Jökull - 01.12.1966, Qupperneq 41

unity, and K assumed in this range. A value of
n can then be derived on the basis of equation
(33). Since the maximum value of e has been
assumed, the value of n computed in this way
will be a low estimate of the actual expansion
coefficient in the experiments of James (1962).
An indication of the behavior of the steam is
obtained by comparison of this exponent with
one of the following five exponents each of
which defines a recognized ideal type of expan-
sion
1) Supersaturated isentropic expansion,
ds = 0, (S)
2) Supersaturated isenthalpic expansion,
dh0 = 0, (S)
3) Equilibrium isentropic expansion,
ds = 0, (E)
4) Equilibrium isenthalpic expansion,
dh0 = 0, (E)
5) Polytropic expansion according to equation
(27) with an exponent np calculated on the
basis of equation (A. 7) in the appendix.
The exponents in (1)—(4) can be calculatecl
directly from Steam Tables.
Figs. 8, (a)—(f) display the calculated values
of n for comparison with the values listed in
(l)—(5) above. Inspection of these curves reveals
the following items of interest.
1) For Fig. 8a where h0 = 376 Btu/lb, the
flow pattern is known (Ryley, 1964) to be
annular-mist, whereas for Figs. 8, b—f, it
is dispersed. The positions of the curves
for Fig. 8a suggest than n is more sensitive
to K than in the remaining cases. This is
to be expected because of the relatively
very low values of x involved.
2) For all the curves, the higher the value
of p, the greater is gg and the more finely
are liquid masses disrupted by the high
accelerative forces towards the exit. The
finer the mist, the greater the paddle work
done by the steam, the greater the energy
dissipation and the lower the value of the
exponent n.
3) For a given value of K and x, the void
fraction Rg will increase with pressure re-
duction and the expansion if rapid will
more nearly conform to the ideal index
n = 1.3, since liquid volume has a failing
influence.
4) As h0 (and therefore x) increases the effect
of K on the exponent n will diminish and
finally disappear as x approaches unity.
5) The proportional effect of pressure upon
the exponent n due to reheating conse-
quent on paddle work by the liquid, di-
minishes with rising values of h0.
6) In spite of the assumption e = 1, the de-
rived coefficient n exceeds np for all as-
sumed values of K in the cases of the
higher enthalpy values h0 = 578 to 1200
Btu/lb. This indicates that the actual ex-
pansivity of the mixture is lower than
would be expected on the basis of the
above assumptions. The discrepancy can
be explained as a result of thermal disequi-
librium at the exit. The expansion is too
, Figs- 8- (a)-(f).
Comparison of Calculated Values
of Expansion Exponent
with Those for Ideal Expansion Types.
JÖKULL 195