Functional Integration pp 51-66 | Cite as

# Physics on and near Caustics

## Abstract

*on*caustics is obtained by studying physics

*near*caustics. Physics on caustics is at the cross-roads of calculus of variation and functional integration. More precisely consider a space

*P*

**M**

^{ d }of paths

*x*: [

*t*

_{ a },

*t*

_{ b }] →

**M**

^{ d }, on a

*d*-dimensional manifold

**M**

^{ d }Consider two of its subspaces:

- i)
The 2

*d*-dimensional space*U*of critical points of an action functional*S*:*P***M**^{ d }→ ℝ \(q \in U \Leftrightarrow \frac{{\delta S}}{{\delta q}}q = 0\)]. - ii)
The space

*P*_{¼},_{ v }**M**^{ d }of paths satisfying*d*initial conditions (μ) and*d*final conditions (*v*).The nature of the intersection*P*_{μ},_{ v }**M**^{ d }∩*U*is the key to identifying and analyzing caustics. Caustics occur when the roots of*S′*(*q*), = 0, for*S*restricted to*P*_{μ},_{ v }**M**^{ d }are not isolated points in*U*. We consider two cases:- i)
The roots define a subspace of non zero dimension ℓ of

*U*. For example, the classical system with action*S*is constrained by conservation laws. - ii)
There is a multiple root of

*S′*(*q*) = 0. For example the classical flow has an envelope.In both situations there is, at least, one nonzero Jacobi field along

*q*with*d*initial vanishing boundary conditions (μ) and*d*final vanishing boundary conditions (*v*), i.e.,*q*(*t*_{ a }) and*q*(*t*_{ b }) are conjugate points along*q*. In both cases we say “*q*(_{ b }) is on the caustics.”The strict WKB approximations of functional integrals for the system

*S*“break down” on caustics, but the full semiclassical expansion, including contributions from*S′*(*q*) and*S‴*(*q*) (and possibly higher derivatives) yield the physical properties of the system*S*on and near the caustics. Caustics display classical physics as a limit of quantum mechanics.We work out glory scattering cross-sections because, there, both situations occur simultaneously: the system is constrained by conservation laws, and the classical flow is caustic forming. The cross-section is given in closed form in section 5.

- i)

## Keywords

Functional Integration Conjugate Point Probability Amplitude Airy Function Jacobi Operator## Preview

Unable to display preview. Download preview PDF.

## References

- [1]C. DeWitt-Morette, G. Low, L. S. Schulman, A. Y. Shiekh: “Wedges I,”
*Foundations of Physics***16**, 311–349 (1986).MathSciNetADSCrossRefGoogle Scholar - [2]The properties of Jacobi fields in configuration space and phase space, and the corresponding WKB approximations are scattered in several papers of Cécile DeWitt-Morette, John La Chapelle, Maurice Mizrahi, Bruce Nelson, Benny Sheeks, Alice Young, and Tian-Rong Zhang. For a summary of results obtained prior to 1984, see references [3, 4]. For more recent results, see reference [5].Google Scholar
- [3]Cécile DeWitt-Morette and Tian-Rong Zhang: “Path integrals and conservation laws,”
*Phys. Rev D***28**, 2503–2516 (1983).MathSciNetADSCrossRefGoogle Scholar - [4]Cécile DeWitt-Morette: “Feynman path integrals, from the prodistribution definition to the calculation of glory scattering,”
*Acta Physica Austriaca Suppl.***XXVI**, 101–170 (1984).MathSciNetGoogle Scholar - [5]John La Chapelle: “Functional Integration on Symplectic Manifolds,” (Ph. D. Dissertation, The University of Texas at Austin, May 1995).Google Scholar
- [6]Cécile DeWitt-Morette and Tian-Rong Zhang: “Feynman-Kac formula in phase space with applications to coherent state transitions,”
*Phys. Rev. D***28**, 2517–2525 (1983).MathSciNetADSCrossRefGoogle Scholar - [7]Cécile DeWitt-Morette, Bruce Nelson, and Tian-Rong Zhang: “Caustic problems in quantum mechanics with applications to scattering theory,”
*Phys. Rev. D***28**, 2526–2546 (1983).MathSciNetADSCrossRefGoogle Scholar - [8]Pierre Cartier and Cécile DeWitt-Morette: “A new perspective on functional integration,”
*J. Math. Phys.***36**, 2237–2312 (1995). For strict WKB approximations, see the calculation leading to (III.55).MathSciNetADSzbMATHCrossRefGoogle Scholar - [9]We use
*h*rather than ℏ =*h*/2π, in order that expressions such as Gaussian, Fourier transforms, etc... on ℝ^{d}do not depend on*d*.Google Scholar - [10]C. DeWitt-Morette, A. Maheshwari, and B. Nelson: “Path integration in non-relativistic quantum mechanics,”
*Phys. Rep.***50**, 266–372 (1979). This article includes a summary of earlier articles.MathSciNetCrossRefGoogle Scholar - [11]Cécile DeWitt-Morette and Bruce L. Nelson: “Glories—and other degenerate points of the action,”
*Phys. Rev. D***29**, 1663–1668 (1984).MathSciNetADSCrossRefGoogle Scholar - [12]Cécile DeWitt-Morette: “Catastrophes in Lagrangian systems” and C. DeWitt-Morette and P. Tshumi: “Catastrophes in Lagrangian systems. An example,” in
*Long Time Prediction in Dynamics*Eds. V. Szebehely and B. D. Tapley (D. Reidel Pub. Co. 1976) pp. 57-69. Also Y. Choquet-Bruhat and C. DeWitt-Morette*Analysis, Manifolds and Physics*Vol. I pp. 105-109, and pp. 277-281 (North Holland 1982–1996).Google Scholar - [13]R. A. Matzner, C. DeWitt-Morette, B. Nelson, and T.-R. Zhang: “Glory scattering by black holes”
*Phys. Rev. D***31**, 1869–1878 (1985).ADSCrossRefGoogle Scholar - [14]Tian-Rong Zhang and Cécile DeWitt-Morette: “WKB cross-section for polarized glories of massless waves in curved space-times,”
*Phys. Rev. Letters***52**, 2313–2316 (1984).MathSciNetADSCrossRefGoogle Scholar

## References Added in Proof

- Ph. Choquard and F. Steiner: “The story of Van Vleck’s and Morette-Van Hove’s determinants”
*Helv. Phys. Acta***69**, 636–654 (1969) analyzes carefully the contributions of Cécile Morette (“On the definition and approximation of Feynman’s path integral,”*Phys. Rev.***81**, 848-852 (1951)), Leon Van Hove (“annexe” to his doctorate thesis) and Wolfang Pauli (unpublished notes) to the strict WKB approximation of Feynman’s path integral. The appearance of singular determinants (“WKB breaks down”) was first noted by Ph. Choquard in*Helv. Phys. Acta***28**, 89-157 (1955) on pages 114-119.MathSciNetGoogle Scholar