Ultrasonic weld quality testing using audio
Various embodiments of a technique are provided for determining the quality of a weld for a battery contact using audio signals. In some embodiments, a system for determining the quality of a weld for a battery contact comprises a processor configured to receive a snapshot that includes a rendering of an audio signal obtained during a welding process by monitoring a device being welded. A weld-quality indicator for the device being welded is determined based at least in part on features extracted from the snapshot and a weld-quality score for the device being welded is determined in response to the weld-quality indicator. In this case, the weld-quality score is indicative of a quality of the weld. The system also includes a memory coupled to the processor and configured to provide the processor with instructions.
1. A system for determining the quality of a weld, the system comprising:
a processor configured to:
receive an input generated in response to sound produced by a welding process that produced a weld, wherein the welding process is performed by an ultrasonic welder having an excitation frequency, wherein the input is indicative of intensity amplitudes of the sound over a range of frequencies including odd harmonics of half the excitation frequency;
process the input to identify features of the sound produced at a time T1, wherein the identified features are based on sound intensity amplitudes and a corresponding set of frequencies of the sound produced at the time T1;
process the identified features of the sound to determine a weld-quality indicator for the weld; and
determine a weld-quality score for the weld based on the weld-quality indicator, wherein the weld-quality score is indicative of a quality of the weld; and
a memory coupled to the processor and configured to provide the processor with instructions.
2. The system recited in claim 1 , wherein:
the weld is applied to a battery contact of a device;
identified features of the sound include a plurality of peak intensity amplitudes of the sound; and
the corresponding set of frequencies includes odd harmonics of half of the excitation frequency.
3. The system recited in claim 1 , wherein:
the weld is applied to a battery contact of a device;
the identified features of the sound include a plurality of peak intensity amplitude differences, each of the peak intensity amplitude differences corresponding to a difference in intensity amplitude between an excitation peak intensity amplitude and each of a plurality of peak intensity amplitudes of the sound in the corresponding set of frequencies at the time T1; and
the corresponding set of frequencies includes odd harmonics of half of the excitation frequency.
4. The system recited in claim 1 , wherein:
the weld is applied to a battery contact of a device;
the input is generated in response to sound produced by the welding process at a plurality of times including the time T1;
the input is indicative of intensity amplitudes of the sound over a range of frequencies including odd harmonics of half the excitation frequency at each of the plurality of times; and
the processor is configured to:
process the input to identify features of the sound produced at each of the plurality of times, wherein the identified features at each of the plurality of times are based on sound intensity amplitudes and a corresponding set of frequencies of the sound produced at the respective time; and
process the identified features of the sound produced at each of the plurality of times to determine the weld-quality indicator for the weld.
5. The system recited in claim 1 , wherein:
the weld is applied to a battery contact of a device;
the input is generated in response to sound produced by the welding process at a plurality of times including the time T1;
the input is indicative of intensity amplitudes of the sound over a range of frequencies including odd harmonics of half the excitation frequency at each of the plurality of times; and
the processor is configured to:
process the input to determine peak intensity amplitude dissipation rates of the sound produced by the welding process, wherein each of the peak intensity amplitude dissipation rates is indicative of a rate of decrease in peak intensity amplitude of a respective peak intensity amplitude at a respective frequency over the plurality of times; and
process the peak intensity amplitude dissipation rates to determine the weld-quality indicator for the weld.
6. The system recited in claim 1 , wherein:
the weld is applied to a battery contact of a device;
the input is generated in response to sound produced by the welding process at a plurality of times including the time T1;
the input is indicative of intensity amplitudes of the sound over a range of frequencies including odd harmonics of half the excitation frequency at each of the plurality of times; and
the processor is configured to:
process the input to determine dissipation rates of the sound produced by the welding process, wherein each of the dissipation rates corresponds to a difference in intensity amplitude between an excitation peak intensity amplitude and each of a plurality of peak intensity amplitudes of the sound appearing at a set of frequencies at the plurality of times.
7. The system recited in claim 6 , wherein the set of frequencies includes odd harmonics of half the excitation frequency.
