Measuring and testing – Vibration – By mechanical waves
Reexamination Certificate
2002-04-22
2004-06-29
Williams, Hezron (Department: 2856)
Measuring and testing
Vibration
By mechanical waves
C073S628000
Reexamination Certificate
active
06755083
ABSTRACT:
BACKGROUND OF THE INVENTION
Component parts, whether they are lenses for optical systems or components of sensitive weapon systems, must be inspected regularly as part of a scheduled maintenance and monitoring program, especially during the assembly process. These inspections are critical, because even minute defects can propagate into still larger flaws that lead to a malfunction or possible catastrophic destruction of the system. Most materials, especially in optical components, have many small defects that are distributed throughout; therefore, it is important to select the largest defects that create the strongest returns. A nondestructive evaluation (NDE) technique would be of considerable value for inspecting optics in advanced laser systems and component parts for stockpile stewardship, for noninvasive medical treatments, and for locating underground targets.
A project is underway at the Lawrence Livermore National Laboratory for developing an NDE technique for dynamically focusing acoustical energy for both detecting and characterizing flaws in parts undergoing ultrasonic testing. This project applies a systematic approach that incorporates detailed simulations, algorithm development, hardware, proof-of-principle NDE experiments, and the design of a flaw-detection/localization/imaging system.
The effort enables one or a few strongly scattering targets or defects to be selected from a larger group while the larger group is itself imbedded in a cluttered environment. Current methods for attacking this problem use a time-domain method (which is quite fast) to locate the strongest scatterer or defect reliably. But, in order to distinguish other scatterers or defects, the existing methods require transformation of the data into the frequency domain, and then a difficult procedure (singular value decomposition) is followed to find the eigenvectors and eigenvalues of a complex matrix in the frequency domain. There is clearly a need to design a method that works wholly in the time domain, and is purely empirical, as is the step that determines the location of the first (and strongest) scatterer, defect or target.
The present invention provides a solution to the above referenced problem and involves a method of time-reversal acoustics which uses an iterative process to determine the optimum signal for locating a strongly reflecting target or scatterer in a cluttered environment, similar to the prior techniques, to determine locations of other targets. The present invention provides a means of localizing other targets in the time domain by using properties of constructive and destructive interference of the sound waves. In this method, after locating the stronger (larger) scattering target, instead of sending back the same signals, which are the time-reversed signals, half of the signals will additionally be reversed in sign and, a new send/receive, send-time-reversal/receive iteration process can then proceed with confidence that the stronger (larger) scatterer is no longer contributing to the received signal.
SUMMARY OF THE INVENTION
It is an object of the present invention to distinguish multiple targets using time-reversal acoustics.
A further object of the invention is to provide a method for locating a number of targets imbedded in a cluttered environment in the time domain rather than in the frequency domain.
A further object of the invention is to provide a method for locating different targets by time-reversal acoustics using an interative process which operates in the time domain.
Another object of the invention is to provide a method for sequentially locating targets of smaller size or scattering strength using the time domain, wherein after locating a larger target, instead of sending back the same signals, which are time reversed signals, the method sends back these signals but with half of them reversed in sign.
Another object of the invention is to enable localizing smaller or more distant targets in the time domain by using properties of constructive and destructive interference of the sound waves by reversing in sign half of the time-reversed signals.
Other objects and advantages of the present invention will become apparent from the following description. The invention involves distinguishing multiple targets using time-reversal acoustics. The invention enables localizing both the larger (stronger) and smaller (weaker) targets in the time domain, thereby eliminating the prior problems of locating the smaller (weaker) targets using the frequency domain of the prior known techniques. In the present invention, after the locating a strongly reflecting target in a cluttered environment using the conventional or standard time-reversal acoustics, the invention provides for localizing other targets in the time domain by using properties of constructive and destructive interference of the secondary waves. After the largest eigenvalue/eigenvector combination has been determined, instead of sending back the same time-reversed signals, half of the signals are reversed in sign. There are various ways of choosing which half of the signals to reverse. One choice is to reverse every other signal in a linear array, or as in a checkerboard pattern for a 2D array of transducers. These reversing signal choices are made to enhance the likelihood that the signal received at the strongest scatterer is weak due to destructive interference of the arrivals. Then, a new send/receive, send-time-reversed/receive iteration can proceed with confidence that the largest (strongest) scatterer is no longer contributing to the received signal. Many choices of the pattern of sign reversal of the signals may be constructed. Discrete analogs of sines and cosines along the array provide one class of choices. After the second strongest (smallest) scatterer or target has been located using this method, then the procedure can be repeated always in time domain using known sequential orthogonalization techniques. Often, the first iteration in this sequence will be close to the desired signal from a second target. In some cases, orthogonalization procedures must be implemented to assure the returned signals are in fact orthogonal to the first eigenvector found.
REFERENCES:
patent: 5092336 (1992-03-01), Fink
patent: 5428999 (1995-07-01), Fink
patent: 5431053 (1995-07-01), Fink
patent: 6161434 (2000-12-01), Fink et al.
patent: 6202489 (2001-03-01), Beffy et al.
patent: 6490469 (2002-12-01), Candy
Carnahan L. E.
Miller Rose M.
Scott Eddie E.
The Regents of the University of California
Thompson Alan H.
LandOfFree
Method for distinguishing multiple targets using... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for distinguishing multiple targets using..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for distinguishing multiple targets using... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3365173