Materials Science & Engineering A 797 (2020) 140072
Available online 12 August 2020
0921-5093/© 2020 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Contents lists available atScienceDirect
Materials Science & Engineering A
journal homepage:www.elsevier.com/locate/mseaA comparison study of the dwell-fatigue behaviours of additive and
conventional IN718: The role of dislocation substructure on the cracking
behaviour
Dunyong Deng, Ru Lin Peng, Johan Moverare
∗Division of Engineering materials, Department of Management and Engineering, Linköping University, SE 58183, Linköping Sweden
A R T I C L E
I N F O
Keywords:IN718
Selective laser melting (SLM) Electron beam melting (EBM) Forged
Dwell Fatigue Creep
Environmentally assisted grain boundary attack Crack propagation
A B S T R A C T
The dwell-fatigue responses of high temperature materials, such as IN718, manufactured via additive manu-facturing processes with different microstructures is of practical interest in terms of time-dependent cracking resistance at elevated temperature. In the present study, the dwell-fatigue cracking behaviours of IN718 manufactured via selective laser melting (SLM) with different heat treatments, and via electron beam melting (EBM) with different scanning strategies were compared at 550◦C and with a long 2160 s dwell-holding
period. Comparison has also been made with a conventional forged counterpart. Detailed microstructure characterizations have been done to correlate the role of dislocation substructures on the dwell-fatigue damage mechanisms and cracking resistances. A mechanism regarding the susceptibility of the dislocation cell substructure in SLM materials to creep damage is proposed. In addition, the effects of other microstructure features on the dwell cracking resistance are also discussed.
1. Introduction
Nickel-base superalloys are of specific interest and importance for the high temperature applications in the aero and energy industries. However, Ni-base superalloy components are usually costly and dif-ficult to manufacture or machine via conventional routines. With a layer-by-layer fashion and offering great design flexibility, additive manufacturing (AM) is a very promising solution and has attracted significant attention in the past years. Among the commercial AM techniques, selective laser melting (SLM) and electron beam melting (EBM) are two of the most widely used for metallic materials. Some key differences regarding the processing between SLM and EBM include beam source, base plate/chamber temperature, atmosphere, powder size, etc. These lead to very different solidification conditions of the melt pools in these two processes, and therefore the microstructures and mechanical behaviours also differs.
Inconel 718 (IN718) is a Ni-base superalloy strengthened principally by 𝛾′′(Ni
3Nb). One of its main applications is as turbine disc material.
And due to IN718’s excellent weldability, SLM and EBM have been successfully applied to manufacturing IN718 with almost full density. The research on both SLM and EBM IN718 has been mostly focusing on the microstructure and tensile properties at both room temperature and elevated temperature [1–11]. These monotonic tensile tests show comparable strengths and ductilities of EBM and SLM IN718 to the
∗ Corresponding author.
E-mail address: johan.moverare@liu.se(J. Moverare).
conventional counterpart after appropriate heat treatments. However, fatigue and creep properties might be more relevant and important since, e.g. turbine engines have been designed to serve certain amount of cycles or hours, but so far the fatigue and creep data at elevated tem-perature are very limited for AM materials. The rough as-built EBM and SLM surfaces and internal defects can significantly the fatigue strength and statistics [12]. Aside from these defects, the intrinsic microstruc-ture and loading orientation (relative to the building direction) also play an important role in the fatigue and creep performance [13–17].
For the turbine disc material, dwell-fatigue crack propagation be-haviour is one of the most important considerations, since the dwell period holding at the maximum load at elevated temperature dur-ing each operation cycle might alert crack growth mode to be time-dependent intergranular and significantly accelerate crack growth rate, comparing to the pure cyclic one. Such a detrimental dwell effect has been well recognized in IN718 [18–27] and other Ni-base disc superal-loys. It is widely accepted that environmentally assisted grain bound-ary attack, i.e. dynamic embrittlement (DE) [28] and stress assisted grain boundary oxidation (SAGBO) [29], is responsible for the dwell effect. Grain boundary microstructure (precipitate [30,31] and misori-entation/character [32,33]) can significant influence the dwell-fatigue cracking resistance.
https://doi.org/10.1016/j.msea.2020.140072
Table 1
Nominal compositions of major elements of the powders used for EBM and SLM process in the present study.