8. The system recited in claim 1 , wherein:
the weld is applied to a battery contact of a device;
the processor is configured to process the input to compare the input with a baseline input generated in response to sound produced by the welding process during the generation of a good weld; and
the processor is configured to compare the input with the baseline input using an image similarity algorithm.
9. The system recited in claim 1 , wherein:
the weld is applied to a battery contact of a device;
the weld-quality indicator comprises a plurality of weld-quality indicators, wherein at least one of the plurality of weld-quality indicators is determined from the input in response to at least one of the group consisting of:
a plurality of peak intensity amplitudes corresponding to a plurality of peak intensity amplitudes appearing at a set of frequencies at the time T1, wherein the set of frequencies includes odd harmonics of half of the excitation frequency;
a plurality of peak amplitude differences corresponding to a difference in amplitude between an excitation peak intensity amplitude and each of a plurality of peak intensity amplitudes appearing at a set of frequencies at the time T1, wherein the set of frequencies includes odd harmonics of half of the excitation frequency; and
a comparison of the input with a baseline input generated in response to sound produced by the welding process during the generation of a good weld, wherein the comparison is performed by using an image similarity algorithm.
10. The system recited in claim 1 , wherein:
the weld is applied to a battery contact of a device;
the input is generated in response to sound produced by the welding process at a plurality of times including the time T1;
the input is indicative of intensity amplitudes of the sound over a range of frequencies including odd harmonics of half the excitation frequency at each of the plurality of times; and
the weld-quality indicator comprises a plurality of weld-quality indicators, wherein at least one of the plurality of weld-quality indicators is determined in response to at least a plurality of peak intensity amplitudes determined from at least one of the plurality of times, the plurality of peak intensity amplitudes corresponding to a plurality of peak intensity amplitudes appearing at a set of frequencies at a time T1, wherein the set of frequencies includes odd harmonics of half of the excitation frequency.
11. A method for determining the quality of a weld, the method comprising:
receiving an input generated in response to sound produced by a welding process that produced a weld, wherein the welding process is performed by an ultrasonic welder having an excitation frequency, wherein the input is indicative of intensity amplitudes of the sound over a range of frequencies including odd harmonics of half the excitation frequency;
processing the input to identify features of the sound produced at a time T1, wherein the identified features are based on sound intensity amplitudes and a corresponding set of frequencies of the sound produced at the time T1;
processing the identified features of the sound to determine a weld-quality indicator for the weld; and
determining a weld-quality score for the weld based on the weld-quality indicator, wherein the weld-quality score is indicative of a quality of the weld.
12. The method of claim 11 wherein:
the weld is applied to a battery contact of a device; and
the identified features of the sound include a plurality of peak intensity amplitudes of the sound, wherein the corresponding set of frequencies includes odd harmonics of half of the excitation frequency.
13. The method of claim 11 wherein:
the weld is applied to a battery contact of a device;
the identified features of the sound include a plurality of peak intensity amplitude differences, each of the peak intensity amplitude differences corresponding to a difference in intensity amplitude between an excitation peak intensity amplitude and each of a plurality of peak intensity amplitudes of the sound for the corresponding set of frequencies at the time T1; and
the corresponding set of frequencies includes odd harmonics of half of the excitation frequency.
14. The method of claim 11 wherein:
the weld is applied to a battery contact of a device;
the input is generated in response to sound produced by the welding process at a plurality of times including the time T1;
the input is indicative of intensity amplitudes of the sound over a range of frequencies including odd harmonics of half the excitation frequency at each of the plurality of times; and
the input is processed to identify features of the sound produced at each of the plurality of times, wherein the identified features are based on sound intensity amplitudes and a corresponding set of frequencies of the sound produced at the time T1; and
the identified features of the sound produced at each of the plurality of times are processed to determine the weld-quality indicator for the weld.
