\mathit{We}_{\mathrm{cri}} = 4\pi:

dDddt=−0.245 ∣ur∣ρcρm,We=ρc∣ur∣2Ddσm>4π\begin{aligned} \frac{\mathrm{d}D_{\mathrm{d}}}{\mathrm{d}t} &= -0.245\, |\mathbf{u}_{\mathrm{r}}| \sqrt{\frac{\rho_c}{\rho_{\mathrm{m}}}}, \\ \mathit{We} &= \frac{\rho_c |\mathbf{u}_{\mathrm{r}}|^2 D_{\mathrm{d}}}{\sigma_{\mathrm{m}}} > 4\pi \end{aligned}

When radiative and boiling heat losses form a solidified outer shell of thickness δ\delta, elastic bending resistance restrains deformation. Fragmentation then requires the modified aeroelastic number Ae∗\mathit{Ae}^* (with shell elastic modulus EE and Poisson ratio ε\varepsilon) to exceed the critical threshold:

Ae∗=12(1−ε2)ρc∣ur∣2E(Ddδ)3>8π3\mathit{Ae}^* = \frac{12(1 - \varepsilon^2)\rho_c |\mathbf{u}_{\mathrm{r}}|^2}{E} \left(\frac{D_{\mathrm{d}}}{\delta}\right)^3 > 8\pi^3

Boiling, radiative heat transfer, and interfacial phase change

Convective and film-boiling heat transfer between dispersed melt qq and coolant phase cc follows Qqc=hqcaqc(Tq−Tc)Q_{qc} = h_{qc} a_{qc}(T_q - T_c). Thermal radiation across the vapor film is evaluated from a two-surface grey-body formulation weighted by local liquid fraction αl\alpha_{\mathrm{l}}:

Qrad=σSBadϵmϵlϵm+ϵl−ϵmϵl×(Tm4−Tl4)10αl1+10αl\begin{aligned} Q_{\mathrm{rad}} &= \sigma_{\mathrm{SB}} a_{\mathrm{d}} \frac{\epsilon_{\mathrm{m}}\epsilon_{\mathrm{l}}}{\epsilon_{\mathrm{m}} + \epsilon_{\mathrm{l}} - \epsilon_{\mathrm{m}}\epsilon_{\mathrm{l}}} \\ &\quad \times \left(T_{\mathrm{m}}^4 - T_{\mathrm{l}}^4\right) \frac{10\alpha_{\mathrm{l}}}{1 + 10\alpha_{\mathrm{l}}} \end{aligned}

Accounting for spectral absorption in water, 90%90\% of QradQ_{\mathrm{rad}} is deposited into the liquid bulk and 10%10\% acts directly at the vapor–liquid interface. Net heat reaching the saturation interface QintQ_{\mathrm{int}} determines the evaporation or condensation mass source Γlv\Gamma_{\mathrm{lv}}:

Qint=Qm→i+Qv→i−Qi→l,Γlv=QintHvint−Hlint\begin{aligned} Q_{\mathrm{int}} &= Q_{\mathrm{m}\to\mathrm{i}} + Q_{\mathrm{v}\to\mathrm{i}} - Q_{\mathrm{i}\to\mathrm{l}}, \\ \Gamma_{\mathrm{lv}} &= \frac{Q_{\mathrm{int}}}{H_{\mathrm{v}}^{\mathrm{int}} - H_{\mathrm{l}}^{\mathrm{int}}} \end{aligned}

Numerical method

Interfacial area concentration aqa_q is advected alongside phase volume fraction αq\alpha_q so local Sauter mean diameter Dq=6αq/aqD_q = 6\alpha_q / a_q tracks stripping and secondary breakup continuously across cells. Heat-transfer coefficients, crust growth, and phase-change rates are updated inside the outer pressure-correction loop at each time step, preventing explicit thermal overshoot when local void fraction changes rapidly around the melt jet.

Verification

  • KTH melt-jet breakup tests: validated across six jet breakup experiments (D0=4–20 mmD_0 = 4\text{--}20\ \mathrm{mm}, We=9–189.8\mathit{We} = 9\text{--}189.8) in Sun et al. (2026), achieving a mean absolute relative difference of 7.7%7.7\% in dimensionless breakup length and capturing how coolant subcooling modifies the velocity dependence of Lb/D0L_{\mathrm{b}}/D_0.
  • FARO L-24 prototypic corium–water experiment: validated in Sun et al. (2027) for 177 kg177\ \mathrm{kg} of 80 wt% UO2–20 wt% ZrO280\ \mathrm{wt\%}\ \mathrm{UO_2}\text{--}20\ \mathrm{wt\%}\ \mathrm{ZrO_2} melt at 3023 K3023\ \mathrm{K} pouring into 2.02 m2.02\ \mathrm{m} of water at 425 K425\ \mathrm{K} and 0.51 MPa0.51\ \mathrm{MPa} on a 108×108×324108 \times 108 \times 324 medium grid. In the steam-only calculation, predicted vessel pressure peaked at 2.16 MPa2.16\ \mathrm{MPa} compared with 2.25 MPa2.25\ \mathrm{MPa} in the experiment (root-mean-square error of 0.18 MPa0.18\ \mathrm{MPa} across 1515 comparison times), and the reconstructed median fragment size D50=2.6 mmD_{50} = 2.6\ \mathrm{mm} closely matched the measured 2.5 mm2.5\ \mathrm{mm}.