Blending of phased array data

Arno Duijster*, Gert Jan Van Groenestijn, Paul Van Neer, Gerrit Blacquière, Arno Volker

*Corresponding author for this work

Research output: Chapter in Book/Conference proceedings/Edited volumeConference contributionScientificpeer-review


The use of phased arrays is growing in the non-destructive testing industry and the trend is towards large 2D arrays, but due to limitations, it is currently not possible to record the signals from all elements, resulting in aliased data. In the past, we have presented a data interpolation scheme 'beyond spatial aliasing' to overcome this aliasing. In this paper, we present a different approach: blending and deblending of data. On the hardware side, groups of receivers are blended (grouped) in only a few transmit/recording channels. This allows for transmission and recording with all elements, in a shorter acquisition time and with less channels. On the data processing side, this blended data is deblended (separated) by transforming it to a different domain and applying an iterative filtering and thresholding. Two different filtering methods are compared: f-k filtering and wavefield extrapolation filtering. The deblending and filtering methods are demonstrated on simulated experimental data. The wavefield extrapolation filtering proves to outperform f-k filtering. The wavefield extrapolation method can deal with groups of up to 24 receivers, in a phased array of 48 × 48 elements.

Original languageEnglish
Title of host publication44th Annual Review of Progress in Quantitative Nondestructive Evaluation, Volume 37
PublisherAmerican Institute of Physics
Number of pages11
ISBN (Electronic)9780735416444
Publication statusPublished - 20 Apr 2018
Event44th Annual Review of Progress in Quantitative Nondestructive Evaluation, QNDE 2017 - Provo, United States
Duration: 16 Jul 201721 Jul 2017


Conference44th Annual Review of Progress in Quantitative Nondestructive Evaluation, QNDE 2017
Country/TerritoryUnited States


Dive into the research topics of 'Blending of phased array data'. Together they form a unique fingerprint.

Cite this