You are on page 1of 4

Mathematical Tripos Part III GENERAL RELATIVITY Problems 3: Lie derivatives, Killing vectors and linearized theory Please

send comments/amendments etc. to

Dr O. Rinne Michaelmas 2008

Copyright © 2008 University of Cambridge. Not to be quoted or reproduced without permission.

The Leibniz rule for Lie derivatives

Let V and ω be vector and covector fields and f a real function, all defined in the neighbourhood of a point p. Show that at p LX (f ω ) = (Xf )ω + f LX ω.


The Lie derivative and contraction

For any vector field X show that LX δ a b = 0. Deduce that Lie derivation commutes with contraction, i.e., if Y is a vector field and ω a covector field, then at any point p, LX ω, Y = LX ω, Y + ω, LX Y .


The Lie derivative of a covector

In the lectures, definitions of the Lie derivatives of scalars and vectors were given in direct manifestly chart-free terms. The definition of the Lie derivative of a covector was obtained using these and the Leibniz property. Write out a direct manifestly chart-free definition of the Lie derivative of a covector, and demonstrate its consistency with that of the lectures.


Lie derivatives in electrodynamics

Two of the Maxwell equations governing electrodynamics are ∇ · B = 0, ∂B = −∇ ∧ E, ∂t

and within a magnetohydrodynamic fluid of high conductivity E + v ∧ B = 0, where v is the velocity of the fluid. The fluid is incompressible, i.e., ∇ · v = 0. Show that in steady-state situations the 3-vector equation of motion for B can be expressed succinctly as Lv B = 0. For unsteady motions we define a non-relativistic 4-velocity V = ∂ ∂ + vα α , ∂t ∂x

and a non-relativistic 4-vector magnetic field B = (0, B). Show that the 3-vector equation of motion for B can be rewritten as LV B = 0 . [An open-ended extension is to examine your favourite branch of physics looking for other examples of the Lie derivative at work.]

1). ∂θ ∂φ Copyright © 2008 University of Cambridge.. when Q is either a function or a vector field.6 An identity for Killing vectors Show that Killing covectors satisfy the equation ∇a ∇b Kc = Rd abc Kd .5 The algebra of Killing vectors Let X and Y be two vector fields. Take Tij = ǫM δ (x)Ui Uj . i. ∂φ ∂ ∂ + cot θ cos φ . Compare it with the solution derived in the lectures using the scalar-vectortensor decomposition of gravitational perturbations. Using the result of problem 6. Show that LX (LY Q) − LY (LX Q) = L[X.e. . Write down Killing’s equation. Identify the associated conserved quantities along a timelike geodesic. sin φ are Killing vectors. 3. 0)T .General Relativity Problems 3 2 3. Let K be a Killing vector. 3. Demonstrate that if a space has two “independent” isometries then it has a third.] Deduce that in Minkowski spacetime the components of Killing covectors are linear functions of the coordinates. use the traditional approach to linearized theory to derive the linearized metric.7 Killing vectors and conserved quantities in Minkowski space Consider Minkowski spacetime with metric ηij = diag(−1. derive the general solution.Y ] Q. Not to be quoted or reproduced without permission. and hence obtain the Poincar´ e invariance group of special relativity.8 The linearized Schwarzschild solution revisited Consider the external gravitational field of a static spherically symmetric body of mass ǫM situated at the coordinate origin x = y = z = 0. 1. 1. [Hint: use the identity Ra [bcd] = 0. Consider the unit sphere with metric ds2 = dθ2 + sin2 θ dφ2 . its 4-velocity is U i = (1. Assuming that the field is both weak and a function of R = (x2 +y 2 +z 2 )1/2 only. and define what is meant by independent here. Deduce that the result holds if Q is a tensor field. Show that ∂ . and that M/R is small. What is the third? Are there any more? 3.

Look for a static solution using the traditional approach to linearized theory.General Relativity Problems 3 3 3. and change the angular coordinate to obtain ds2 = dt2 − dz 2 − dr′ − r′ dφ′ . −Ωy. Compute the corresponding energy-momentum tensor. Not to be quoted or reproduced without permission. The energy-momentum tensor is T ab = ρU a U b . and a Cartesian 4-velocity U a = (1. −1). y plane centred on the origin. finding h11 = h22 = −λ as the only non-zero components of the perturbed metric tensor. so that terms of O(R2 Ω2 ) can be neglected. moving in a circular Newtonian orbit of radius R in the x. where λ ≡ 8µ ln(r/r0 ). Is this Minkowski spacetime? Show intuitively how a distant object may give rise to double images. the quadrupole moment and the metric perturbation. . and r0 is a scale length. 0). Show that their positions may be taken to be xa = ±(R cos Ωt. where Ω2 = m/(4R3 ). 3. Ωx. Terms involving powers of µ are to be ignored. 2 2 2 2 2 Copyright © 2008 University of Cambridge. 0.9 A cosmic string In Cartesian coordinates the energy momentum tensor of a straight cosmic string aligned along the z -axis is Tij = µδ (x)δ (y )diag(1. 0).11 The Lense-Thirring effect Consider a large thin shell of mass M and radius R which rotates slowly about the z -axis with angular velocity Ω. each of mass m. where µ is a small positive constant. Make a change of radial coordinate given by (1 − λ)r2 = (1 − 8µ)r′ 2 to obtain ds2 = dt2 − dz 2 − dr′ − (1 − 8µ)r′ dφ2 . where r2 = x2 + y 2 + z 2 . 3. r = x2 + y 2 .10 Gravitational waves from an orbiting binary Consider two stars. Introduce a shell density ρ = M δ (r − R)/(4πR2 ). R sin Ωt. 0. Rewrite the line element as ds2 = dt2 − dz 2 − (1 − λ)(dr2 + r2 dφ2 ).

˙ z ˙ ). This is the Lense-Thirring effect (1918). x. −x.General Relativity Problems 3 4 Show that T0α is a solenoidal vector. 0) for r < R. . where Copyright © 2008 University of Cambridge. ˙ y. Now consider a freely falling particle moving slowly with velocity x ˙ a = (1. the inertial frames are dragged around by the rotating shell. First solve the linearized Einstein equations for the scalar part. Obtain the geodesic equations using the connection components given on the handout. You should find that Bα = ω (y. Thus we can regard this problem as a superposition of two linearized perturbations. Not to be quoted or reproduced without permission. and a vector one for which only T0α is nonzero. How does the result differ from Newtonian theory? Next turn to the vector part.e. i. a scalar one for which only T00 is nonzero. ω = 4M Ω/(3R). Thus show that the inertial frames are rotating with angular velocity ω with respect to the background Minkowski frame. Peruse the handout showing the curvature tensors for these two types of perturbations.