BAM Dissertationsreihe
Filtern
Dokumenttyp
- Dissertation (2)
Sprache
- Englisch (2) (entfernen)
Referierte Publikation
- nein (2)
Schlagworte
- Austenitic Stainless Steel (1)
- Austenitischer Rostfreier Stahl (1)
- Beugung (1)
- Data fusion (1)
- Diffraction (1)
- Duplex stainless steels (1)
- Eigenspannungen (1)
- Heat Treatment (1)
- Hydrogen assisted cracking (1)
- Powder Bed Fusion of Metals with Laser Beams (1)
- Pulverbettbasiertes Laserstrahlschmelzen (1)
- Residual Stress (1)
- Time-of-Flight secondary ion mass spectrometry (1)
- Wärmbehandlung (1)
Organisationseinheit der BAM
- 9 Komponentensicherheit (2) (entfernen)
173
Additive manufacturing processes offer extensive advantages for the design freedom of structures through layer-by-layer production. This enables high weight savings as well as the integration of functions such as cooling channels. This technology thus offers great potential to contribute to a sustainable future. The pioneer among these manufacturing processes is the powder bed fusion of metals with laser beams (PBF-LB/M). This process is characterised by high laser scanning speeds and highly localised heat input, which have a strong effect on the microstructure and thus also on the mechanical properties. For example, the austenitic steel 316L exhibits a cellular structure at the subgrain level. This microstructure feature leads to higher yield strengths and comparable ductility to conventionally processed 316L. In addition to the traditional applications of 316L steel in the petrochemical and nuclear industries, this enables new applications such as medical stents or bipolar plates for fuel cells with proton exchange membranes. However, the layer-by-layer production with high scanning speeds and localised heat input induces cooling rates in the order of 106 K.s-1. The large temperature gradients and the shrinkage restraints of each weld bead and layer lead to the development of complex residual stress fields. These reduce the material performance and can even lead to premature failure. Thus, the fatigue properties are severely affected by rapid crack growth or prematurely developing cracks. Furthermore, specimens may warp during PBF-LB/M or immediately when the components are separated from the build plate. Therefore, residual stress is one of the main disadvantages of PBF-LB/M, making it difficult for this technology to be more widely accepted in the industry. Based on the current state of the literature, the procedure for determining residual stress employing diffraction methods, the influence of the component geometry, as well as the inter-layertime (ILT) on residual stress and, lastly, suitable heat treatment strategies for relaxing residual stress in PBF-LB/M/316L, were identified as insufficiently researched areas.
Determining residual stress is a major challenge. X-ray and neutron diffraction are particularly suitable for filigree structures, which can preferably be produced using PBF-LB/M. Here, the microscopic strain of the lattice planes is used to calculate the macroscopic residual stress. These methods are nondestructive and allow the spatial resolution of the bi-axial and tri-axial residual stress. In the present work, in-situ neutron diffraction tensile tests were performed to analyse the micromechanical behaviour of PBF-LB/M/316L. The suitability of the lattice planes for calculating the macroscopic residual stress was investigated. The (311) lattice plane was found to be the best option for determining the macroscopic residual stress in PBF-LB/M/316L. Furthermore, it was shown that the Kröner model can be used to calculate the X-ray diffraction constants despite the texture. Currently, both aspects are common practices in the determination of residual stress. The results presented here support the validity of this approach and increase the confidence in the experimentally determined residual stress, which has a positive effect on the assessment of quality concerning the safety of a component manufactured by PBF-LB/M.
The geometry of a structure manufactured by PBF-LB/M determines the component stiffness and influences the thermal gradients during manufacture and ultimately the residual stress. The effect of smaller or larger dimensions (larger than 10 mm) on the residual stress is rarely considered. To investigate this aspect, representative test specimens with different thicknesses and lengths were produced. Hence, the influence of the geometry i.e., component stiffness on the residual stress was evaluated. The residual stress was determined using X-ray and neutron diffraction. The analysis of the residual stress showed that an increase in thickness leads to overall higher residual stress. In addition, it was shown that increasing the sample dimension leads to smaller residual stress gradients. Above a threshold value of a few millimetres, no significant change in the residual stress was observed.
The ILT is inherent in every PBF-LB/M construction job and influences the thermal gradients during production and thus the residual stress. A change in wall thickness in a geometrically complex structure or a variation in the number of specimens in the construction process leads directly to a change in the ILT. To simulate this, specimens with different ILT were produced. The residual stress was determined by X-ray and neutron diffraction. The use of a short ILT resulted in higher surface residual stress, but lower volume residual stress. Here, the surface residual stress and the residual stress in the volume showed contrary behaviour. This was attributed to the complex heat conduction during the process, as shown by the thermographic measurements.
To avoid distortion of the specimens or real components upon separation from the build plate or during post-processing steps, stress relief annealing is usually performed after the PBF-LB/M process. Based on standards for heat treatment of welded austenitic steels, heat treatments were performed at low (450 °C for four hours) and high (800 °C and 900 °C for one hour) temperatures. The results show that the heat treatment at 450 °C relaxed the residual stress by only 5 %. This low relaxation is due to the stability of the cell structures. The high-temperature heat treatment showed that 900 °C is required to dissolve the cell structure and achieve a relaxation of about 85 %. This result is in good agreement with the standards for stress relief annealing of welded austenitic steels.
