2.11 Example Metric For 3D Flat Space

2.
11 Example metric for 3D flat space

lets do an example to make this all less abstract! The position vector between
two points close together in flat 3D space is simply dr = dxi + dyj + dzk.
So the distance between these two points is ds2 = dr.dr = dx2 + dy 2 + dz 2
which is exactly what we expect from Pythagoros!
and we can do this our new was as well as the dot product of basis vectors.
these are e1 = i, e2 = j, e3 = k where x1 = x, x2 = y and x3 = z. then
g11 = gxx = i.i = 1, g22 = gyy = j.j = 1 and g33 = gzz = k.k = 1, and all
cross terms are zero. so we get our distance by expanding our double sum
ds2 = (gij dxi )dxj explicitally as
= (g1j dx1 + g2j dx2 + g3j dx3 )dxj
= g1j dx1 dxj + g2j dx2 dxj + g3j dx3 dxj

= g11 dx1 dx1 + g12 dx1 dx2 + g13 dx1 dx3 +
g21 dx2 dx1 + g22 dx2 dx2 + g23 dx2 dx3 +
g31 dx1 dx1 + g32 dx1 dx2 + g33 dx3 dx3
all the gij = 0 except where i = j for this coordinate system so
= g11 dx1 dx1 + g22 dx2 dx2 + g33 dx3 dx3 = dx2 + dy 2 + dz 2
but we also looked at how we could do a coordinate transformation to

spherical polar coordinates, where x = r sin θ cos φ, y = r sin θ sin φ and
z = r cos θ. last lecture we saw that this gave us a new set of basis vectors,
er , eθ and eφ which we could define in terms of our old basis vectors i,j,k as
∂r
er = = sin θ cos φ i + sin θ sin φ j + cos θ k
∂r
∂r
eθ = = r cos θ cos φ i + r cos θ sin φ j − r sin θk
∂θ
∂r
eφ = = −r sin θ sin φ i + r sin θ cos φ j
∂φ
from above we can calculate how distance depends on these new coordinates
by taking the dot products
grr = er .er = sin θ2 cos2 φ + sin θ2 sin2 φ + cos2 θ = sin2 θ + cos2 θ = 1
1
gθθ = eθ .eθ = r 2 cos2 θ cos2 φ+r 2 cos2 θ sin2 φ+(−r)2 sin2 θ = r 2 cos2 θ+r 2 sin2 θ = r 2
gφφ = (−r)2 sin2 θ sin2 φ + r 2 sin2 θ cos2 φ = r 2 sin2 θ
there are also the cross terms grθ = gθr etc. but these are all zero. so then
our utterly general (and completely opaque!) distance is ds2 = gij dxi dxj .
Expanding out this double sum explicitally gives
ds2 = (gij dxi )dxj = (g1j dx1 + g2j dx2 + g3j dx3 )dxj
= g1j dx1 dxj + g2j dx2 dxj + g3j dx3 dxj
= g11 dx1 dx1 + g12 dx1 dx2 + g13 dx1 dx3 +
g21 dx2 dx1 + g22 dx2 dx2 + g23 dx2 dx3 +
g31 dx1 dx1 + g32 dx1 dx2 + g33 dx3 dx3
this coordinate system is x1 = r, x2 = θ and x3 = φ and in in this system all
the gij = 0 if i 6= j. so we can throw out most of these terms to get
ds2 = grr dr 2 + gθθ dθ2 + gφφ dφ2
ds2 = dr 2 + r 2 dθ2 + r 2 sin2 θdφ2
which is instantly recognissable as the distance between two points in 3D flat
space in spherical polar coordinates. So this abstract mathematical machi-
nary really does connect to what we already know!
Curvature is completely defined by the metric tensor! its the property
of the space, how distance relates to position. BUT, we still have a way to
go as this is NOT the sort of way we want to define curvature. it contains
the important information about the real curvature BUT it also contains
extraneous information about the coordinate system we are using. Our first
cartesian set of coordinates obviously gave us flat space but the second is
more subtle. at first glance you might think that it represents a curved
surface as distances are no longer doing a Pythagoros law - the metric tensor
components are not given by a diagonal, unit matrix. BUT - there exists
a coordinate transformation that gets us back to a metric in the form of
ds2 = dx2 + dy 2 + dz 2 - the underlying space is FLAT.
The difference in a real curved space is that there is no transformation we
can make to get this metric back to a flat space one. But its not necessarily
immediately apparent from the components of the metric tensor which ones
will allow coordinate transformations to get us to the unit matrix.
2
2.12 Kronekar delta and invariance of tensor equations
we saw that the basis vectors transform as eb = ∂xa /∂xb ea . This means that
any quantity A = Aa ea in another frame,
b ∂xb a ∂xd
A eb = a A ed
∂x ∂xb
where we changed a to d in the implied sum for the basis vectors as we’ve
already used a - we can call them anything we like in a sum, but they are
BOTH sums. So finally we get
∂xd a
A ed = δad Aa ed = Aa ea
∂xa
where the δba = 1 for a = b and 0 otherwise is termed the Kronekar delta
function. So if we have something that transforms as the coordinate differ-
ences, then this means its tensor equation looks the same in ANY frame!!!
A quick way to see if you have dropped any indices is to count them -
cancel out the indices which have an impled sum, and then the indices we
have on the left hand side MUST be equal to the indices we have on the
right.
2.13 Basis vectors for covariant components

