The post-processing of ultrasonic array data using the total focusing method

The use of ultrasonic arrays for NDT has increased significantly in recent years. This is due to the flexibility of array systems which can be electronically configured to produce plane, focused, steered and steer-focused beams. In this way, one array transducer can do the job of many standard single-element transducers. Currently, the approach is to have independently controlled parallel pulser-receiver channels connected to a number of array elements. These channels allow timing delays to be applied to each element on both transmission and reception. In this way a specific set of transmit-receieve delays are applied to an array to create a specific beam profile. This paper describes an alternative approach in which the complete raw data set of time domain signals from every transmitter-receiver pair is collected, stored and post-processed. Theoretically, the time taken to acquire this data is approximately the same as that taken to perform a B-scan with the array. The key advantage is that post-processing of the complete raw data set enables any beam profile to be recreated. Additionally, this approach allows novel inspections to be performed which would be impossible with single-element transducers and impractical using the traditional phased array controller methodology. This paper concerns one such post-processing algorithm, the Total Focusing Method (TFM). In the TFM, an image is created in which the beam has been focused on every point within the field of view. This optimises the focusing performance of a given array. The use of the TFM is then demonstrated on a number of test structures and is proposed as a superior alternative to conventional array test techniques.