Published: 2001
Total Pages:
Get eBook
Nitrogen concentrations far in excess of Sieverts' Law calculations and as high as 0.2 wt.% have been obtained in steel welds during arc welding. Such high concentrations of nitrogen in the weld metal can originate from a variety of sources, depending on the welding operation in question. One such mechanism involves the interaction between the surrounding atmosphere, which is about 80% nitrogen, and the plasma phase above the weld pool. Impingement of the surrounding atmosphere into the arc column, which is primarily composed of an inert shielding gas, can be due, in part, to insufficient shielding of the weld metal. In other cases, nitrogen can be purposefully added to the shielding gas to enhance the microstructural evolution of the weld metal. The mechanisms responsible for enhanced nitrogen concentrations are of significant interest. In both arc melting and welding operations, a plasma phase exists above the liquid metal. This plasma phase, which is composed of a number of different species not normally observed in gas-metal systems, significantly alters the nitrogen absorption reaction in liquid iron and steel. Monatomic nitrogen (N) is considered to be the species responsible for the observed enhancements in the nitrogen concentration. This role for monatomic nitrogen is based on its significantly higher solubility in iron with partial pressures many orders of magnitude less than that for diatomic nitrogen. It has also been proposed that the total amount of nitrogen present in the liquid metal is the balance of two independent processes. Monatomic nitrogen is absorbed through the interface between the arc and the liquid metal. Once a saturation level is reached at any location on the metal surface, nitrogen is then expelled from the surface of the liquid metal. This expulsion of nitrogen from the weld pool surface occurs via a desorption reaction, in which bubbles form at the surface and other heterogeneous nucleation sites in the liquid melt. These bubbles are filled with nitrogen gas, which has been rejected from the liquid iron. Outside the arc column, the nitrogen in solution in the iron is in equilibrium with diatomic nitrogen rather than monatomic nitrogen, which dominates the arc column. Models based on the role of the plasma phase in producing these enhanced nitrogen concentrations have also been developed. For example, Gedeon and Eaga have proposed that the diatomic gas introduced into the plasma phase in the arc column partially dissociates at a temperature higher than that at the sample surface. The monatomic species is then transported to the liquid metal surface, where it is absorbed at the temperature on the liquid metal surface. Mundra and DebRoy have used this same methodology to develop a semi-quantitative model to describe the temperature at which the diatomic gas dissociates in the plasma phase. In the two-temperature model, a hypothetical temperature, T{sub d}, equal to the temperature at which the equilibrium thermal dissociation of diatomic nitrogen produces the partial pressure of monatomic nitrogen in the plasma, is defined. This dissociation temperature is in a range of 100 to 300 K higher than the temperature at the metal surface, T{sub s}, and is a measure of the partial pressure of the atomic nitrogen in the plasma. This methodology provides an order-of-magnitude agreement between the calculated and experimental nitrogen concentrations but does not strictly provide a capability for predicting the nitrogen concentration. No quantitative means for predicting the nitrogen concentration in the weld metal currently exists. In developing a quantitative model, it must be recognized that nitrogen dissolution into the weld pool is intimately tied to several simultaneously occurring physical processes. These processes include the formation of various nitrogen species in the plasma phase above the weld pool, reactions at the interface between the plasma phase and the weld pool surface, and the transport of nitrogen within the weldment by convection and diffusion. A mathematical model, which combines calculations describing each of these processes into a single model, has been developed here. The validity of this model has also been tested by comparing the modeling results with those from a series of GTA welding experiments with pure iron.
