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

Subscribe >
Skip to content

Detection of fatigue crack initiation on welded joints

By Yoann Lage

TWI has completed a project investigating detection of fatigue crack initiation on welded joints under bending loading using acoustic emission.

The project was carried out by TWI’s Condition and Structural Monitoring (CSM) section, which provides a complete monitoring solution to multiple industry sectors using technologies including acoustic emission (AE), guided waves and spectrum analysis. Engineers in CSM specialise in monitoring the integrity of structures and machines and using the resulting information to predict failures.

 

Acoustic emission advancements 

In recent years the AE technique has increasingly been employed in structural health monitoring to increase the safety and reliability of structural components. Already a well-established technique to identify and monitor acoustic activity propagated through materials, advancements in AE have reached the stage where it is now possible to identify the initiation of damage, as well as its type and location. This research project at TWI aimed to demonstrate the advantages of using AE monitoring in standard fatigue bending tests of welded joints.

The tests involved were based on standard fatigue test requirements in which specimens are subjected to four-point bending load conditions to determine S‑N curves (Figures 1 and 2).

Objectives

This project’s objectives were to  detect crack initiation and propagation in welded steel structures (specimens) during fatigue bending testing using AE, and then quantify the accumulated damage of the structure, using the recorded data to inform future AE pattern recognition for damage welded structures.

Programme of work

Tests were performed according to a specific standard bending test for welded joints. They were monitored with a Vallen AE acquisition system, using four-channel VS150 sensors. The work programme comprised a set of fatigue tests to verify the repeatability and accuracy of the accumulated damage identification and crack initiation.

During the fatigue tests a soap solution was applied over the cracked zone to track its growth. Results can be seen in Table 1. The comparison between Table 1 and the AE results shown in Figure 3 proves the good correlation between the crack growth and the AE energy. This shows clearly the advantage of using this monitoring method for early identification of crack initiation, which is not possible using the traditional strain measurement or other parameters.

Table 1. Crack size propagation identified using the soap solution at different numbers of cycles.

Cycles 50mm 60mm 78mm
2c (mm) a (mm) 2c (mm) a (mm) 2c (mm) a (mm)
50000            
100000 9 0.3 5.3 0.4    
150000 11.1 0.5 7.0 0.6 3.5 0.3
200000 22 1.0 4.4 0.5
250000 27 1.8 5.5 0.8
300000 54 2.8
340368 71 3.7
349159 74 4.4

 

With the adopted AE software it was also possible to identify the locations of the accumulated damage to the structure. Figure 4 shows the location of the acoustic emission events associated to the damaged weld, based on the propagation time of the signals from four different sensor positions.

Future developments

The AE monitoring technique represents an accurate method of identifying crack initiation and crack propagation, and a potential solution for structural monitoring. To continue the presented work, a new set of experiments will be prepared to investigate the correlation between the acoustic emission signals and the crack growth phenomena. The results obtained from this work will also support new core research projects examining pipework fatigue by induced vibrations, an area of investigation of interest to the oil and gas sector.

Figure 1. Experimental setup with Vallen AE system and bending machine
Figure 1. Experimental setup with Vallen AE system and bending machine
Figure 2. Joint welded steel specimen with the installed VS150 AE sensors
Figure 2. Joint welded steel specimen with the installed VS150 AE sensors
Figure 3. Acoustic emission results: amplitude peak energy on the left axis and cumulative energy on the right, per number of cycles
Figure 3. Acoustic emission results: amplitude peak energy on the left axis and cumulative energy on the right, per number of cycles
Figure 4. Damage locations' identification
Figure 4. Damage locations' identification

For more information on the results of this project or TWI’s other work relating to condition and structural monitoring, email contactus@twi.co.uk

Avatar Yoann Lage Project Leader – Condition and Structural Monitoring

}