Element(wt.%) Ni Cr Fe Nb Mo Co EBM Bal. 19.1 18.5 5.04 2.95 0.07 SLM 50∼55 17.0∼21.0 Bal. 4.75∼5.5 2.8∼3.3 <1.0 Element(wt.%) Ti Al Mn Si Cu EBM 0.91 0.58 0.05 0.13 0.1 SLM 0.65∼1.15 0.20∼0.80 <0.35 <0.35 <0.3
Therefore, the dwell-fatigue cracking resistance is very sensitive to the microstructure. Our previous studies [6,9] have showed the very different microstructures of EBM and SLM IN718 in both as-built and heat-treated conditions. And, even within one EBM build, the microstructure of the contour (surface) region is also largely different to that of the hatch (bulk) region. Recently, dwell-fatigue crack prop-agation tests have been performed under 2160s-dwell and 550◦C for
conventionally heat treated EBM (just the hatch part [34]) and SLM IN718 [35] at both parallel and perpendicular to building direction (BD) orientation in our previous studies, showing superior and inferior cracking resistance to the forged counterpart, respectively. This is attributed to the considerably different microstructures of the tested materials, and the resulted susceptibility to oxidation or creep damage. This inspires the present authors to extend the dwell-fatigue tests on a few more heat-treated AM microstructures, to get a more complete microstructural evidences for the dwell-fatigue cracking behaviours for the different AM microstructures. Specifically, a short-term homoge-nized and a long-term homogehomoge-nized SLM microstructures, a standard heat-treated EBM hatch and a standard heat-treated EBM hatch + contour microstructures, will be included and compared in this study. 2. Experimental
2.1. Material
For the nominal chemical compositions of the powders and general processing parameters for EBM and SLM processes, please refer to our previous work in [6,9]. The nominal compositions of major elements of the powders used for EBM and SLM processes for manufacturing samples are listed in Table 1. It is worth to note that in both EBM and SLM processes, contour and hatch scanning strategies are usually applied to surface and bulk regions, respectively. However, due to the very different settings, EBM contour and hatch microstructures are dramatically different [9], while the microstructure difference between these two regions in SLM samples is almost ignorable [36]. Detailed sample conditions, with regards to the heat treatment routines and the hatch/contour usages, are listed in Table 2. The purpose of applying the homogenization routines for SLM samples is to remove the segre-gation, and more importantly to remove dislocations/residual strains from the manufacturing process, which plays a significant role in the dwell fatigue cracking behaviours shown in the following. Applying the solution step after homogenization is to precipitate 𝛿 phase at grain boundary for better high temperature performance. And ageing step is to precipitate the strengthening phases 𝛾′and 𝛾′′for the peak strength.
And the idea of applying different heat treatment routines to EBM and SLM samples is to keep the characteristic features from the processes, which might give hints for applications of these processes.
2.2. Dwell-fatigue crack propagation tests
Compact tension (CT) samples (see Fig. 1) were made for the dwell-fatigue crack propagation tests. As can been seen inFig. 1, the notch is machined as parallel to the building direction, and the dwell intergranular crack developed during testing would continue on the
Fig. 1. CT specimen geometry showing that the notch is parallel to building direction and the resulted dwell cracking is on the notch/pre-crack plane.
notch plane. This ensures the mode slowromancapi@ fracture and the quantitative measurement of cracking propagation rate, as per ASTM E647. Note that, all the as-built EBM and SLM blocks are about 10 mm thick. The EBM Hatch sample was prepared by firstly ground the as-built EBM block to about 5 mm thickness to fully remove the contour region. All the CT specimen thicknesses are listed inTable 3.
In order to generate a sharp crack tip for the dwell-fatigue testing, all the as-machined CT specimens were first pre-cracked to about 1.5 mm in length from notch tip. Pre-cracking was conducted under pure fatigue condition with a load ratio of 𝑅 = 𝑃𝑚𝑖𝑛∕𝑃𝑚𝑎𝑥 = 0.05 and 10 Hz frequency at room temperature. Load applied during pre-cracking was lower than for the later dwell fatigue testing, as per ASTM E647.
The dwell-fatigue tests were performed at 550◦C in lab air with
a Zwick KAPPA 50 DS system. The waveform of each dwell fatigue cycle was trapezoidal, including 10 s ramping up to peak load, 2160 s of dwell (holding) at the peak load and 10 s of ramping down. The load ratio was 𝑅 = 𝑃𝑚𝑖𝑛∕𝑃𝑚𝑎𝑥= 0.05. Due to the different thicknesses, different load ranges were adapted to ensure a similar starting stress intensity factor (starting 𝛥K ∼ 20 MPa*m0.5) and noticeable crack
growth within reasonable experiment time. The detailed loadings for each sample are listed in Table 3. A direct current potential drop (DCPD) system with a pulsed 10 A current was used to monitor the real-time crack length during testing. When the tests were finished, fracture surfaces of all specimens were also inspected to correct the crack growth data measured from DCPD. For the fracture surface inspection, the pre-cracking and dwell-cracking can be easily identified under optical microscope, and then the actual crack growth can be measured. The crack length correction was performed as per ASTM E647. 2.3. Microscopy
A Leica M205C optical stereo microscope was used to inspect the fracture surface. And for the microstructure and crack path study, a Hitachi SU70 FEG scanning electron microscope (SEM) with EDS and EBSD systems was used. To study the dislocation substructure, spec-imens were also characterized with transmission electron microscopy
Fig. 2. Grain morphologies in the (a) contour and (b) hatch region after heat treatment of EBM samples. (c) and (d) showing the grain boundary precipitates in the contour and hatch region, respectively.
