Determination of Resistance Spot Weld Quality in Real Time Using Reflected Acoustic Waves. Comparison with Through-transmission Mode

The determination of quality parameters of resistance spot weld in real time is of great importance in today’s industry. Acoustic waves can solve this problem in the sense of measuring certain features directly related to the weld quality – degree of heating of the weld zone, moment of beginning of melting and some others. The ultrasonic transducer incorporated into an electrode sends short acoustic pulses into the weld area and receives the signals reflected from different interfaces in heat affected zone during welding process. The specific image and signal processing algorithms applied to the received signals enable us to localize weak reflections from internal interfaces, recognize their nature and use this information to characterize the weld quality. Introduction: The inspection of resistance spot weld quality in real time is quite a challenging and at the same time promising field of NDT. It enables one to monitor the dynamics of the spot weld properties right at the moment of the weld manufacturing. It is extremely important on the industrial floor as it gives the operator information about the quality of the product and thus enables him to correct the mistake before it is too late. Quality inspection of “hot” welds is relatively new field of NDT studies which is potentially capable of providing 100% quality output of the production line. It becomes especially important as long as implementation of the in-line quality inspection techniques will help save a lot of time by elimination or, at least, reduction of costs of after-production inspection and fixing bad parts. Our approach to the above mentioned problem is using the acoustic waves which pass through the spot weld structure during welding. On their way through the setup the waves are modified by the media and thus carry some information about the processes taking place inside the system. The setup consists of a set of one or two ultrasonic transducers incorporated into the electrodes of the spot welder. The ultrasonic transducers are driven by pulser-receiver; the last is controlled by the computer. The schematic view of the setup is presented in Figure 1. Figure 1. Schematic view of the experimental setup. 1 – incident wave, 2 – reflected wave, 3 – transmitted wave. The detailed description of the operational process of the setup can be found in [1, 2]. In this article we will present the most important results obtained previously in the transmission mode and compare them with the reflection mode. The comparison of the two methods and advantages of application of either method will be discussed in this paper. Results: In transmission mode a series of pulses 1 is sent through the weld zone (see Figure 1) and is picked up by the receiving transducer as waves 3. As the welding current heats and melts the media, the material properties at different moments of welding change much. For this reason the waves passing through the weld at different moments of welding differ much from each other. The comparison of the wave characteristics provides useful information about the processes inside the stack-up. The finite-difference model of spot weld formation described in [2] shows the dynamics of the temperature and some material properties inside the weld zone during the spot weld formation. Figure 2 shows the temperature distributions in the spot weld at different phases of welding procedure. The calculations have shown very high inhomogeneity of the temperature distributions along the vertical direction in the weld setup. The temperature gradients in the middle of the structure reach 700-800 K/mm. The longitudinal velocity gradients at the same region are in the order of 1000 m/s per mm. It means that in the center of the liquid (white) area the velocity of sound propagation is around 3800 m/s while at the steel-copper interface the velocity is close to 5000 m/s. The same high gradients are for the other material properties such as stiffness, thermal expansion coefficient, density, etc. The acoustic impedance mismatch at all interfaces is also dependent on the temperature distribution at the given moment. Acoustic reflectivity and acoustic transparency of the structure also fluctuate during welding, thus affecting the amplitude and the magnitude spectrum of the signal. All these factors give rise to the modifications the wave experiences on its way through the weld zone. Previous analysis [1] has shown that among half a dozen of different parameters of the signal the time of flight is the most reliable one. There was found high correlation of throughtransmitted signal delay with the nugget size of the spot weld. Figure 3 shows the signature of the weld obtained in transmission mode. Such an image is sometimes called M-scan; here each vertical line represents one A-scan obtained consequently during 500 ms. In the given picture there are 200 waveforms put together to form an image.