African Fusion March 2015

Figure 13: LM micrograph of Weld D2 showing a martensitic structure in the COST FB2 base material.

Figure 14: LM micrograph of Weld D2, which also reveals a martensitic structure in the COST F base material.

Figure 12: The crossweld hardness profile of Weld C after PWHT.

obtained for the interlayer and the 27NiCrMoV base material respectively. Local maxima were again found in heat affected zones – the highest in HAZ of 27NiCrMoV base material (up to 350 HV near the face of weldment), lower in the HAZ of the interlayer (up to 300 HV) and about 320 HV in HAZ of COST F base material. This means that carbon enrichment increases hardness, yet the increase is not essential for macroscopic mechanical properties. The interlayer was again the weakest part of the weldment. Dissimilar weld joint D The last trial weld joint investigated, the weldment of COST F and COST FB2 steels, was produced in two variants with different weld metals – Thermanit MTS 3 (D1 weld joint) and Thermanit MTS 616 (D2 weld joint), which are both based on modified 9Cr steels. Thermanit MTS 3 contains 1.0 wt.% of molybdenum, while Thermanit 616 contains 0,4 wt.% of mo- lybdenum and 1.7 wt% of tungsten. Each variant underwent three different heat treatment schedules. The best one was selected on the basis of results of mechanical testing and a detailed microstructural study was performed using LM, SEM and TEM. Using that optimised technology a new weld joint was produced and long term creep testing was carried out. Results of creep testing together with fractographic analysis and TEM of carbon replicas and thin foils are summarised in another paper of this proceedings (Kasl J, Jandov á D:Testing of dissimilar weld joint of steels COST FB2 and COST F). Microstructures in all zones of theweldment corresponded to heavy tempered martensite with a high density of precipi- tates (Figures 13 to 15). The COST F base material unaffected by welding was significantly coarser than that of FB2. Small islands of δ -ferrite were sporadically observed in the base materials and more often in the weld metal, especially in the root. Density of precipitates decreased in the sequence from FB2 to COST F base materials and then to the weld metal. No differences between D1 and D2 variants were observable us- ing LMand SEM. The variant with the weldmetal of Thermanit 616 showed better mechanical properties; therefore this one underwent additional microstructural study and testing under creep conditions. Crossweld hardness profiles of all six variants of weld D weremeasured (twoweld joints after threemethods of PWHT). The hardness of the D2 weld after optimal heat treatment is shown in Figure 16. The hardness of both base materials unaffected by welding is approximately the same – about 240 HV 10 – and decreased in the HAZ, especially in FB2 steel to around 200 HV 10. Higher hardness (up to 280 HV 10) was found in the weld metal. This hardness profile corresponded to results fromtensile testing. All specimens ruptured in either the FB2 or COST F base materials.

Figure 15: LM micrograph of Weld D2 showing a martensitic structure in the weld metal.

Figure 16: Crossweld hardness profile of Weld D after optimal PWHT.

References 1 Kern TU, Mayer KH, Donth B, Zeiler G, Digianfrancesco A: The European Efforts in Development of New High Temperature Rotormaterials – COST536; Proceeding of the 9 th Liege Conference, Materials for Advanced Power Engineering, 2010. 2 J Lecomte-Beckers, Q Contrepois, T Beck, B Kuhn Eds: Forschungzcentrum J ü lich GmbH, Proceeding of the 9 th Liege Conference, Materials for Advanced Power Engineering, 2010. pp. 27-36. After qualification tests and finalisation of the weld joint designs, these procedures will be used in real production of combined rotors for turbines for fossil fuel power plants Acknowledgements This work was supported by Grant project TE01020118 from the Technology Agency of the Czech Republic. Conclusions Four trial weld joints were investigated. Microstructure was evaluated using light, scanning and transmission electron microscopy and thesewere comparedwith results ofmechani- cal testing. It then became possible to optimise consumables and PWHT as well as heat treatments of the base materials. Full conditions of the heat treatments of individual weld joints, however, cannot be published.

27

March 2015

AFRICAN FUSION