Structural Integrity Assessment of 316L Stainless Steel Produced by Additive Manufacturing

  • Mehran Shahriarifar

Student thesis: Doctoral ThesisDoctor of Philosophy

Abstract

Additive manufacturing (AM) is an emerging technology that has recently gained large
interest due to its potential to produce customised components with complex geometries.
Laser powder bed fusion (L-PBF) is a mainstream AM process for fabricating 3D metallic
parts where the powder bed is melted selectively by a high energy density fibre laser. The
layer-wise production technique in L-PBF provides a high degree of freedom of design to
produce a wide variety of objects with complex geometries. The material investigated in this
study is austenitic stainless steel 316L with a wide variety of applications in different
industrial sectors, making it an attractive material for L-PBF technology. However, there is
currently a lack of data and understanding on the effect of different post-process heattreatments on the material performance under static and fatigue loading and the correlation
between mechanical properties with microstructure. Moreover, the effect of residual stress
and microstructure on fatigue crack growth rate (FCGR) is not fully understood. Both the
effect of build orientation and heat-treatment on residual stress distribution and subsequently
on fatigue crack growth need to be investigated. In addition, the effect of defects, anisotropy,
and post-process heat-treatments on fatigue strength and cracking mechanism requires
further insight on the role of defect on the anisotropic behaviour.
This research aims to provide a holistic experimental assessment of the structural integrity
performance of L-PBF 316L with various post-process heat-treatment conditions. The
primary objectives are: 1) to study the effect of different post-process heat-treatments such
as stress-relieving, annealing, and hot isostatic pressing (HIPing) on the microstructure and
mechanical performance of the material; 2) To address the anisotropic properties taking into
account vertical and horizontal samples; 3) to investigate the hardness and static properties
and make a correlation with microstructure; 4) to investigate the fatigue strength and evaluate
different damage mechanisms in terms of the initiation and propagation of the crack; 5) to
conduct fatigue crack propagation test to investigate the effect of residual stress as well as
microstructure on fatigue crack growth rate (FCGR).
Microstructure evaluation revealed that the material in the as-built condition has a fine grain
cellular structure. Such fine grain microstructure was not found after annealing and HIPing
where a more homogenised microstructure was present. Higher hardness was measured for
the as-built condition but a significant reduction after both annealing and HIPing was
correlated to the grain coarsening after post-processing. Tensile test results further confirmed
the microstructure and microhardness observations. The as-built samples had higher yield
strength, but lower ductility compared to the annealed and HIPed samples.
Fatigue strength tests were carried out on vertical and horizontal samples in the as-built and
HIPed conditions. The as-built horizontal samples exhibited higher fatigue strength than their
vertical counterparts. Fatigue crack initiation occurred from lack-of-fusion or rarely gas pores
near the surface in the as-built samples. The alignment of the lack-of-fusion killer defects
with respect to loading direction resulted in a larger cross-section area for the vertical
samples, hence reducing their fatigue strength compared with the horizontal samples with a
smaller elongated-shape cross-section area of the defect. Significant improvement in the fatigue strength with less anisotropy was observed after HIPing. This improvement was
attributed to the densification of the material after HIPing as well as the higher ductility of
the HIPed samples. Fractography of the HIPed samples revealed that fatigue crack was
initiated from microstructure or defects that were much smaller than the killer defects in the
as-built samples.
FCGR testing was conducted to evaluate the effect of residual stress and microstructure. For
the former factor, as-built samples were compared to the stress-relieved samples, where the
stress-relieving at 650°C had a negligible effect on the microstructure but changed the
residual stress distribution. Residual stress measurements showed compressive residual
stresses dominance in the mid-thickness but tensile residual near the surface in the as-built
condition for both vertical and horizontal samples. Stress-relieving reduced the residual
stresses but not fully eliminated them. Despite this, no significant difference was observed in
the crack growth rate of the two conditions. To study the microstructure effect, stressrelieved, annealed, and HIPed conditions were considered, where both annealing and HIPing
caused a change in the microstructure. The FCGR test results showed that the material
resistance to crack propagation was improved after annealing and HIPing. The ∆𝐾𝑡ℎ was
improved by 60% and 75% after annealing and HIPing respectively. Also, the crack growth
rate in the Paris law regime was found to be 20-50% lower compared to stress-relieved and
as-built samples.
The findings of this research showed that post-process heat-treatments can have significant
effects on the microstructure and material properties of L-PBF 316L. A detailed
microstructure and defect characteristic was performed to link to mechanical properties. Both
annealing and HIPing were found to improve static and fatigue properties, except the yield
strength which was higher in the as-built material due to the presence of fine-grain
microstructure and high dislocation. In terms of fatigue strength, in the as-built condition, the
alignment of lack-of-fusion defects to loading resulted in a larger cross-section area in the
vertical samples, hence lower fatigue strength. The almost fully-dense HIPed samples with a
high ductility showed higher fatigue strength compared to the as-built samples. In terms of
fatigue crack growth rate, the effect of residual stresses was found to be negligible for the
SENB configuration that was used in this study. Both annealing and HIPing, however, were
found to change the microstructure and improved the resistance of the material to crack
growth
Date of Award2023
Original languageEnglish
Awarding Institution
  • Coventry University
SupervisorXiang Zhang (Supervisor), Kashif Khan (Supervisor) & Matthew Doré (Supervisor)

Cite this

'