VALIDATION CASE:
LINE CHILLDOWN USING LIQUID HYDROGEN
This
validation case compares SINDA/FLUINT predictions
with a 1966 test by the National Bureau of Standards
(NBS, now National Institute of Standards and Technology,
NIST). This is an update to a previous comparison
performed in 1988. The updated comparison uses SINDA/FLUINT
4.7 and also makes use of C&R’s Thermal
Desktop® and FloCAD® to facilitate model building
and postprocessing.
In
the NBS tests, a pressurized LH2 tank was used with
a volume (300 liters) sufficient to assure nearly
constant supply temperature and pressure. The tank
was isolated from an empty line (open to the atmosphere)
by a series of valves. The line was vacuum-jacketed
3/4” and made with an unspecified copper alloy.
At time zero, a valve was opened (presumably the furthest
downstream) and LH2 was allowed to flow until the
line was completely full and liquid hydrogen was discharged
from exhaust end of the pipe.
Comparison
with Test Data
The transient profile below presents the main results
of this paper: a comparison of SINDA/FLUINT predictions
with NBS test data. The predictions show the inlet
side of the line filling slightly faster than actually
happened in test, but overall the comparison is excellent,
especially for the parameter that matters most: the
total time (and therefore LH2 expended) to completely
chilldown the line.
It
is notoriously difficult to make comparisons with
tests performed by third parties in advance of analysis.
This case turns out to be the exception: almost all
uncertainties turned out to have little to no effect
on the final results. The two exceptions to this statement
are the following unknown parameters, the specific
heat of copper and the wall roughness factor.
Specific
Heat--Variation of specific heat with temperature
was critical to achieving a successful comparison.
This sensitivity is to be expected considering the
nature of the problem: the quenching of the pipe wall
over a 250K range of sensible cooling. This model
used a Cp vs. T profile for OFHC (oxygen free high
conductivity) copper, a specialty alloy not normally
available for the off-the-shelf tubing that was almost
certainly used by NBS. Nonetheless, lacking better
data no adjustments to this data were made. Scaling
factors applied to this profile will, however, cause
the final temperature histories to shift noticeably
left and right (in time).
Wall
Roughness--Another nonnegligible uncertainty
was wall roughness, although it had a small effect.
The results shown in the prior section were made using
e/D = 0.0004 (WRF in FLUINT). This value is only slightly
higher than should be experienced in off-the-shelf
copper tubing (e/D of about 10-4), but this higher
value may be justified on the basis of joints or seams
not described in the NBS report. Note that assuming
a smooth wall (e/D=0) speeds up the chill-down time
by only 3.5%, so the uncertainty in this parameter
is still not significant.
Additional
information available for download:
Customization
and Consulting
C&R
also provides consulting and
custom software solutions to specifically meet
your needs. |
