Subscribe to our newsletter to receive the latest news and events from TWI:

Subscribe >

Nonlinear ultrasonic phased array imaging of a fatigue crack

There is a major industrial push to be able to detect ‘closed’ defects such as fatigue cracks and other tightly closed cracks.  These may be invisible to currently available NDT inspection techniques which are based on principles of linear wave propagation.  In this project, a nonlinear technique using standard ultrasonic equipment was investigated.  The technique is based on the movement of energy away from the transmission bandwidth, to other frequencies, in a nonlinear regime.

The majority of ultrasonic inspection techniques rely on the assumption of linear wave propagation and the derived properties.  However fatigue cracks and certain other ‘closed’ defects may exhibit nonlinear responses due to contact between the surfaces, friction or ‘clapping’ (Solodov, 2009 and Wright et al, 2017). Therefore nonlinear ultrasonic approaches are being developed which may have the potential to detect such flaws even in the presence of a much stronger linear reflector.  However, such nonlinear approaches have not yet become widely available within the NDT marketplace. 

One nonlinear approach that was developed at the University of Bristol (Potter et al, 2014 and Cheng et al, 2017) has the advantage that it uses standard ultrasonic phased array equipment combined with a post-processing technique which detects a migration of energy outside the transmission bandwidth to the harmonic frequencies.

The core objective of the project was to implement, and evaluate, the technique of Potter et al (2014) in terms of its potential application in industry to detect closed defects.  The specific objectives were to:

  • Implement MATLAB code to allow for data acquisition from the Peak NDT Micropulse hardware, in accordance with the requirements of the technique.
  • Write and develop algorithms to process and image the acquired data using the proposed nonlinear method.
  • Evaluate the technique through application to a fatigue crack.
  • Determine the suitability for industrial use.
Aluminium sample with fatigue crack used in the experiments. All dimensions in millimetres
Aluminium sample with fatigue crack used in the experiments. All dimensions in millimetres
  • Code was implemented within MATLAB to allow for data acquisition from the MicroPulse in 16bit data format for both Full Matrix Capture (FMC) and phased array modes.
  • These two modes are used for focusing at each pixel (x,y) of the image area.  Diffuse field data is collected by application of a  time delay prior to reception..
  • Frequency domain imaging algorithms were developed within MATLAB in accordance to the proposed technique.  For each pixel  of the image, a nonlinear metric, γ(x,y), is calculated by subtracting the FMC and phased array data.  (In a linear regime the FMC and phased array data would be equal, however in a nonlinear regime, this is no longer the case.  The phased array test has higher incident energy and so the nonlinear effects at the fatigue crack tip are stronger.)
  • Tests were carried out on an aluminium sample containing a fatigue crack. The suitability of the technique was evaluated.

The nonlinear imaging technique of Potter et al (2014) has been successfully implemented and experimentally proven.  On the sample used, the signal from the crack tip had been just visible using the (linear) FMC technique, but was much weaker than the notch and back wall signals (see Figure 2a – over).

On the other hand, in the nonlinear imaging technique the fatigue crack produced a significantly nonlinear response with nonlinear metric γ=22.5%. There was no nonlinear response form notch or back wall, so the clarity with which the crack was seen was much improved.

However, this result was only achieved after very many repeat trials with varying experimental parameters.  Results were extremely sensitive to the choice of parameters, such as time delays and acquisition intervals, as well as assumed velocities and the exact probe position.  The optimum parameters will vary on a case-by-case basis according to the material properties, specimen geometry, specimen condition, and random noise (Cheng et al, 2017).  If non-optimum parameters were selected, we found that the nonlinear technique gave no detectable response above the linear methods.

Furthermore, the data acquisition for the method is very slow.  The most time-intensive part is the need for a separate phased array firing for each image pixel.

Due to the extreme sensitivity to variables and the slow acquisition speed, it is not recommended that the technique be adopted for industrial application.

References

Cheng J, Potter J N, Croxford A J and Drinkwater B W, ‘Monitoring fatigue crack growth using nonlinear ultrasonic phased array imaging’, Smart Materials and Structures 26 (2017)

Potter J N, Croxford A J and Wilcox P D, ‘Nonlinear Ultrasonic Phased Array Imaging’, Physical Review Letters 113, 144301 (2014)

For further information please email contactus@twi.co.uk.

Image of the fatigue crack using (a) FMC, (b) the nonlinear technique
Image of the fatigue crack using (a) FMC, (b) the nonlinear technique
Media /

Related Insights

View all insights
}