where expertise comes together - since 1996

Web's Largest Portal of Nondestructive Testing (NDT)
Open Access Database (Conference Proceedings, Articles, News), Exhibition, Forum, Network

All Forum Boards
Technical Discussions >
Phase Array for mis-orientated si
Career Discussions
Job Offers
Job Seeks
Classified Ads
About NDT.net
Articles & News

713 views
02:04 Jan-16-2007

Charles Truskolasky

Test Examiner Mil-2132, Level III
Lehigh Heavy Forge,
USA,
Joined Jan 2007
10
Phase Array for mis-orientated si

What are the depth and sweep angle limitations for finding 2mm flaws with sig/noise ratio limit of 6dB. I need to find a system capable of finding such flaws in fine grain mild steel, in cylindrical bodies with diameters to 48-in and depths extending to the center of these diameters. I want to use longitudinal w. sweeps 0 to 35-deg. and 45-deg shear w., in the circumferential directions. The sweep is necessary to locate flaws that lie at various angles to the surface. Is anyone doing this?


 
01:12 Jan-18-2007

Michael Moles †2014 *1948

,
Joined

Re: Phase Array for mis-orientated si ----------- Start Original Message -----------
: What are the depth and sweep angle limitations for finding 2mm flaws with sig/noise ratio limit of 6dB. I need to find a system capable of finding such flaws in fine grain mild steel, in cylindrical bodies with diameters to 48-in and depths extending to the center of these diameters. I want to use longitudinal w. sweeps 0 to 35-deg. and 45-deg shear w., in the circumferential directions. The sweep is necessary to locate flaws that lie at various angles to the surface. Is anyone doing this?
------------ End Original Message ------------

It is difficult to answer this question without more information. First, are these defects omni-directional reflectors like inclusions, or planar like cracks? If the latter, you need to tailor the incident angles to the defect orientation; if the former, this is less of a big problem. Assuming inclusions or something similar, the frequency will be set by the material properties. (You can check this manually, and usethe same frequency for your array definition.) The arry element spacing (gap) will define the amount of sweep that you can manage; let's assume you have many small elements (e.g. 32) which will allow 0-35 L-waves. These are calculatable from standard physics equations, or measurable following ASTM E-2491-06 for PA set-ups. Lastly, how can you get 6 dB SNR on small defects? Probably by focusing with your PA, which can give ~10x the amount of power. Please post typical ultrasonic frequencies, defect types, and planned arrays for further comments.


 
02:04 Jan-18-2007
Charles Truskolasky
Re: Phase Array for mis-orientated si ----------- Start Original Message -----------
: : What are the depth and sweep angle limitations for finding 2mm flaws with sig/noise ratio limit of 6dB. I need to find a system capable of finding such flaws in fine grain mild steel, in cylindrical bodies with diameters to 48-in and depths extending to the center of these diameters. I want to use longitudinal w. sweeps 0 to 35-deg. and 45-deg shear w., in the circumferential directions. The sweep is necessary to locate flaws that lie at various angles to the surface. Is anyone doing this?
: It is difficult to answer this question without more information. First, are these defects omni-directional reflectors like inclusions, or planar like cracks? If the latter, you need to tailor the incident angles to the defect orientation; if the former, this is less of a big problem. Assuming inclusions or something similar, the frequency will be set by the material properties. (You can check this manually, and use the same frequency for your array definition.) The arry element spacing (gap) will define the amount of sweep that you can manage; let's assume you have many small elements (e.g. 32) which will allow 0-35 L-waves. These are calculatable from standard physics equations, or measurable following ASTM E-2491-06 for PA set-ups. Lastly, how can you get 6 dB SNR on small defects? Probably by focusing with your PA, which can give ~10x the amount of power. Please post typical ultrasonic frequencies, defect types, and planned arrays for further comments.
------------ End Original Message ------------

The required frequency of the process specification is 2.0 MHz with options to got to 4.0 MHz for increased sensitivity. The requirement is to find flaws of equivalent flat bottom holes that lie at various un-predictable orientations. Reporting and rejection criteria is based on amplitudes meeting the levels of 1mm equivalent FBH diameter in the near surface region, eg. within 10-in of surface and finding 2mm flaws near the center region. The process requires the system to orientate the beam with an angle normal to the flaw. The spec. doesn't care at this point what the origin of the flaw is but does it meet the FBH requirements. I've calculated an individual .25-in wide element in the array would give me at least a +/- 30-deg beam but I'm concerned that the portable systems available would not have the required amplification or channels for the transducer array.



 
05:22 Jan-22-2007

Michael Moles

,
Joined

Re: Phase Array for mis-orientated si ----------- Start Original Message -----------
: : : What are the depth and sweep angle limitations for finding 2mm flaws with sig/noise ratio limit of 6dB. I need to find a system capable of finding such flaws in fine grain mild steel, in cylindrical bodies with diameters to 48-in and depths extending to the center of these diameters. I want to use longitudinal w. sweeps 0 to 35-deg. and 45-deg shear w., in the circumferential directions. The sweep is necessary to locate flaws that lie at various angles to the surface. Is anyone doing this?
: : It is difficult to answer this question without more information. First, are these defects omni-directional reflectors like inclusions, or planar like cracks? If the latter, you need to tailor the incident angles to the defect orientation; if the former, this is less of a big problem. Assuming inclusions or something similar, the frequency will be set by the material properties. (You can check this manually, and use the same frequency for your array definition.) The arry element spacing (gap) will define the amount of sweep that you can manage; let's assume you have many small elements (e.g. 32) which will allow 0-35 L-waves. These are calculatable from standard physics equations, or measurable following ASTM E-2491-06 for PA set-ups. Lastly, how can you get 6 dB SNR on small defects? Probably by focusing with your PA, which can give ~10x the amount of power. Please post typical ultrasonic frequencies, defect types, and planned arrays for further comments.
: The required frequency of the process specification is 2.0 MHz with options to got to 4.0 MHz for increased sensitivity. The requirement is to find flaws of equivalent flat bottom holes that lie at various un-predictable orientations. Reporting and rejection criteria is based on amplitudes meeting the levels of 1mm equivalent FBH diameter in the near surface region, eg. within 10-in of surface and finding 2mm flaws near the center region. The process requires the system to orientate the beam with an angle normal to the flaw. The spec. doesn't care at this point what the origin of the flaw is but does it meet the FBH requirements. I've calculated an individual .25-in wide element in the array would give me at least a +/- 30-deg beam but I'm concerned that the portable systems available would not have the required amplification or channels for the transducer array.
------------ End Original Message ------------

You can probably live quite comfortably with a portable system. Amplification is not usually the key issue - it is signal-to-noise ratio, even in fine-grained steel. Normally a smaller pulser voltage can be compensated by increasing the gain. The electrical noise is usually well below the grain noise, so this is unimportant.

As for number of channels, our instruments are configurable. If you really need a lot of channels (say 64/256), move from OmniScan up to Focus LT, which is not a lot more expensive. Also, with either OmniScan or Focus LT, you can "merge" the data fields in TomoView togive you a complete image. We have done this with titanium castings, for example. Let me send you some images on your company e-mail to show the concept.
Michael




 


© NDT.net - The Web's Largest Portal of Nondestructive Testing (NDT) ISSN 1435-4934

Open Access Database, |Conference Proceedings| |Articles| |News| |Exhibition| |Forum| |Professional Network|