Anisotropy and dispersion in periodically layered media

Geophysics ◽  
1984 ◽  
Vol 49 (4) ◽  
pp. 364-373 ◽  
Author(s):  
K. Helbig

Elastic waves propagating in a periodically layered medium exhibit transverse isotropy, provided the wavelength is long compared to the spatial period of the layer sequence. The elastic constants of the long‐wave equivalent transversely isotropic medium can be calculated in several ways, all of which are based on the low‐frequency limit: one either determines from the outset the macroscopic (in average homogeneous) strain response of a representative block of the periodically layered medium to (in average homogeneous) stresses, or one determines the dispersion equation for such media and evaluates this equation for low frequencies. Both approaches yield the same replacement medium. The replacement of a layered medium by a homogeneous transversely isotropic medium is justified if all wavelengths are sufficiently long. High‐resolution techniques, the increasing use of shear waves, and attention to stratigraphic detail require a quantitative evaluation of what is sufficiently long, as well as a study of what happens for wavelengths between that limit and the limit of resolution. Such information can be obtained through a numerical evaluation of the general dispersion equation: one obtains the frequency as a function of the spatial wave vector k. The phase velocity is the v=ωk/(k⋅k) and the group velocity [Formula: see text]. In general, both dispersion and anisotropy are to be expected. This method has been applied to SH-waves in periodically layered media in general, and the dispersion equation has been evaluated numerically for two representative media. The most significant result is that the long‐wavelengths approach, i.e., a nondispersive transversely isotropic replacement medium, is strictly valid for wavelengths larger than three times the spatial period of layering. For small angles against the vertical, dispersion for shorter wavelength is significant. However, for directions making an angle of more than about 30 degrees with the vertical, dispersion sets in at much shorter wavelengths and is in general much more gentle.

Geophysics ◽  
2000 ◽  
Vol 65 (4) ◽  
pp. 1305-1315 ◽  
Author(s):  
Hongbo Zhou ◽  
George A. McMechan

An analytical formula for geometrical spreading is derived for a horizontally layered transversely isotropic medium with a vertical symmetry axis (VTI). With this expression, geometrical spreading can be determined using only the anisotropy parameters in the first layer, the traveltime derivatives, and the source‐receiver offset. Explicit, numerically feasible expressions for geometrical spreading are obtained for special cases of transverse isotropy (weak anisotropy and elliptic anisotropy). Geometrical spreading can be calculated for transversly isotropic (TI) media by using picked traveltimes of primary nonhyperbolic P-wave reflections without having to know the actual parameters in the deeper subsurface; no ray tracing is needed. Synthetic examples verify the algorithm and show that it is numerically feasible for calculation of geometrical spreading. For media with a few (4–5) layers, relative errors in the computed geometrical spreading remain less than 0.5% for offset/depth ratios less than 1.0. Errors that change with offset are attributed to inaccuracy in the expression used for nonhyberbolic moveout. Geometrical spreading is most sensitive to errors in NMO velocity, followed by errors in zero‐offset reflection time, followed by errors in anisotropy of the surface layer. New relations between group and phase velocities and between group and phase angles are shown in appendices.


Geophysics ◽  
1994 ◽  
Vol 59 (10) ◽  
pp. 1605-1612 ◽  
Author(s):  
Björn E. Rommel

Many real rocks and sediments relevant to seismic exploration can be described by elastic, transversely isotropic media. The properties of plane waves propagating in a transversely isotropic medium can be given in an analytically exact form. Here the polarization is recast into a comprehensive form that includes Daley and Hron’s normalization and Helbig’s full range of elastic constants. But these formulas are rather lengthy and do not easily reveal the features caused by anisotropy. Hence Thomsen suggested an approximation scheme for weak transverse isotropy. His derivation of the approximate polarization, however, is based on a property that is not suitable to measure small differences between an isotropic and a weakly transversely isotropic medium. Therefore the approximation of the polarization is recast. The corrected approximation does show a dependence on weak transverse isotropy. This feature can be viewed as an additional rotation of the polarization with respect to the wavenormal. It depends on the anisotropy as well as the inverse velocity ratio. An approximate condition of pure polarization, which occurs in certain directions, is also obtained. The corrected approximation results in a better match of the approximate polarization with the exact one, providing the assumption of weak transverse isotropy is met. When comparing the additional rotation with the deviation of the (observable) ray direction from the wavenormal, the qSV‐wave indicates transverse isotropy most clearly.


1984 ◽  
Vol 51 (4) ◽  
pp. 811-815 ◽  
Author(s):  
Y. M. Tsai

The stress distribution produced by the identation of a penny-shaped crack by an oblate smooth spheroidal rigid inclusion in a transversely isotropic medium is investigated using the method of Hankel transforms. This three-part mixed boundary value problem is solved using the techniques of triple integral equations. The normal contact stress between the crack surface and the indenter is written as the product of the associated half-space contact stress and a nondimensional crack-effect correction function. An exact expression for the stress-intensity is obtained as the product of a dimensional quantity and a nondimensional function. The curves for these nondimensional functions are presented and used to determine the values of the normalized stress-intensity factor and the normalized maximum contact stress. The stress-intensity factor is shown to be dependent on the material constants and increasing with increasing indentation. The stress-intensity factor also increases if the radius of curvature of the indenter surface increases.


Sign in / Sign up

Export Citation Format

Share Document