Processing-microstructure-property relationships in additively manufactured duplex stainless steels

Metal additive manufacturing (AM) has seen rapid expansion in the last decades. One family of alloys that are becoming more popular in AM applications are duplex stainless steels (DSS), known for their excellent mechanical and corrosion properties. These alloys are commonly used in the oil and gas, chemical, and marine industries. However, since AM can drastically change the properties and microstructure of materials compared to traditional methods, the processing-microstructure-property relationships of AM DSS materials are not fully understood. To address this knowledge gap, this work investigates the influence of processing parameters in laser powder directed energy deposition (LP-DED) and wire arc additively manufactured (WAAM) DSS alloys. A series of LP-DED samples at varying laser powers and scanning speeds were characterized to elucidate the complex influence processing parameters have on the microstructure and properties, specifically the defect content, phase fraction, and microindentation hardness. Various scanning strategies including raster, zigzag, spiral, and a serpentine pattern were investigated to understand their influence on the microstructure of AM DSS materials, specifically the grain size, grain shape, and texture. To boost the speed and efficiency of the characterization of processing-microstructure-property relationships for DSS alloys, thermomechanical simulations to simulate AM thermal histories using Joule heating were performed and the thermal histories were validated with a finite element analysis (FEA) model. The various thermal histories that were seen during these simulations lead to a wide variety of microstructures and properties with which preliminary microstructure-property relationships were obtained, showing that the phase boundary area density is an important feature that determines the mechanical properties. To expand upon the versatility of quick thermomechanical simulations, thermomechanical simulations were used as a ‘proxy’ dataset to expand on an important ‘reference’ dataset. A relatively large dataset from the thermomechanical simulations on cheap wrought materials was used to bolster a more expensive and relatively smaller dataset from a WAAM build. Using machine learning with both the ‘proxy’ and ‘reference’ dataset, interpretable microstructure-property relationships were attained, showing that 65% of the mechanical properties are determined by the intrinsic strength (internal friction stress, solid solution strengthening, and dislocation strengthening) while the rest of the mechanical properties are determined by the microstructure with 20% coming from the phase fraction and 15% coming from the microstructure refinement. To show a full breadth of experiments, a case study on the deposition of WAAM DSS is included. This study shows the prevalence of residual stress in large-scale AM and the importance of and process optimization while showing the potential for a low-nickel wire feedstock replacement for normal DSS wire feedstocks. This research shows the challenges that are inherent to large-scale AM and some of the methods executed to circumvent these challenges.

Name thesis:degree_name

Doctor of Philosophy

Level thesis:degree_level

dissertation

Discipline thesis:degree_discipline

Materials Science

Department dc:contributor.department

Department of Materials Science and Engineering

Grantor

Iowa State University

Year dc:date.issued

2025

Author dc:creator

• Johnson, Grant A

Advisors dc:contributor.advisor

• Collins, Peter C
• LeSar, Richard
• Roy, Sougata
• Johnson, Duane D
• Napolitano, Ralph E
• Narra, Sneha P

Language dc:language.iso

en

DOI

https://doi.org/10.31274/td-20260223-114

OAI identifier oai:identifier

oai:dr.lib.iastate.edu:20.500.12876/dvmqBX3v