Laser welding needs precise measurement of weldgap position to avoid weld defects. Most often, optical measurement methods are used, but well-aligned narrow gaps canbe difficult to detect. An improved inductive method capable of detecting zero gaps in square butt joints is proposed. The new method uses two eddy current coils, one on each side of the gap, and measures the complex response of the individual coils, i.e. both the inductive and resistive response. By combining the coil responses, both the position and the geometry of the weld gap can be estimated. The method was experimentally investigated by traversing a single coil over an adjustable gap between two plates and combining the measured coil responses into a simulated two-coil probe. The gap was adjusted in both misalignment and gap width up to 0.4 mm. Comparing the results to known settings and positions shows that gap position is measured to within 0.1 mm, if the probe is within a working area of 1 mm from the gap in both position and height. Results from the new method were compared to simulations, from the same experimental data, of a previously reported method where the coils were electrically combined by wiring them together. The previous method can give accurate results but has a much smaller working area and depends on servo actuation to position the probe above the gap. The improved method gives better tolerance to varying misalignment and gap width, which is an advantage over previous inductive methods.

The use of automated laser welding is a key enabler for resource efficient manufacturing in several industrial sectors. One disadvantage with laser welding is the narrow tolerance requirements in the joint fit-up. This is the main reason for the importance of joint tracking systems. This paper describes anevaluation of four non-contact measurement methods to measure the position, gap width and misalignment between superalloy plates. The evaluation was carried out for increased knowledge about the possibilities and limitations with the different methods. The methods are vision-, laser-line-,thermography- and inductive probe systems which are compared in an experimental setup representing a relevant industrial application. Vision is based on a CMOS camera, where the image information is used directly for the measurements. Laser-line is based on triangulation between a camera and a projected laserline. Thermography detects the heat increase in the gap width due to external heat excitation. Inductive probe uses two eddy current coils, and by a complex response method possibilities to narrow gap measurement is achieved. The results, evaluated by comparing the data from the different systems, clearly highlights possibilities and limitations with respective method and serves as a guide in the development of laser beam welding.

This paper presents an improvement to a recently presented inductive gap measurement method, using a model to reduce systematic errors. Gap measurement is important in laser keyhole welding, where the laser beam and the resulting weld seam are very narrow, requiring high precision in alignment and gap preparation. The previously reported method for gap measurement uses one inductive coil on each side of the gap, each measuring distance to the gap and lift off above a plate, to estimate the position, width and alignment of the gap in a square butt joint. The method can detect zero width gap and shows position error less than 0.1 mm, but gap width and alignment measurement suffer from systematic errors. The improvement is based on a model that is designed to describe these systematic errors as functions of the gap dimensions. The model relies on observations of experimental data, and is calibrated to a small set of measurements. Using the model with the initial estimate of the gap dimensions to compensate the coil measurements, an improved estimate of the gap dimensions can be calculated. The errors in the compensated results are within 0.1 mm except for gap width, which still suffers from the effect of combined gap width and misalignment.

Finite element modelling of high frequency inductive coils and metal plates was used to investigate the impedance response to a narrow gap between the plates. The model was used to predict distance sensitivity for both coil-to-gap and coil-to-plate. Different coil axis orientations, along and across the gap and normal to the plate, were modeled, as well as different coil lengths. The model can be used to predict the working range and select orientation and geometry of inductive seam tracking probes with zero gap capability for precision laser beam welding. Different materials, thicknesses and frequencies can be used.