Information Technology Reference
In-Depth Information
(
)
=++
2
1
c
,
U
B Bwksk
k
=−
w
τ
r
,
r
r
,
r
c
r
which substituted in Eq. (21b) implies that
(
)
(22)
FqBx
=−
τ
=
q
τ
B B
τ
rc
,
r
,
c
,
r
,
c
,
c
,
r
with the form Eq. (21b), meaning that
τ
and B correspond to functions
β
and
. Expression (22) was first obtained by Plebañski [6].
Now we shall consider the eigenvalue problem of F . For this purpose,
ψ
we stem from Eq. (21b), and due to this
r
kk
Uk
r
=−
1
=
0
,
Then, we immediately have one of the two null proper vectors of a nonnull
field (different type from C ) [23]:
proper value
=
qw
2
F
k
m
=
qw
2
k
,
(23a)
rm
r
This suggests that U r may be an eigenvector, but it is not:
(
)
b
2
2
2
2
FU
=−
qw U
qw
a
B k
,
(23b)
rb
r
r
(
)
Nevertheless, if we multiply Eq. (23a) by
and add the resulting
equation to Eq. (23b), we obtain the other null proper vector [24, 25]:
2
2
1
2
aB
m
2
qw
2
F
η
=−
qw
η
,
proper value
=−
with
(23c)
rm
r
(
)
η η
r
=
0
η
r
=+
Ua
r
2
Bk
2
r
1
2
,
;
r
It is not usual to find
explicitly in the literature. It is clear that these two
proper vectors are independent because:
η
r
k
r
η
=
1
(23d)
r
γ
r
Remember that two null vectors
r
and
are proportional if and only
ξ
r
if
ξγ =
0
, therefore Eq. (23d) implies the nonparallelism of such proper
r
vectors.
With Eqs. (10) and (21b), it is possible to prove that
 
Search WWH ::




Custom Search