9 views

Uploaded by Faisal Rahman

tensor

save

- generalizedcoords
- Full Text of _An Introduction to Tensor Calculus & Relativity
- 0308005
- General Relativity in a mathematical nutshell
- Notes 05 Coordinates Dirac Delta
- Lecture 22
- TensorCalculus Maple
- A 0340107
- L10 Elastic Aniso 3Oct11 (1)
- Fundamental Solutions in the Theory of Microfluids
- 554midterm2 Solutions
- Product Specifications(2)
- 3. Stress
- PART2
- Introduction to Tensor Calculus
- Strain Rate
- What is Gradient
- ANSYS TOPOPT
- lab4_ionization2015
- 220-HW7
- Taylor
- Jose P. S. Lemos and Francisco S. N. Lobo- Plane symmetric traversable wormholes in an anti-de Sitter background
- Vasp manual
- 1701.01262
- Mathematical Economics w 04
- Pannel Method
- Eigenvalues
- Boura Tinos 2006
- Sahyd c17 Eng
- Forces in General Relativity
- 1167 Hochschulwerbung Brunel Absolv
- 1204 Enterprise - Praktikum
- Flyer Master
- Chapter 4
- 1195 Axel Springer - Shk
- Tensor
- 001 CFD Foundation
- ch6
- Chapter 4
- Splines
- 08 Finite Elements Basisfunctions
- 1179 Mrs.sporty Qual Stud
- Arrays
- MATH45061_Ch2
- German Cv
- Master Wiwi
- Interpolation Chap 3
- Matrices
- Programming in Maple the Basics
- Nst Mmii Chapter3
- Mat Lab Basics Turner
- Tensor Techniques in Physics
- Cubic Spline Example
- CH03_5C_4
- Vrr
- Master Wiing
- Tensor Calculus
- 15 Trace
- Lecture 5
- l304_2

You are on page 1of 5

Dr. Christian Salas November 15, 2009

1

Introduction

In this note I try to provide some intuition for the properties of tensors and in particular why the metric tensor can be used to raise and lower indices when applied to other tensors.

2

Deﬁnition of tensors in terms of vectors

It is helpful to think of all tensors as being composed of contravariant and/or covariant vectors. A contravariant vector Ak deﬁned with regard to a system ¯k deﬁned with of coordinates {xk } is one which is transformed into a vector A regard to a system of coordinates {z k } by the transformation law

k ¯k = ∂z Am A ∂xm

(k, m = 1, . . . , N )

(1)

Contravariant vectors are always identiﬁed by a single upper index. In con¯k by trast, a covariant vector Bk is one which is transformed into a vector B the transformation law

m ¯k = ∂x Bm B ∂z k

(k, m = 1, . . . , N )

(2)

Covariant vectors are always identiﬁed with a single lower index. We can now deﬁne all types of tensors in terms of these two vectors. For example, a second-order contravariant tensor T uv with respect to the coordinate system {xk } can be deﬁned as T uv = Au B v 1 (u, v = 1, . . . , N ) (3)

x2 . . 2. . N ) (10) (11) 2 . a third-order mixed tensor with two contravariant components and one covariant component can be deﬁned as uv Tw = Au B v C w (u. . . 3 The metric tensor ds2 = (dy 1 )2 + (dy 2 )2 + · · · + (dy N )2 = δjk dy j dy k (9) The metric of the usual Euclidean metric space can be written as If instead of the Descartes orthogonal coordinates {y k } we introduce the arbitrary generalised coordinates {xk } by means of the equations y k = y k (x1 . N ) (7) Using (1) and (2) we see that it transforms according to the law u v p v p u uv ¯w ¯u B ¯vC ¯w = ∂z ∂z ∂x Am B n Cp = ∂z ∂z ∂x T mn T =A ∂xm ∂xn ∂z w ∂xm ∂xn ∂z w p (8) In the same way. . v. w = 1.This can be thought of as the N 2 matrix T = A ⊗ B. N ) (5) and using (2) we see that it transforms into a second-order covariant tensor ¯uv with respect to the coordinate system {z k } according to the law T m m n n ¯uv = A ¯u B ¯v = ∂x ∂x Am Bn = ∂x ∂x Tmn T ∂z u ∂z v ∂z u ∂z v (6) We can deﬁne tensors of any other order and type by adding vector factors. xN ) then we can write ∂y k 1 ∂y k 2 ∂y k N ∂y k m dx + 2 dx + · · · + N dx = dx dy = ∂x1 ∂x ∂x ∂xm k (k = 1. . Using (1) we see that ¯uv with respect T uv is transformed into a second-order contravariant tensor T to the coordinate system {z k } by the transformation law v v u u ¯uv = A ¯u B ¯ v = ∂z ∂z Am B n = ∂z ∂z T mn T ∂xm ∂xn ∂xm ∂xn (4) Similarly. v = 1. . . . . we can deﬁne a second-order covariant tensor Tuv with respect to the coordinate system {xk } as Tuv = Au Bv (u. we can deﬁne tensors of arbitrary order and type. . . . . For example. . .

