Numerical Analysis Of Porous Cylindrical Pipe Simulating Source In Acoustic Logging And Study On Application Of Inversion With Data From Cross-dipole Acoustic Logging

Numerical Analysis of Porous Cylindrical Pipe Simulating Source in Acoustic Logging and Study on Application of Inversion with Data from Cross-dipole Acoustic logging Author: Weiguo LvAdviser: Kexie Wang Subject: Theoretical PhysicsIt is important to constitute a similar model in order to simulate acoustic logging system. Whichever models approximate of acoustic logging system. There are many holes on the acoustic logging tool that can eliminate direct waveform. It is difficult to establish math model to simulate the holes. In much previous theoretical research of acoustic logging, the used source models were point source or rigid cylindrical model, which were too simplifying the acoustic logging system. This paper studies mainly a porous source model to simulate acoustic logging when porocity is great. Transmitter is cylindrical source. In the first section of this thesis we completed the theoretical formulation of the acoustic field excited porous source in fluid-filled borehole that surrounded by uniform formation. We accomplish the simulation of field in three different formations and two different porous sources. Results of numerical analysis show: There are extensional mode and inner-stonely wave in porous shell when characteristic velocity of compressional and shear waves are small, and it is evidence in slow formation. The velocity of extensional mode is 0.9 times of the characteristic compressional velocity in the porous shell. The velocity of inner-stonely is bigger than velocity of stonely in formation. Because of compressional wave be fast in the fast formation, extensional mode doesn't be observed in compressional waveform. Similarly, inner-stonely wave is in the shear waveform. When porocity increases, the extensional wave is disappear but inner-stonely is appear, and it simulates holes in acoustic logging tool. To identify components in the system, we analyze all dispersion of modes wave. The modes are composed of two parts. One parts are relative to formation closely that are stonely wave and pseudo-Rayleigh wave, the other is relative to porous source which are extensional mode ,inner-stonely and a unknown mode wave. Extensional mode isn't cut-off mode, and its velocity is 0.9 times of the characteristic compressional in the porous shell. As the frequency is lowered, velocity of extensional mode is decreases. It is excited greatly in low frequency and not in high frequency. When porocity decreases, both characteristic compressional and extensional mode velocity increase. But the velocity of extensional mode is still 0.9 times of the characteristic compressional velocity in the porous shell. Characteristic of dispersion have changed. Velocity of the unknown mode is dispersion and it isn't a cut-off mode. As frequency is raised, the velocity increasing in high frequency equals to frame-shear velocity in porous shell. Its amplitude excited is small and can't be observed in the waveform. Because velocity of characteristic shear equals to nearly that of frame-shear as porocity is reduced, dispersion curve of the unknown mode become a line. The inner-stonely mode hasn't cut-off frequency. As frequency increases, its velocity increases. As frequency is raised further, the velocity reduces to a constant that is less than velocity of liquid in borehole. The excited spectrums of compressional wave and shear wave not only relate to medium parameters, but also to source model (or model parameter). Weighted spectral semblance (WSS) is a simple yet efficient technique to measure dispersion curve of guided modes from array data. But The method only can extract a mode from array data. We can't pick up more than one mode at each frequency and even dispersion curve extracted is confused and other modes are loss. Prony's method can extract multiple mode, but a disadvantage of this method is that it often generates spurious estimate, another shortcoming is that it requires regularly-spaced receiver. It isn't fit to processing array data that gets rid of distortion data. In this paper Weighted spe...