Page 113 - Welding Robots Technology, System Issues, and Applications

99
Sensors for Welding Robots

for most cases, control measures should be taken within 5 mm weld, which usually
is in the order of 0.5 s. It should also be noted that a process disturbance in real
welding usually comes gradually and in such cases, it is important for the
monitoring system to detect this and make a smooth adjustment of the weld
parameters.

From a practical point of view there are many reasons to customize and tune the
control of the process. In most cases, any changes that may affect the quality
gradually increase, like a change in the gap. If the change is abrupt, we might not
be able to compensate for quality by control actions but have to stop the process.
However, in such cases, we can accept waiting for some time with the decision
from the monitoring system. This indicates that it is a matter of compromise
between measured and detected irregularities on the one hand and resulting faults
as defined in the WPS on the other, including trade-offs between quality and
productivity. Unwanted stops during a weld may introduce quality concerns as well
as decrease productivity.

Thus, the use of monitoring techniques should be an integrated part of the robot
welding system including the definition of Welding Procedure Specifications. The
algorithms used to monitor the process will depend on the process and what critical
fault to detect. Another issue is the sensitivity to errors, which must be handled
with an in-depth understanding of the purpose of the monitoring system and what
is does.

3.9 Monitoring System Development – An Example

Some examples from experiments will be described here to show the benefits in
quality control of the welding operation performed by the robot [1]. For this
purpose, it is important to apply a method that is fast and robust with respect to
detection time and number of false alarms. The setting of these parameters should,
for each case, be possible to adapt according to the requirements as there is
generally a trade-off between quality vs productivity.

3.9.1 Short-circuiting GMAW

In order to produce weld joints of uniform weld quality, it is desirable that the weld
process is stable. Experiments have shown that in the short-circuit mode, optimal
process stability occurs when the short-circuit frequency equals the oscillation
frequency of the weld pool. Optimal process stability can also be described to
have:

x A maximum short-circuit rate (Number/s)
x A minimum standard deviation of the short-circuit rate