0% found this document useful (0 votes)
50 views18 pages

Covariant and Contravariant Vectors

This document introduces tensors and their properties under coordinate transformations. It defines contravariant and covariant vectors, which transform differently under rotations. Tensors are objects with multiple indices that transform according to the type (contravariant or covariant) of each index. The document discusses tensor operations like summation, outer products, and differentiation of tensors. It also covers metrics, raising and lowering indices, and tensors in Minkowski spacetime.

Uploaded by

lila
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
50 views18 pages

Covariant and Contravariant Vectors

This document introduces tensors and their properties under coordinate transformations. It defines contravariant and covariant vectors, which transform differently under rotations. Tensors are objects with multiple indices that transform according to the type (contravariant or covariant) of each index. The document discusses tensor operations like summation, outer products, and differentiation of tensors. It also covers metrics, raising and lowering indices, and tensors in Minkowski spacetime.

Uploaded by

lila
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

IntroductiontoTensors

Contravariantandcovariantvectors
Rotationin2space:

x'=cosx+siny
y'=sinx+cosy

Tofacilitategeneralization,replace(x,y)with(x1,x2)
Prototypecontravariantvector: dr=(dx1,dx2)
=cosdx1+sindx2
Similarlyfor
Sameholdsforr,sincetransformationislinear.

2
Compactnotation:
(generalizestoanytransformationinaspaceofanydimension)
Contravariantvector:
Nowconsiderascalarfield(r):Howdoestransformunder
rotations?

appearsratherthan

ForrotationsinEuclideannspace:
where=anglebtwn

and

axes

Itisnotthecaseforallspacesandtransformationsthat
sowedefineanewtypeofvectorthattransformslikethegradient:
Covariantvectors:

ExplicitdemonstrationforrotationsinEuclidean2space:

WhataboutvectorsinMinkowskispace?

but

=>contravariantandcovariant
vectorsaredifferent!

Recap(forarbitraryspaceandtransformation)
Contravariantvector:
Covariantvector:
Forfutureconvenience,definenewnotationforpartialderivatives:

Note:
=Kroneckerdelta=1ifi=j,0ifij

Tensors
ConsideranNdimensionalspace(witharbitrarygeometry)and
anobjectwithcomponents
inthe

inthe

coordsystemand

coordsystem.

Thisobjectisamixedtensor,contravariantini...kandcovariantin
l...n,underthecoordtransformation

if

Rankoftensor,M=numberofindices
Totalnumberofcomponents=NM
Vectorsarefirstranktensorsandscalarsarezeroranktensors.

IfspaceisEuclideanNspaceandtransformationisrotationof
Cartesiancoords,thentensoriscalledaCartesiantensor.
InMinkowskispaceandunderPoincartransformations,tensorsare
Lorentztensors,or,4tensors.
Zerotensor0hasallitscomponentszeroinallcoordsystems.
Maintheoremoftensoranalysis:
Iftwotensorsofthesametypehavealltheircomponentsequalin
onecoordsystem,thentheircomponentsareequalinallcoordsystems.
Einstein'ssummationconvention:repeatedupperandlower
indices=>summation
e.g.:

AiBicouldalsobewrittenAjBj;indexisadummyindex
Anotherexample:
jandkaredummyindices;iisafreeindex
Summationconventionalsoemployedwith

,etc.

Exampleofasecondranktensor:Kroneckerdelta

SP5.12

TensorAlgebra(operationsformakingnewtensorsfromoldtensors)
[Link]:addcomponents:
Proofthatsum
isatensor:
(foronecase)

[Link]:multiplycomponents:e.g.,
[Link]:replaceonesuperscriptandonesubscriptbya
dummyindexpair
e.g.,
Resultisascalarifnofreeindicesremain.
e.g,

10

11

[Link]:contractioninconjunctionwithouterproduct
e.g.:
Again,resultisascalarifnofreeindicesremain,e.g,
[Link]:e.g.,
SP5.35
DifferentiationofTensors
Notation:

,etc.

12

IFtransformationislinear
(sothatp'sareallconstant)
=>derivativeofatensorwrtacoordinateisatensoronlyfor
lineartransformations(likerotationsandLTs)
Similarly,differentiationwrtascalar(e.g.,)yieldsatensorfor
lineartransformations.

13

NowspecializetoRiemannianspaces
characterizedbyametric
Assume

If

with

issymmetric:

(nolossofgenerality,since
theyonlyappearinpairs)
,thenspaceisstrictlyRiemannian
(e.g.,EuclideanNspace)

Otherwise,spaceispseudoRiemannian(e.g.,Minkowskispace)
iscalledthemetrictensor.
Notethatthemetrictensormaybeafunctionofpositioninthespace.

Proofthat

14

isatensor:
(sincedxiisavector)
(2setsofdummyindices)

=>

It'stemptingtodivideby

andconclude

Butthere'sadoublesumoverk'andl',sothisisn'tpossible.
Instead,suppose
=>

=1ifi'=1
=0otherwise
Similarlyfor

,etc.

15
Nowsuppose

=1ifi'=1or2
=0otherwise

Onlycontributingtermsare:k'=1,l'=1k'=1,l'=2
k'=2,l'=1k'=2,l'=2
0

since

issymmetric.
sinceiandjaredummyindices.

=>

Similarlyforall

16

Generaldefinitionofthescalarproduct:
Define

astheinversematrixof

isalsoatensor,sinceapplyingtensortransformationyields
,whichdefines

astheinverseof

Raisingandloweringofindices:anothertensoralgebraicoperation,
definedforRiemannianspaces=innerproductofatensorwiththe
metrictensor
e.g.:

Note:covariantandcontravariantindicesmustbestaggeredwhen
raisingandloweringisanticipated.

4tensors
InallcoordsystemsinMinkowskispace:

=>
e.g:

17

18

UnderstandardLorentztransformations:

Alltheotherp'sarezero.

e.g.:

SP5.610

You might also like