160
For almost 150 years it is known that hydrogen has a deleterious effect on the mechanical properties of metallic components. Nowadays, the problem of hydrogen assisted degradation is highly relevant in energy related fields due to the massive use of steel as a structural component in these applications and its sensitivity to hydrogen. Since the discovery of hydrogen assisted cracking (HAC), researchers studied intensively and suggested possible explanations and mechanisms in order to define how hydrogen is affecting the material. In general, it is considered that hydrogen changes the mechanical properties more in terms of ductility (deformation capacities) than in strength (load capacities). Hydrogen concentration is one of three crucial factors in the degradation process, together with the microstructure of the material and the internal/external mechanical load. The relatively high concentration of hydrogen resulting in this loss of ductility can originate during production or before service (e.g. welding processes) and during service (i.e. catholically protected systems to eliminate corrosion processes in sour environments).
In parallel to the theoretical work, tremendous efforts were, and are still, invested in searching for a proper method to elucidate, map and quantify the hydrogen in the microstructure, which is the basis for this work. For steels, the focus is mainly on the observations of diffusion processes and the interaction of hydrogen with the microstructure
in regions with high local stresses/strains (for example around evolving cracks). The challenge for reaching this goal arises from the fact that accurate indication of hydrogen by means of position, unlike heavier atoms, can be made only by mass spectrometry or by interaction with another element (e.g. silver decoration, special coating and resonant nuclear reaction by nitrogen). In addition to this, the difficulty recording the hydrogen behavior while it rapidly diffuses through the material, leaving only the unpredicted failure, should be taken into account.
Although using powerful characterization methods, models and computational simulations, the key to defining the mechanisms behind HAC is still under debate and not fully understood. The relationship between material and hydrogen is determined by three factors, i.e., the material structure and microstructure – determining the physical properties, the mechanical load applied on the material and the hydrogen concentration. It is well known that in order to have a complete definition of HAC these three factors must be examined locally with the minimal scale and the maximal resolution reachable. The major gap is the lack in such a characterization method or a technique by which one has the ability to detect and observe the hydrogen in the metallic microstructure. The commonly used techniques nowadays are capable of characterization of the microstructure without the ability to observe the hydrogen distribution. Global hydrogen concentration and localized hydrogen observation are possible by some techniques which are incapable of indicating a change in the structure or microstructure therefore a comprehensive overview can be gained only by combining several methods.
In the presented research, secondary ion mass spectrometry (SIMS) was adopted as the main tool to detect and locally map the hydrogen distribution in two types of duplex stainless steel grades: EN 1.4462 (standard 2205 duplex stainless steel) and EN 1.4162 (2101 lean duplex stainless steel). The term duplex stainless steel (DSS) refers to the austenitic-ferritic microstructure of the steel where the combination of physical and mechanical properties of the two phases is achieved. The DSS was selected as a case study for this work due to the wide use of this grade in many energy and the lack of knowledge on hydrogen behavior in two-phase containing microstructures. ToFSIMS was exploited in-situ and ex-situ in three experimental approaches during or following
an electrochemical charging procedure. This type of hydrogen charging was selected as it simulated a procedure of cathodic protection of most sub-water oil and gas extraction and delivery systems. The experimental procedures were:
1. Ex-situ charging followed by ToF-SIMS imaging for basic understanding of hydrogen distribution.
2. Ex-situ charging followed by in-situ mechanical loading to obtain information on hydrogen behavior around a propagating crack.
3. In-situ permeation of hydrogen through a steel membrane inside the ToF-SIMS to obtain information on diffusion behavior of hydrogen in a two-phase microstructure.
The comprehensive view of the effect of hydrogen on steel was gained by using supplementary methods, such as high resolution scanning electron microscopy (HR-SEM), focused ion beam (FIB) and electron back-scattered diffraction (EBSD). The state of the art in this work lies in applying both: in-situ experimental approaches and data treatment of the ToF-SIMS raw data. The data treatment includes the combination of data from several sources (data fusion).
The results for the ex-situ charging followed by static sample imaging and data fusion showed that when the analyzed surface is directly exposed to the electrolyte the degradation is pronounced differently in the ferrite, austenite and interface. The degradation mechanisms in the ferrite and austenite were reflected by the formation of cracks on the surface of both, where a high concentration of hydrogen was obtained. This result supports the assumption that hydrogen is attracted to highly deformed regions. The advantage of using in-situ charging/permeation in comparison to ex-situ charging is that the effect of hydrogen on the ferrite and austenite phases when the hydrogen is evolving from within the microstructure is realized, in comparison to when the analyzed surface is initially exposed directly to the electrolyte. In both experiments the ferrite was observed as a fast diffusion path for the hydrogen. The faster diffusion of hydrogen through the ferrite is expected due to the higher diffusion coefficient, however, a direct proof for the diffusion sequence in this scale was never shown. Most significant results were achieved by the ‘core’ experiments of this research. These experiments included the design of a novel dynamic mechanical loading device to apply an external load during SIMS imaging of a hydrogen precharged-notched sample. For the first time it was shown that plastic deformation induced by applying a mechanical load is resulting in a redistribution of hydrogen locally around the notch.