
247
ENSI Erfahrungs- und Forschungsbericht 2011
head dominates over thermal expansion in the flow
limiter area. In Figure 2b the change in distances
between the points with respect to time. The max-
imum distance between the edge of the flow lim-
iter and the IGT casing is given at 0.25 mm. Thus,
clamping is defined as
Di (t) -0.5 mm, for any
pair i at any time t > 0 but before the nozzle welds
fail. Analysis of results presented in Figure 2b sug-
gests that at given melt pool depth of 1.9 m and a
location of the IGT close to the bottom center, the
IGT is not clamped in its housing during the entire
time until global vessel wall failure takes place. This
result suggests that in the considered case of 1.9 m
melt pool, the dominant vessel failure mode is IGT
failure and it happens at least 1 hour before the
global vessel failure. Other melt pool depths and/or
other locations of the IGT might result otherwise,
but are not covered here and will be the subject of
further inquiry. In another piece of work the effect
of low fluid Prandtl number (Pr) on heat transfer
from a melt pool to the vessel wall has been exam-
ined [3]. The integral effect of the local heat trans-
fer enhancement due to corium melt low Prandtl
number is found to be significant, and the vessel
wall temperature is considerably increased. How-
ever, we also found that this effect occurs later
than the rapid acceleration of vessel creep in con-
sidered case of 0.7 m deep melt pool. Thus global
vessel failure is not sensitive to the local heat trans-
fer enhancement caused by fluid low Pr.
2. Progress in DEFOR Activity
A conservative-mechanistic approach to predict
the debris agglomeration is proposed, based on
simulations of the VAPEX FCI code and validated
by the DEFOR-A experimental data (see Figure 3).
It is assumed that mass fraction of agglomerated
debris is proportional to the mass fraction of com-
pletely liquid droplets and thin-crust particles («liq-
uid» or «glue» particles).
Figure 1:
(a) Mesh of the IGT
section with 11553
tetrahedral elements
and 18435 nodes,
and (b) geometry with
the location of the flow
limiter.
Figure 2:
(a) Displacement of points
along the flow limiter area
and four pairs of points for
which the distances bet-
ween points are measured
with respect to time;
(b) The change in distances
D
i
(t) = D
i
(t) – D
i
(t
0
) for
each pair i and time t
where D
i
(t
0
) is the distance
between the points in pair
i at time t = 0. The vertical
lines correspond to the
estimated time of nozzle
welds creep acceleration,
nozzle welds melting, and
global vessel wall creep
acceleration.
(a) (b)
(a) (b)
Comentarios a estos manuales