The susceptibility to solidification cracking of unstabilized and stabilized ferritic stainless steels was investigated using self-restrained Houldcroft, Modified Varestraint-Transvarestraint (MVT), and hot tensile testing. Five experimental steel grades comprising an unstabilized, two mono stabilized (Ti or Nb), and two dual stabilized (Ti + Nb), and two commercial unstabilized and a dual stabilized (Ti + Nb), and another dual stabilized containing-Mo alloy (nine different alloys in total) were used in this study. Seven steel grades comprising an unstabilized, two mono stabilized (Ti and Nb) respectively, three dual stabilized (Ti + Nb) and a dual stabilized containing Mo were used for the self-restrained Houldcroft method. Autogenous gas tungsten arc welding at a speed of 6 mm/s, 3 mm/s, and 1 mm/s was done. The unstabilized ferritic stainless steel was resistant to solidification cracking. Ti addition to ferritic stainless steel resulted in a minor increase to susceptibility to solidification cracking. Nb in ferritic stainless steel increased solidification cracking. The addition of Ti and Nb resulted in a decreased susceptibility to solidification cracking compared to an alloy containing only Nb. The weld metal microstructures were a mixture of columnar and equiaxed grains. The interdendritic crack surfaces were enriched in Nb, Ti, Mn, Si, Al, Mn, and Mo. The MVT test was used for the test of an unstabilized, a Nb stabilized and two (Ti + Nb) dual stabilized ferritic stainless steels. Two different welding speeds of 6 mm/s and 3 mm/s using autogenous gas tungsten arc welding were employed. The high content (Ti + Nb) steel at a welding speed of 3 mm/s had the greatest sensitivity to solidification cracking. The Nb stabilized steel at both welding speeds (6 mm/s and 3 mm/s) and high content (Ti + Nb) steel at a welding speed of 6 mm/s showed intermediate sensitivity to solidification cracking. The unstabilized and low content (Ti + Nb) grades were the least sensitive to solidification cracking. The weld metal microstructures transverse to the welding direction revealed columnar grains in all the samples for both welding speeds. Three experimental Ti-, Nb-, and dual Ti + Nb stabilized ferritic stainless steels were used for hot tensile testing using a Gleeble-1500D thermo-mechanical machine at testing temperatures of 1200°C, 1250°C, and 1300°C. The dual stabilized ferritic stainless steel showed a high and fairly constant hot ductility with an increasing testing temperature. The Ti stabilized alloy revealed a slightly lower ductility compared to the dual stabilized steel but much higher ductility than the Nb stabilized ferritic stainless steel. The SEM images of the intergranular cracking showed interdendritic morphologies. EDX analysis showed the elements Al, Mn, Ti, Si, Ni, S, Nb, and Ni to be associated with the fractured surfaces. The hot tensile test results were inconclusive, due to the small number of samples and an acquisition frequency that was too low. The MVT test was better than the self-restrained Houldcroft, and the self-restrained Houldcroft was better than the hot tensile tests in quantifying the susceptibility of a specific ferritic stainless steel alloy to solidification cracking. The cracking response of Houldcroft seemed to be dominated by welding speed. Cracking response of MVT test seemed to be dominated by the Nb content. The effect of Nb and Ti on the susceptibility cracking could be explained in terms of the effect of these two alloying elements on the difference between the liquidus and the solidus. Nb was found to segregate strongly to the grain boundaries (low k value) which resulted in a significant increase in the difference between the liquidus and the solidus. This difference increased BTR which results in a high susceptibility to solidification cracking. Ti has a higher k value and segregates less than Nb during solidification. / Thesis (PhD)--University of Pretoria, 2018. / Materials Science and Metallurgical Engineering / PhD / Unrestricted