. . dxN ∂y N ∂y N ∂xN ∂x1 ∂y N ∂y N ∂xN ∂x2 ∂y N ∂y N ∂xN ∂xN = dx1 dx2 · · · dxN N i=1 N i=1 . ∂y i ∂xN ∂y i ∂xN dx1 dx2 . . . . . ··· ∂y 1 ∂x1 ∂y 1 ∂x2 . ∂y N ∂x2 ∂y N ∂x2 ··· ··· . . . . . . . . . ··· ∂y 2 ∂x1 ∂y 2 ∂x2 . . . . . . . ∂y N ∂xN ∂y N ∂xN dx1 dx2 . ··· N i=1 N i=1 ∂y i ∂x1 ∂y i ∂x2 . . . ··· ∂y N ∂x1 ∂y N ∂x2 . 2. . . . . . . . ∂y 1 ∂x1 ∂y 1 ∂x1 ∂y 1 ∂x1 ∂y 1 ∂x2 . . dxN ∂y 2 ∂y 2 ∂xN ∂x1 ∂y 2 ∂y 2 ∂xN ∂x2 ∂y 2 ∂y 2 ∂xN ∂xN + · · ·+ dx1 dx2 · · · dxN ∂y N ∂x1 ∂y N ∂x2 . . . ∂y i ∂x1 ∂y i ∂x2 ∂y i ∂x1 ∂y i ∂x1 N i=1 N i=1 . ∂y 2 ∂x2 ∂y 2 ∂x2 ··· ··· . ∂y i ∂x1 ∂y i ∂x2 ∂y i ∂x2 ∂y i ∂x2 ··· ··· . . ∂y 2 ∂x1 ∂y 2 ∂x1 ∂y 2 ∂x1 ∂y 2 ∂x2 . dxN ∂y 1 ∂y 1 ∂xN ∂x1 ∂y 1 ∂y 1 ∂xN ∂x2 ∂y 1 ∂y 1 ∂xN ∂xN + dx 1 dx 2 ··· dx N ∂y 2 ∂x1 ∂y 2 ∂x2 . . .for k = 1. dxN N ∂y i ∂y i i=1 ∂xN ∂x1 N ∂y i ∂y i i=1 ∂xN ∂x2 N ∂y i ∂y i i=1 ∂xN ∂xN This can be written more compactly in tensor notation as ds2 = gmn dxm dxn 3 (12) . . and the metric form can be written as ds2 = ∂y 1 1 ∂y 1 2 ∂y 1 N dx + dx + · · · + dx ∂x1 ∂x2 ∂xN 2 + ∂y 2 1 ∂y 2 2 ∂y 2 N dx + dx + · · · + dx ∂x1 ∂x2 ∂xN 2 2 ∂y N 1 ∂y N 2 ∂y N N + ··· + dx + dx + · · · + N dx ∂x1 ∂x2 ∂x = dx1 dx2 · · · dxN ∂y 1 ∂x1 ∂y 1 ∂x2 . . ∂y 1 ∂xN ∂y 1 ∂xN dx1 dx2 . . ∂y 1 ∂x2 ∂y 1 ∂x2 ··· ··· . . . N . . . ∂y N ∂x1 ∂y N ∂x1 ∂y N ∂x1 ∂y N ∂x2 . . . ∂y 2 ∂xN ∂y 2 ∂xN dx1 dx2 .

. gmn = δjk N ∂y i ∂y i i=1 ∂xN ∂x1 N ∂y i ∂y i i=1 ∂xN ∂x2 ··· N ∂y i ∂y i i=1 ∂xN ∂xN 4 Raising and lowering indices gmn An = Am (15) Premultiplying a contravariant vector by gmn lowers the index i. Now. 0 4 0 ··· 0 0 . . . . . . .e. we have g mn An = Am (17) To get an intuitive feeling for why this is so. 0 0 · · · gN N Then the contravariant metric tensor is simply 1 0 ··· g11 1 0 ··· g22 mn −1 [g ] = [gmn ] = . the inverse of the (covariant) metric tensor gmn gives the contravariant metric tensor g mn . . . . . . . . . (18) . . . . . . . . . g11 0 · · · 0 0 g22 · · · 0 [gmn ] = . . This has the eﬀect of raising the index on a covariant vector i.e. .where ∂y j ∂y k (13) ∂xm ∂xn The symmetric second-order system gmn is a covariant second-order tensor called the metric tensor. the full matrix form of gmn is N ∂y i ∂y i N ∂y i ∂y i N ∂y i ∂y i · · · 1 1 1 2 i=1 ∂x ∂x i=1 ∂x1 ∂xN i=1 ∂x ∂x N ∂y i ∂y i N ∂y i ∂y i N ∂y i ∂y i ··· i=1 ∂x2 ∂x1 i=1 ∂x2 ∂x2 i=1 ∂x2 ∂xN [gmn ] = (14) .e. . As can be seen from the calculations above. suppose we are dealing with an orthogonal system so that gmn is diagonal i. . . . . 1 gN N (19) . we have We can see that this is so by writing gmn An = δjk ∂y j ¯k ∂y k ¯k ∂y j ∂y k n A = δ A = A = Am jk ∂xm ∂xn ∂xm ∂xm (16) (by the transformation law for covariant vectors). .

