Nondestructive methods and systems for detecting and/or characterizing damage
A nondestructive method for detecting damage in parts and/or characterizing effective material properties may include: exposing a material to one or more nondestructive stimuli; measuring a response of the material to the stimuli; selecting at least one of a specific length scale or a specific time scale; and analyzing the measurement of the response with a scale-aware single- or multi-physics model to identify anomalies in the measurements as compared to an expected response of the material to the stimuli, wherein the scale-aware single- or multi-physics model is based on the at least one of the specific length scale or the specific time scale.
1. A method comprising:
exposing, via a surface acoustic wave (SAW) wedge transducer or an ultrasonic transducer, a material sample having different material granularities to one or more nondestructive stimuli;
measuring, via a laser device, a response of the material sample to the one or more nondestructive stimuli; and
analyzing the measurement of the response with a scale-aware single- or multi-physics model to identify anomalies in the measurement as compared to an expected response of the material sample to the one or more nondestructive stimuli, wherein:
the scale-aware single- or multi-physics model is based on at least one of a specific length scale or a specific time scale; and
the scale-aware single- or multi-physics model comprises governing equations comprising scale-agnostic first principles and scale-aware phenomenological laws that must hold over any finite or infinitesimal regions in a space or a time, and the scale-aware phenomenological laws are single- or multi-physics constitutive or interaction laws that must be learned from data at the at least one of the specific length scale or the specific time scale.
2. The method of claim 1 , wherein the constitutive laws are discretized or integrated, and the scale-aware phenomenological laws are parameterized to yield parameters, and wherein optimization is used to fit the parameters to minimize a residual error for the governing equations, a solution error, or an error in observables.
3. The method of claim 1 , wherein the constitutive laws are discretized or integrated, and the scale-aware phenomenological laws are parameterized, and wherein machine learning is used to fit the parameters to minimize a residual error for the governing equations, a solution error, or an error in observables.
4. The method of claim 1 , wherein the at least one of the specific length scale or the specific time scale are selected manually by a user.
5. The method of claim 1 , wherein the at least one of the specific length scale or the specific time scale are selected by a computer based on a goodness of fit of the scale-aware single- or multi-physics model to the measurement of the response for different length scales and/or different time scales.
6. The method of claim 1 , wherein the one or more nondestructive stimuli comprise one or more of ultrasound waves or acoustic waves.
7. A system comprising:
a memory coupled to one or more processors, the memory having instructions stored thereon, the instructions executable to cause the one or more processors to perform:
exposing, via a surface acoustic wave (SAW) wedge transducer or an ultrasonic transducer, a material sample having different material granularities to one or more nondestructive stimuli;
measuring, via a laser device, a response of the material sample to the one or more nondestructive stimuli; and
analyzing the measurement of the response with a scale-aware single- or multi-physics model to identify anomalies in the measurements as compared to an expected response of the material sample to the one or more nondestructive stimuli, wherein:
the scale-aware single- or multi-physics model is based on at least one of a specific length scale or a specific time scale; and
the scale-aware single- or multi-physics model comprises governing equations comprising scale-agnostic first principles and scale-aware phenomenological laws that must hold over any finite or infinitesimal regions in a space or a time, and the scale-aware phenomenological laws are single- or multi-physics constitutive or interaction laws that must be learned from data at the at least one of the specific length scale or the specific time scale.
8. The system of claim 7 further comprising:
a nondestructive stimuli source; and
a detector capable of the measuring of the response of the material sample to the one or more nondestructive stimuli.
9. A method comprising:
exposing, via a surface acoustic wave (SAW) wedge transducer or an ultrasonic transducer, a material sample having different material granularities to one or more nondestructive stimuli;
measuring, via a laser device, a response of the material sample to the one or more nondestructive stimuli; and
analyzing the measurement of the response with a scale-aware single- or multi-physics model to characterize one or more effective material properties based on the measurement of the response of the material sample to the one or more nondestructive stimuli, wherein:
the scale-aware single- or multi-physics model is based on the at least one of a specific length scale or a specific time scale; and
the scale-aware single- or multi-physics model comprises governing equations comprising scale-agnostic first principles and scale-aware phenomenological laws that must hold over any finite or infinitesimal regions in a space or a time, and the scale-aware phenomenological laws are single- or multi-physics constitutive or interaction laws that must be learned from data at the at least one of the specific length scale or the specific time scale.
10. The method of claim 9 , wherein the constitutive laws are discretized or integrated, and wherein the scale-aware phenomenological laws are parameterized to yield parameters, and wherein optimization is used to fit the parameters to minimize a residual error for the governing equations, a solution error, or an error in observables.
11. The method of claim 9 , wherein the constitutive laws are discretized or integrated, and the scale-aware phenomenological laws are parameterized, and wherein machine learning is used to fit the parameters to minimize a residual error for the governing equations, a solution error, or an error in observables.
12. The method of claim 9 , wherein the at least one of the specific length scale or the specific time scale are selected manually by a user.
13. The method of claim 9 , wherein the at least one of the specific length scale or the specific time scale are selected by a computer based on a goodness of fit of the scale-aware single- or multi-physics model to the measurement of the response for different length scales and/or different time scales.
14. The method of claim 9 , wherein the one or more nondestructive stimuli comprise one or more of ultrasound waves or acoustic waves.
15. A system comprising:
a memory coupled to one or more processors, the memory having instructions stored thereon, the instructions executable to cause the one or more processors to perform:
exposing, via a surface acoustic wave (SAW) wedge transducer or an ultrasonic transducer, a material sample having different material granularities to one or more nondestructive stimuli;
measuring, via a laser device, a response of the material sample to the one or more nondestructive stimuli; and
analyzing the measurement of the response with a scale-aware single- or multi-physics model to characterize one or more effective material properties based on the measurement of the response of the material sample to the one or more nondestructive stimuli, wherein:
the scale-aware single- or multi-physics model is based on at least one of a specific length scale or a specific time scale; and
the scale-aware single- or multi-physics model comprises governing equations comprising scale-agnostic first principles and scale-aware phenomenological laws that must hold over any finite or infinitesimal regions in a space or a time, and the scale-aware phenomenological laws are single- or multi-physics constitutive or interaction laws that must be learned from data at the at least one of the specific length scale or the specific time scale.
16. The system of claim 15 further comprising:
a nondestructive stimuli source; and
a detector capable of the measuring of the response of the material sample to the one or more nondestructive stimuli.