Covariant components came from ∇φ - but this in cartesian coordinates is
just
∂φ ∂φ ∂φ
∇φ = i+ j+ k = Ai ei
∂x ∂y ∂z
where Ai = ∂φ/∂xi are the components, and since we know that these are
covariant, the basis vectors must have the index high.
and we would get these basis vectors in a different way - e1 = i = ∇x1 ,
e = j = ∇x2 and e3 = k = ∇x3
2
so the basis vectors are the same as before, but we’d get them in a different
way. ei = ∇xi which is not necessarily equal to ei = ∂r/∂xi .
show what happens if we go to a new coordinate frame xj
∂φ i
∇φ = e
∂xi
3
∂φ ∂xj i
= e
∂xj ∂xi
∂φ
= j ∇xj
∂x
so this tells us how these new basis vectors transform
∂xj i
ej = ∇xj = e
∂xi
and then A = Aa ea will transform to another frame as
∂xb ∂xb c ∂xc

Ab eb = Aa e = Aa ec = δac Aa ec = Aa ea
∂xa ∂xc ∂xa
So our tensor equations look the same in all frames!
FOR MATHEMATICALLY INCLINED PEOPLE ONLY!! This gives us
another vector space of the manifold - call it TP∗ the dual or cotangent space
of the manifold at P as opposed to our first set which defined the tangent
space TP to the manifold at P
Back to everyone. The basis vectors themselves look identical if we have
an orthonormal set of coordinates (they don’t have to be reclilinear, just be
90 degrees where they meet). BUT THEY ARE NOT IDENTICAL IF THE
COORDINATES ARE NOT ORTHOGANAL.
e.g. in 3D spherical polars r 2 = x2 + y 2 + z 2 so
∂r ∂r ∂r
er = ∇r = i+ j+ k
∂x ∂x ∂z
easy way is to say ∂r 2 /∂x = 2r∂r/∂x so ∂r/∂x = 1/(2r)×∂r 2 /∂x = 1/(2r)×
∂(x2 + y 2 + z 2 )/∂x = 2x/(2r) = x/r
similarly ∂r/∂y = y/r and ∂r/∂z = z/r
so er = (x/r)i + (y/r)j + (z/r)k. but we know x = r sin θ cos φ etc so
er = sin θ cos φi + sin θ sin φj + cos θk
etc for the rest.

2.11 Example Metric For 3D Flat Space

Uploaded by

Copyright:

Available Formats

2.11 Example Metric For 3D Flat Space

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2.11 Example Metric For 3D Flat Space

Uploaded by

Copyright:

Available Formats

2.

11 Example metric for 3D flat space

= (g1j dx1 + g2j dx2 + g3j dx3 )dxj

= g1j dx1 dxj + g2j dx2 dxj + g3j dx3 dxj

but we also looked at how we could do a coordinate transformation to

grr = er .er = sin θ2 cos2 φ + sin θ2 sin2 φ + cos2 θ = sin2 θ + cos2 θ = 1

2.13 Basis vectors for covariant components

∂xb ∂xb c ∂xc

You might also like