A typical non-zero element is g so we have g mm Am = mm = 1 gmm ∂xm ∂xm = ∂y j ∂y j (20) ∂xm ∂xm ∂xm ¯ A = Aj = Am m ∂y j ∂y j ∂y j (21) (by the transformation law for contravariant vectors). Now. we can apply this raising and lowering of indices technique to tensors by remembering the deﬁnition of tensors in terms of contravariant and covariant vectors. (END OF NOTE) 5 . For example. we can write uv g mk Tk = T uvm (22) because uv g mk Tk = Au B v g mk Ck = Au B v C m = T uvm (23) The same applies to tensors of arbitrary order and type.

- generalizedcoordsUploaded bymagixter
- Full Text of _An Introduction to Tensor Calculus & RelativityUploaded byanamen
- 0308005Uploaded byMaiman Lato
- General Relativity in a mathematical nutshellUploaded byBernard SCHAEFFER
- Notes 05 Coordinates Dirac DeltaUploaded byPratikGujarathi
- Lecture 22Uploaded byDavid Rinaldi
- TensorCalculus MapleUploaded bylatec
- A 0340107Uploaded byInternational Organization of Scientific Research (IOSR)
- L10 Elastic Aniso 3Oct11 (1)Uploaded byShafeeq Shabaz
- Fundamental Solutions in the Theory of MicrofluidsUploaded byMeng-Chien Wu
- 554midterm2 SolutionsUploaded byCarlos Eduardo
- Product Specifications(2)Uploaded byHoward Ang
- 3. StressUploaded byAnshar Abdullah Jawil
- PART2Uploaded byCHINNU_HIDAYATH
- Introduction to Tensor CalculusUploaded byAbe
- Strain RateUploaded byzidaaan
- What is GradientUploaded bysezenismail
- ANSYS TOPOPTUploaded bynetkasia
- lab4_ionization2015Uploaded byElsa Novi Yanti
- 220-HW7Uploaded bye076016
- TaylorUploaded byGeorge Mathai
- Jose P. S. Lemos and Francisco S. N. Lobo- Plane symmetric traversable wormholes in an anti-de Sitter backgroundUploaded byRtpom
- Vasp manualUploaded byDavey
- 1701.01262Uploaded byCroco Ali
- Mathematical Economics w 04Uploaded byAsiiyah
- Pannel MethodUploaded bytommaso850
- EigenvaluesUploaded byinovabiotec5588
- Boura Tinos 2006Uploaded byBob
- Sahyd c17 EngUploaded byAnonymous v7XdaQu
- Forces in General RelativityUploaded byMarco Jones

- 1167 Hochschulwerbung Brunel AbsolvUploaded byFaisal Rahman
- 1204 Enterprise - PraktikumUploaded byFaisal Rahman
- Flyer MasterUploaded byFaisal Rahman
- Chapter 4Uploaded byFaisal Rahman
- 1195 Axel Springer - ShkUploaded byFaisal Rahman
- TensorUploaded byFaisal Rahman
- 001 CFD FoundationUploaded byFaisal Rahman
- ch6Uploaded byFaisal Rahman
- Chapter 4Uploaded byFaisal Rahman
- SplinesUploaded byFaisal Rahman
- 08 Finite Elements BasisfunctionsUploaded byFaisal Rahman
- 1179 Mrs.sporty Qual StudUploaded byFaisal Rahman
- ArraysUploaded byFaisal Rahman
- MATH45061_Ch2Uploaded byFaisal Rahman
- German CvUploaded byFaisal Rahman
- Master WiwiUploaded byFaisal Rahman
- Interpolation Chap 3Uploaded byFaisal Rahman
- MatricesUploaded byFaisal Rahman
- Programming in Maple the BasicsUploaded byFaisal Rahman
- Nst Mmii Chapter3Uploaded byFaisal Rahman
- Mat Lab Basics TurnerUploaded byFaisal Rahman
- Tensor Techniques in PhysicsUploaded byFaisal Rahman
- Cubic Spline ExampleUploaded byFaisal Rahman
- CH03_5C_4Uploaded byFaisal Rahman
- VrrUploaded byFaisal Rahman
- Master WiingUploaded byFaisal Rahman
- Tensor CalculusUploaded byFaisal Rahman
- 15 TraceUploaded byFaisal Rahman
- Lecture 5Uploaded byFaisal Rahman
- l304_2Uploaded byFaisal Rahman