Numerical Modelling of a Cutting Arc Torch

BEATRIZ MANCINELLI; FERNANDO MINOTTI; LEANDRO PREVOSTO; HÉCTOR KELLY

Computational and Numerical Simulations

InTech Open Access Publisher

Año: 2014; p. 65 - 82

Plasma cutting is a process of metal cutting at atmospheric pressure by an arc plasma jet, where a transferred arc is generated between a cathode and a work-piece (the metal to be cut) acting as the anode. Small nozzle bore, extremely high enthalpy and operation at relatively low arc current (10-200) A are a few of the primary features of these torches. The physics involved in such arcs is very complicated. The conversion of electric energy into heat within small volumes causes high temperatures and steep gradients. Dissociation, ionization, large heat transfer rates (including losses by radiation), fluid turbulence and electromagnetic phenomena are involved. In addition, wide variations of physical properties, such as density, thermal conductivity, electric conductivity and viscosity have to be taken into account. These factors make hopeless the possibility of an analytical solution for such arcs. Recently, numerical plasma modelling has reached a state advanced enough to be of practical use in the study of cutting-arc processes. Plasma modeling by numerical simulation in cutting torches is a powerful tool to predict the values of the fundamental physical quantities, namely the plasma temperature, the particles concentration and the fluid velocity. These numerical codes are employed to understand the relevant physical processes ruling the plasma behavior in order to interpret the experimental results of several plasma diagnostics, and ultimately to obtain optimized designs of such devices. However, no complete predictive power has been possible as yet due to the complexity and variety of the processes involved. In particular, the practical use of cutting torch codes requires the introduction of some numerical coefficient whose value has to be obtained from a comparison between the model predictions and the experiment. For these reasons, the experimental validation of such models is of primary importance. On the other hand, experimental data of cutting arcs are hard to obtain due to the harsh ambient conditions, and also because much of the plasma flow takes place in relatively small, bounded regions inaccessible to experimental probing. In practice, most of the available experimental data are related to spectroscopic and probe measurements in the external plasma region (the outer region between the nozzle exit and the anode, where the pressure has relatively small variations about the atmospheric value), giving information of only part of the variables involved, mostly temperatures and species concentrations. More recently, also measurement of flow velocities in cutting torches have been reported by our Group. Because of the above quoted reasons, the experimental validation of the existing cutting torch models has been restricted to the temperature distribution in the nozzle-anode gap, together with other global parameters easy to obtain as the arc voltage and the arc chamber pressure. That experimental validation consists in obtaining a good matching with the experimental temperature values within the experimental uncertainty, which typically amounts to 10 % of the measured value. However, since in a cutting torch the plasma pressure is close to the atmospheric pressure, the plasma temperature is mainly determined by the energy equation (with the dominant energy loses being due to inelastic electron collisions that produce a sort of anchorage in the temperature), and so it is almost decoupled from the hydrodynamic flow. Hence, the calculated temperature value does not result in practice quite sensitive to changes in the numerical coefficients of the models at least within the temperature experimental uncertainty. In high-energy density cutting torches it appears that the flow velocity is a more sensitive variable than the temperature to the modeling details. This can be understood considering the following argument. The acceleration of the flow is driven by a mostly axial pressure gradient along the torch nozzle, established between the pre-nozzle chamber and the nozzle exit. So, given the pressure profile, the velocity of each fluid element at different positions inside the nozzle depends strongly on its density, which in turn depends on the temperature and, in the high velocity, compressible regime, on the velocity itself. As inside the nozzle the radial temperature profile is very sharp, small variations of its shape, for instance its peak width, lead to appreciable changes in the radial mass distribution that reflect directly on the velocity. All this is more clearly seen and quantified using simple 1-D, two zone models in which only axial dependence is considered of an inner hot zone and an outer cold region. By assuming sonic flow at the nozzle exit, these simple 1-D models are able to predict quite reasonably the measured temperatures (within the experimental uncertainty), but give very imprecise values (100% off) of the flow velocity. The purpose of this chapter is to validate the most frequently used plasma cutting torch models employing not only temperature but also velocity values as the experimental data to be confronted with. In order to do this, a 2-D model similar to those proposed in the literature, was developed and applied to the same 30 A oxygen cutting torch that was used in a previous velocity measurement experiment. In this chapter the plasma torch, model assumptions, governing equations, boundary conditions and the physical details of the model are presented. The calculated distributions of temperature and velocity and its comparison with the experimental data are shown.