TY - GEN
T1 - NUMERICAL MODELING OF ENHANCED NITROGEN DISSOLUTION DURING GAS TUNGSTEN ARC WELDING
AU - Palmer, Todd A.
N1 - Publisher Copyright:
© 2001 59th Electric Furnace Conference and 19th Process Technology Conference Proceedings. All rights reserved.
PY - 2001
Y1 - 2001
N2 - A model to calculate the nitrogen concentration in the weld metal during the GTA welding of iron has been developed here. Nitrogen concentrations in the weld pool are based on the combination of calculations involving the plasma phase above the weld pool, the interface between the weld pool surface and the plasma phase, and the weldment interior. Monatomic nitrogen partial pressures are calculated as a function of the electron temperatures in the plasma phase, and nitrogen concentrations on the weld pool surface are based on both the monatomic nitrogen partial pressure and the weld pool surface temperatures. Once absorbed at the weld pool surface, nitrogen is then transported to the weldment interior predominantly by convection. The presence of turbulence in the weld pool, which is taken into account by increasing the nitrogen diffusion coefficient, further enhances the transport of nitrogen. Nitrogen desorption, which occurs via bubble formation at the liquid metal surface, is characterized by a supersaturation of nitrogen in the weld metal and is also considered in this model. To test the validity of this model, several autogenous GTA welding experiments in pure iron have been performed at two travel speeds with a number of nitrogen additions to the argon shielding gas. Nitrogen concentrations have also been measured in each weld and compared with the modeling results. The general shape and size of the experimental and modeled weld pools are similar. Both the modeling and experimental results produce nitrogen concentrations between 2.7 and 4.7 times higher than Sieverts' Law calculations for a temperature of 2000 K and nitrogen partial pressures between 0.005 and 0.020 MPa. When the modeling and experimental results are compared, both results are equivalent in magnitude for a given set of welding parameters and follow similar trends with changes in the nitrogen addition to the shielding gas and the travel speed. The modeling calculations also display several features, which contribute to these results. Electron temperatures in the plasma phase adjacent to the weld pool in a range around 3000 K are found to produce levels of monatomic nitrogen sufficient to produce nitrogen concentrations in the weld pool equivalent to the experimental results. Levels of nitrogen supersaturation between 50 and 75% higher than the equilibrium nitrogen concentration are required to produce the nitrogen concentrations equivalent to the experimental results. The incorporation of turbulence in the calculations is also a key component in the final results.
AB - A model to calculate the nitrogen concentration in the weld metal during the GTA welding of iron has been developed here. Nitrogen concentrations in the weld pool are based on the combination of calculations involving the plasma phase above the weld pool, the interface between the weld pool surface and the plasma phase, and the weldment interior. Monatomic nitrogen partial pressures are calculated as a function of the electron temperatures in the plasma phase, and nitrogen concentrations on the weld pool surface are based on both the monatomic nitrogen partial pressure and the weld pool surface temperatures. Once absorbed at the weld pool surface, nitrogen is then transported to the weldment interior predominantly by convection. The presence of turbulence in the weld pool, which is taken into account by increasing the nitrogen diffusion coefficient, further enhances the transport of nitrogen. Nitrogen desorption, which occurs via bubble formation at the liquid metal surface, is characterized by a supersaturation of nitrogen in the weld metal and is also considered in this model. To test the validity of this model, several autogenous GTA welding experiments in pure iron have been performed at two travel speeds with a number of nitrogen additions to the argon shielding gas. Nitrogen concentrations have also been measured in each weld and compared with the modeling results. The general shape and size of the experimental and modeled weld pools are similar. Both the modeling and experimental results produce nitrogen concentrations between 2.7 and 4.7 times higher than Sieverts' Law calculations for a temperature of 2000 K and nitrogen partial pressures between 0.005 and 0.020 MPa. When the modeling and experimental results are compared, both results are equivalent in magnitude for a given set of welding parameters and follow similar trends with changes in the nitrogen addition to the shielding gas and the travel speed. The modeling calculations also display several features, which contribute to these results. Electron temperatures in the plasma phase adjacent to the weld pool in a range around 3000 K are found to produce levels of monatomic nitrogen sufficient to produce nitrogen concentrations in the weld pool equivalent to the experimental results. Levels of nitrogen supersaturation between 50 and 75% higher than the equilibrium nitrogen concentration are required to produce the nitrogen concentrations equivalent to the experimental results. The incorporation of turbulence in the calculations is also a key component in the final results.
UR - http://www.scopus.com/inward/record.url?scp=85197494355&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85197494355&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:85197494355
T3 - 59th Electric Furnace Conference and 19th Process Technology Conference Proceedings
SP - 647
EP - 658
BT - 59th Electric Furnace Conference and 19th Process Technology Conference Proceedings
PB - Iron and Steel Society
T2 - 59th Electric Furnace Conference and 19th Process Technology Conference
Y2 - 11 November 2001 through 14 November 2001
ER -