15. The method of claim 11 wherein:
the weld is applied to a battery contact of a device;
the input is generated in response to sound produced by the welding process at a plurality of times including the time T1;
the input is indicative of intensity amplitudes of the sound over a range of frequencies including odd harmonics of half the excitation frequency at each of the plurality of times; and
the input is processed to determine peak intensity amplitude dissipation rates of the sound produced by the welding process, wherein each of the peak intensity amplitude dissipation rates is indicative of a rate of decrease in peak intensity amplitude of a respective peak intensity amplitude at a respective frequency over the plurality of times; and
the peak intensity amplitude dissipation rates are processed to determine the weld-quality indicator for the weld.
16. The method of claim 11 wherein:
the weld is applied to a battery contact of a device;
the input is generated in response to sound produced by the welding process at a plurality of times including the time T1;
the input is indicative of intensity amplitudes of the sound over a range of frequencies including odd harmonics of half the excitation frequency at each of the plurality of times;
the input is processed to determine dissipation rates of the sound produced by the welding process; and
each of the dissipation rates corresponds to a difference in intensity amplitude between an excitation peak intensity amplitude and each of a plurality of peak intensity amplitudes of the sound appearing at a set of frequencies at the plurality of times.
17. The method of claim 16 , wherein the set of frequencies includes odd harmonics of half the excitation frequency.
18. The method of claim 11 wherein:
the weld is applied to a battery contact of a device;
the input is compared with a baseline input generated in response to sound produced by the welding process during the generation of a good weld; and
the input is compared with the baseline input using an image similarity algorithm.
19. The method of claim 11 wherein:
the weld is applied to a battery contact of a device;
the input is generated in response to sound produced by the welding process at a plurality of times including the time T1;
the input is indicative of intensity amplitudes of the sound over a range of frequencies including odd harmonics of half the excitation frequency at each of the plurality of times; and
the weld-quality indicator comprises a plurality of weld-quality indicators, wherein determining at least one of the weld-quality indicators in the plurality of weld-quality indicators includes at least one of:
determining a plurality of peak intensity amplitudes corresponding to a plurality of peak intensity amplitudes appearing at a set of frequencies at a time from at least one of the plurality of times, wherein the set of frequencies includes odd harmonics of half of the excitation frequency;
determining a plurality of peak intensity amplitude differences corresponding to a difference in amplitude between an excitation peak intensity amplitude and each of a plurality of peak intensity amplitudes appearing at a set of frequencies at a time from at least one of the plurality of times, wherein the set of frequencies includes odd harmonics of half of the excitation frequency; and
comparing the input with a baseline input generated in response to sound produced by the welding process during the generation of a good weld by using an image similarity algorithm.
20. The method of claim 11 wherein:
the weld is applied to a battery contact of a device;
the input is generated in response to sound produced by the welding process at a plurality of times including the time T1;
the input is indicative of intensity amplitudes of the sound over a range of frequencies including odd harmonics of half the excitation frequency at each of the plurality of times; and
the weld-quality indicator comprises a plurality of weld-quality indicators, wherein determining at least one of the weld-quality indicators in the plurality of weld-quality indicators includes at least a plurality of peak intensity amplitude differences corresponding to a difference in amplitude between an excitation peak intensity amplitude and each of a plurality of peak intensity amplitudes appearing at a set of frequencies at a time from at least one of the plurality of times, wherein the set of frequencies includes odd harmonics of half of the excitation frequency.
21. A computer program product for determining the quality of a weld, the computer program product being embodied in a tangible computer readable storage medium and comprising computer instructions for:
receiving an input generated in response to sound produced by a welding process that produced a weld, wherein the welding process is performed by an ultrasonic welder having an excitation frequency, wherein the input is indicative of intensity amplitudes of the sound over a range of frequencies including odd harmonics of half the excitation frequency;
processing the input to identify features of the sound produced at a time T1, wherein the identified features are based on sound intensity amplitudes and a corresponding set of frequencies of the sound produced at the time T1;
processing the identified features of the sound to determine a weld-quality indicator for the weld; and
determining a weld-quality score for the weld based on the weld-quality indicator, wherein the weld-quality score is indicative of a quality of the weld.