Page 119 - Welding Robots Technology, System Issues, and Applications

106 Welding Robots
planned one), perform seam tracking and change welding parameters in
real-time. With those capabilities available the system should be capable
of emulating the adaptive behavior showed by the manual welder.
3. Analysis phase: this is normally a post-welding phase where the manual
welder, or welding operator in the case of an automatic welding system,
examines the obtained welds and decides if they are acceptable or if
changes are needed in the two previous phases. In the particular case of an
automatic system, this phase can be performed automatically, or by means
of user input using specific software interfaces. When advance sensors are
used, like laser 3D cameras, this phase can be executed on-line during the
welding phase. This is particularly interesting since evaluation of welding
quality on-line may influence the ongoing welding process.

Consequently, when designing a fully automated robotic welding system all the
above welding phases must be considered as a way to achieve a good performance
and welding quality. The following sections detail some of the relevant problems,
namely: modeling and control the welding process, system interfaces and
programming environments.

4.1 Modeling the Welding Process

Modeling the welding process is basically a theoretical problem (a physics problem
mainly) and a technological problem, i.e., understanding the welding process
requires theoretical studies but also extensive experimentation to obtain the
governing models. Several of the most interesting welding processes were
explained in Chapter 2, giving practical guidelines about the relationships between
the variables and the parameters that characterize the welding process. Part of the
current knowledge on welding is empirical and based on detailed experimentation,
which focuses on technological aspects. Consequently, the strategy used in this
book was to present the most interesting welding processes from a robotics and
automation point of view, focusing on the technological characteristics and
automatic system requirements. The physics of the process is briefly introduced
and the reader referred to other technical publications, as a way to identify the
process parameters relevant for each welding process.

4.1.1 Definition and Detection of the Process Parameters

To design a welding robotic system the first step is to identify the process related
parameters, i.e., the parameters that should be controlled in a way to obtain the
desired quality, also defined by a set of accepted characteristics (Figure 4.1). The
process related input parameters can be classified into three different categories:

1. Primary inputs: variables that can be modified on-line during the
welding process. Taking as example the GMAW process, the primary