Objective

An investigation into the metallurgical properties of the weld and heat affected zone was required.

Results

A specimen from each sample was sectioned, mounted, polished and etched for metallurgical investigation. Figures 1 and 2 show the weld metal in each case. Figure 1 is indicative of the expected austenite/ferrite structure, while Figure 2 shows a single phase ferritic structure (also expected).  No porosity or voids were observed in the weld metal.  In both cases, sufficient weld penetration is evident in the A240 plate (Figures 3 and 4). 

Figures 5 and 6 show the interface between the AISI 409 strip and the weld metal in each case.  No carbide precipitation or evidence of martensite can be seen in the heat affected zone (HAZ) in Sample A.  Although carbide precipitation is not clearly evident in Sample X, there is a greater difference in the grain size in the HAZ and may be evidence of lath martensite in the HAZ.  Vickers microhardness tests taken from the weld through the HAZ and into the parent material show that the HAZ is softer than the parent material. This indicates that the change in microstructural features is a consequence of grain growth.   Further, Figure 6 shows a feature in the microstructure of the weld metal (on the right hand side) which may be evidence of martensite in the weld. 

Examination of the other strip layers is shown in Figures 7 and 8.  Evidence of heavy carbide precipitation can be found in the HAZ in one of the Sample X layers (Figure 8).  Closer examination of the microstructure (Figure 10) shows intergranular precipitation, probably M23C6. These occur when the material is exposed to elevated temperature for a period of time, allowing time for the formation of these carbides on the grain boundaries (of the order of 2-10 minutes above 600ºC).  Such precipitates are sometimes evidence of high heat input. They typically form during the welding time, as well as during cooling, particularly if cooling in the area is slow.  These precipitates drastically reduce the corrosion resistance of the material by taking out large amounts of chromium from the matrix material.  Further, they cause the material to become brittle in this region, making it susceptible to brittle fracture and also lowering the fatigue resistance.  Examination of a similar region in Sample A does not show evidence of such precipitates (Figure 9). 

Conclusions

Ferritic stainless steels can become brittle after welding if M23C6 precipitates are allowed to form on grain boundaries.  Further, if martensite forms in the weld metal or in the HAZ, embrittlement can also occur.  These two phases are thus undesirable as they reduce the corrosion resistance of the material, as well as the toughness and fatigue strength. 
Sample A does not show evidence of the formation of either M23C6 or martensite in the HAZ or in the weld metal.  Sample X, however, shows evidence of the formation of M23C6 precipitates on the grain boundaries in the 3 strips from the surface of the part.  Microscopic examination together with the microhardness tests indicates that softening has occurred in the HAZ, probably as a consequence of grain growth.