The thermal or process efficiency in gas metal arc welding (GMAW) is a crucial input to numerical models of the process and requires the use of an accurate welding calorimeter. In this paper, the authors compare a liquid nitrogen calorimeter with an insulated box calorimeter for measuring the process efficiency of Fronius cold metal transfer, Lincoln surface tension transfer and RapidArc, Kemppi FastRoot and standard pulsed GMAW. All of the controlled dip transfer processes had a process efficiency of y85% when measured with the liquid nitrogen calorimeter. This value was slightly higher when welding in a groove and slightly lower for the RapidArc and pulsed GMAW. The efficiency measured with the insulated box calorimeter was slightly lower, but it had the advantage of a much smaller random error.

Split anode calorimetry investigations of tungsten inert gas (TIG) arcs were conducted using a large diameter shielding gas nozzle. Some conditions displayed arc symmetry whilst others proved distinctly asymmetric. The variation of welding current and electrode tip to workpiece distance (ETWD) was studied. Decreasing the ETWD was found to increase the current density towards the arc axis but similar to a previous study of the authors, Gaussian distributions were not observed. The gas nozzle was designed to produce laminar gas flow and sound shielding behaviour, however, anode surface oxidation was found after welding; presumably caused by shielding gas contaminated with oxygen through the welding sequence. Therefore, axial arc symmetry was influenced by random effects. The conditions and reasons for the observed phenomena are explored within this paper.

Most previous split anode calorimetry research has applied high weld currents which exhibit pseudo Gaussian distributions of arc current and power density. In this paper we investigate low current arcs and show that both the current and power distributions have minima in the centre-varying significantly from the expected Gaussian profile. This was postulated due to the formation of the arc with the copper anode and the tungsten cathode. Furthermore, a number of parameters were varied including the step size between measurements, anode thickness and anode surface condition as well as cathode type and tip geometry. The step size between measurements significantly influenced the distribution profile and the anode thickness needed to be above 7 mm to obtain consistent results.

An investigation was raising the question: "What does 'arc length' mean?" Actually, it is considered expressing a kind of natural relationship between arc voltage and arc column shape. Statements such as "The higher the voltage the longer the arc" or "The arc voltage proves approximately proportional to the arc length", are frequently noticed in this conjunction. However, the author suggests that there is no general possibility of describing 'arc length' over the whole welding process range. Instances are represented in this paper, showing both theoretical attempts of definition and practical observations. This paper intends to contribute to a serious discussion of something trivial, indeed very well-known or used among welding experts, but actually yet hardly understood, at least as when it comes to closer examination