Machine welding can be broken down into two divisions: Mechanised systems, and Automatic systems (see diagram below). Automatic systems can be further divided into Dedicated systems, and Flexible systems (e.g. robotic).
The role and function of sensing in relation to key areas in the diagram are described below.
Mechanised welding requires considerable operator intervention to make process adjustments; for example, performing a joint tracking function, or adjusting welding parameters. In this case, the operator is responsible to perform any necessary sensing and adaptive control functions.
Typical mechanised systems include those using submerged arc welding and utilising a column and boom and tractor welding systems.
Dedicated automatic welding
Dedicated automatic welding systems automate the welding of a very limited range of products. They usually work within the outline of the machine's base. With these systems, an operator will unload the machine, add a new component and then press the start button. No further intervention is required.
In their simplest form, dedicated machines rely on the repeatability of all elements of the system (e.g. piece-parts, jigs, and the welding process) and there is no process-related sensing. More complex systems use welding related sensors (e.g. to provide joint tracking) to help ensure that the welding system works reliably without human intervention. Most systems will have some sensors incorporated as part of their safety systems (e.g. proximity switches). Such systems are widely used in the automotive sub-component supply sector.
Flexible welding (usually robotic welding)
Such systems may work with or without welding related sensor systems (i.e. they may weld repeatable components according to pre-programmed parameter settings). The operator simply performs the unload/load and start operations. Increasingly, robot systems use sensors to perform joint recognition, joint tracking and weld recognition.
Automatic welding becomes adaptive when sensors gather joint information which is then used to modify the way the system operates. For welding, the three important sensing functions are joint recognition, joint tracking and weld recognition.
Joint recognition involves use of vision-based or mechanical sensors to find the weld joint - and its start and end positions. In one method, a voltage is applied to the MIG/MAG welding wire and is used to probe for the joint. The voltage will fall to zero when contact is made with part of the joint, and the machine/robot will remember this position.
Joint tracking involves use of vision, arc parameter or mechanically-based sensors to track the line of a weld joint. The most widely used technique (with the MIG/MAG process) involves measurement of the change in the resistance of the arc and wire extension as a joint edge is approached during weave welding. The technique does not work with aluminium because of its low electrical resistance. Vision-based systems (e.g. Meta, Servo-robot) use forms of triangulation to detect the line of the joint. As shifts in the position of the joint are detected, so the path of the machine (robot) is changed to perform the joint tracking operation.
Weld (joint) recognition. Before changing process parameters to ensure that a joint is filled to the specified shape, the joint's pre-existing features must be recognised. To achieve adaptive control of weld volume/shape it is necessary to have a good understanding of the relationships between joint geometry and the welding parameters. These relationships are not well developed yet and there are few truly adaptive systems making multi-pass welds.
What is automation and when should I consider its use in welding?
For more information, please contact us.