Adaptive welding enables dynamic altering of the welding parameters to compensate for changing environment. Sensors providing process status information in real time are an integral part of such an adaptive system. In this investigation, infrared thermography was used as a sensor to control the position of the arc and the penetration depth of the weld. Preliminary work on infrared thermography showed that variation in these parameters produces a change in the surface temperature distributions of the plates being welded. Subsequently, to achieve computer control of these variables, image analysis techniques have been developed to quantify the changes in the temperature distribution.

Infrared sensing techniques used in are welding and resistance welding process monitoring and control are presented. A variety of IR sensors have been investigated for use in monitoring and controlling weld bead width, penetration depth, seam tracking and cooling rate during are welding, and nugget size and cooling rare during resistance welding. In addition, factors that affect or interfere with infrared sensing are discussed.

The thermal distribution changes associated with gas tungsten are welding process were studied to identify and correct for weld joint offsets for butt joints. These changes were experimentally measured using infrared thermography. In the weld offset conditions the slope of the temperature distributions were found to be altered dramatically. Not only did offset produce asymmetric temperature distributions but a steep increase in gradients was also observed. These changes were then used to detect and correct initial errors in the position of the torch during the welding of straight and curved contours of perfectly fitted butt joints and joints with gap

Sensing elements need to be incorporated in robotic welding systems to enable the robot to perceive and adapt to on-line variations occurring in the welding process. In this work, infrared thermal imaging techniques have been used to track variations produced by inadequate control during the joint preparation and fixturing stages. Variations in two joint parameters, gap and position, were studied. Changes in these parameters were found to have peculiar effects on the surface temperature distributions. The observed effects were used to develop quantitative error signals. These error signals were then used to measure the joint gaps and joint-torch offsets in real-time. The joint torch offset error signal was successfully used to control an initial error in joint position during real-time welding.

Heat is deposited in a target material when it is impacted by a projectile. The feasibility of using an infrared thermographic system to measure the surface temperature profile of composite materials under ballistic impact was demonstrated. Three types of composites were used in this instrumentation study to illustrate the effectiveness and limitation of this technique. The resolution of temperature was found to be better than ±1 K. The total temperature rise in the materials during impact can be attributed to the deformation required to perforate the material as well as frictional heat produced between the projectile and the hole. The heat deposited was calculated based of the temperature profiles. The relative contribution of the frictional effect to the total energy absorption was evaluated using a sequential shot technique. Effects of material properties and impact parameters on the peak temperature and heat deposition were determined. Graphite and PE composites are more effective than Kevlar composite in dissipating heat during the ballistic impact penetration process due to their higher thermal conductivity. The correlation between heat dissipation and extent of damage in composites was also investigated.

Integration of sensors in a welding system enhances the quality of the welds produced. Variations in three welding process parameters-weld bead width, penetration depth, and torch position-were monitored using an infrared sensor. Intentionally induced variations in each of these welding parameters were found to affect uniquely the plate surface temperature distributions measured by the infrared sensor. The effects of weld bead width and torch position perturbations on the temperature distribution were separated so as to identify and control these two weld process parameters simultaneously. Preliminary results suggest that simultaneous penetration depth, bead width, and torch position control is possible.< >

Infrared sensing techniques were investigated to assist remote welding systems to identify and correct weld-joint offsets in real time. During the welding process, the temperature distribution along a line normal to the joint and ahead of the arc was measured. A distinct drop in the measured temperature distribution was observed to coincide with a gap in the joints. In the temperature gradient profile, the gap was characterized by three changes in sign. The second change in sign of the temperature gradient profile was found to correspond to the gap center. The first and third changes in sign were found to coincide with the edges of the joint. These unique changes were used to successfully track curved contours of joints with a gap.

Robotic arc welding, as practiced in industry today, is accomplished by programming a robot to move its welding torch through a set of spatial coordinates that coincide with the position of a part's joint. Joint tracking systems are currently used to reduce the costs of part preparation and fixturing. In this investigation, infrared thermography was used to joint track in gas tungsten arc welding. An infrared camera was used to record the temperature gradients surrounding the welding torch and transmit the images to a central computer. There, using an experimentally derived image processing technique, the computer determined the distance of the torch from the joint and transmitted corrective action to control the torch path

Implementation of robotics into welding is an important step towards higher productivity and better quality control of the fabrication process. However most robots currently perform welding in a "blind" fashion. In order to enhance the intelligence of the robots, several sensing techniques such as laser stripping, through-the-arc sensing and infrared thermography have been investigated. Most of these sensing techniques are capable of monitoring only a single welding parameter. However infrared thermography has shown promise t9 detect several types of impending weld defects. 1,4 The results presented in this paper identify approaches to obtain quantitative relationships to monitor the two major weld parameters, torch position and penetration depth. The asymmetry of the thermal profiles caused by an arc misalignment has been quantified into a torch seam error relationship by comparison of the features of thermal profiles on either sides of the calibrated torch position. To monitor the weld penetration depth the thermal distribution of the plates being welded were fitted to an ellipse using a least squares method. The principle features of the ellipse were found to be sensitive to the penetration depth of the plate being welded.

Penetration depth is a key variable, which needs to be controlled to ensure defect-free welds. One of the major problems involved with adaptive control penetration is the lack of suitable variables which can be viewed directly by the sensor. A proposed sensing technique is to relate the invisible variables, such as penetration depth and thickness of the welded steel plates, to visible variables of thermal images. The sensed infrared information was obtained through digital signal process of the thermal images associated with the high temperature of the molten metal pool and its vicinity during the welding process. Quantitative measurements were conducted to find a relationship between visible and invisible parameters. The thickness of plates being welded was varied and the corresponding changes in both penetration depth and surface temperature distributions were studied quantitatively. A least squares method was used to fit the obtained isotherms to an equation of an ellipse. The penetration depth and thickness of the materials being welded were found to be functions of the minor axes and the area of the ellipse for Gas Tungsten Arc Welding (GTAW). The experimental results can be used to achieve adaptive penetration depth control.

Implementation of robotics into welding is an important step toward higher productivity and better quality control of the fabrication process. However most robots currently perform welding in a "blind" fashion. In order to enhance the intelligence of the robots several sensing techniques such as laser stripping, through-the-arc sensing and infrared thermography have been investigated. Most of these sensing techniques are capable of monitoring only a single welding parameter. However infrared thermography has shown promise to detect several types of impending weld defects. 1,2 The results presented in this paper identify approaches to obtain quantitative relationships to monitor the torch position with respect to the seam of the plates being welded. The asymmetry of the thermal profiles caused by arc misalignment has been quantified into a torch seam error relationship. Two principal comparison techniques have been identified. Averaging methods have been implemented to reduce the noise level in the error signals.