Table 2
Sample conditions and heat treatment routines applied in the present study.
Sample Homogenization Solution Ageing
EBM Hatch – 980◦C/1 h 720◦C/8 h 50◦C/h cool to 620◦C 620◦C/8 h Hatch+contour – SLM HSA 1080 ◦C/1 h 48HSA 1080◦C/48 h
*Samples were cooled down in the lab air after homogenization and solution steps.
Table 3
Summary of sample thicknesses and loads of dwell-fatigue crack propagation tests at 550◦C, load ratio 𝑅 = 0.05, 2160 s dwell period.
Sample condition Thickness 𝛥𝑃 /N 𝑃𝑚𝑎𝑥/N 𝑃𝑚𝑖𝑛/N
EBM SA Hatch ∼5mm 1600 1684 84
EBM SA Hatch+Contour
∼10mm 3000 3157 157
SLM HSA SLM 48HSA
(TEM) on a FEI Tecnai G2 system at 200 kV. The thin foil TEM speci-mens were electropolished in 90 vol.% ethanol + 10 vol.% perchloric acid at −20◦C at 25 V. Note that, all the TEM specimens were not
aged to exclude the coherent strain contrast from 𝛾′ and 𝛾′′, to gain
better contrast from dislocations. All the TEM specimens were before mechanical tests and were in the as heat-treated condition.
3. Results
3.1. Microstructure comparison
The microstructures of EBM contour and hatch regions are shown inFig. 2a and b, respectively. Generally, the contour grains are a mixed of equiaxed and columnar, while the hatch grains are mostly columnar and parallel to the building direction. Grain boundary precipitates are present in both contour and hatch grains, while it seems there are more clean boundaries (very few or without 𝛿 and NbC) in the hatch region (seeFig. 2d) than in the contour region.
The SLM samples are subjected to two different homogenization du-rations at 1080◦C and show correspondingly different microstructures.
The 1 h homogenization (HSA) condition (Fig. 3a) gives macroscop-ically identical grain morphologies to the as-built condition, and the sub-grain structure (Fig. 3c) inherited from the manufacturing process also remains. The HSA microstructure is completely un-recrystallized. Differently, intensive grain boundary migrations and annealing twin-nings happen during the 48 h homogenization (48HSA) (Fig. 3b), but some as-built structure with the sub-grain feature can still be found (see the dashed circle inFig. 3b). The un-recrystallized part with the as-built sub-grain feature is 9.3% by EBSD analysis. There are comparatively more grain boundary 𝛿 precipitates in the HSA condition (Fig. 3c) than in the 48HSA condition (Fig. 3d).
3.2. Crack path and fracture morphology
The fracture surfaces of EBM SA Hatch+Contour and Hatch samples are shown in Fig. 4a and b, respectively. Intergranular cracking is evident, and particularly the facets of columnar grains elongated along the building direction in the hatch region is obvious on both fracture surfaces. The over all crack front is considerably jagged. InFig. 4a it should be noted that the difference between the crack portions developed in the contour and hatch regions: crack propagation rate seems faster in the contour region than in the hatch region, even though contour (surface) region is under plane stress and is supposed to have slower crack propagation rate than that under plane strain [26].
Figs. 5a and b show the fracture surfaces of SLM HSA and 48HSA samples, respectively. The magnified views of fracture surfaces are
Fig. 3. Grain morphologies in the SLM (a) HSA and (b) 48HSA samples. (c) and (d) showing the grain boundary precipitates and the sub-grain structures in the HSA and 48HSA conditions, respectively.
Fig. 4. Fracture surfaces of (a) EBM SA Hatch+Contour and (b) EBM SA Hatch samples. The hatch and contour regions are marked separately in (a). Source:(b) is adapted from [34].
included inFig. 5c and d for SLM HSA and 48HSA samples, showing the identical grain morphology features as in that inFig. 3, and indicating that intergranular cracking is dominant.
3.3. Dwell crack growth rate
The crack growth rate 𝑑𝑎∕𝑑𝑁 is plotted versus the stress intensity factor range 𝛥𝐾 inFig. 6, and a forged counterpart adapted from [25] is also included inFig. 6for comparison. It clearly shows that the dwell crack growth rate is very sensitive to the microstructure. Generally, crack propagation rates of SLM specimens are faster than that of EBM specimens. SLM HSA condition has the fastest crack growth rate, and by applying the 48 h of homogenization step the crack growth rate is
reduced by a few times and is similar to the forged counterpart. For the EBM specimens, with the contour region the crack growth rate is significantly accelerated, compared to only the hatch condition. This is consistent with the fracture surface inspection of longer dwell crack propagation in the contour region than in the hatch region as shown in Fig. 4a.
4. Discussion
4.1. Cracking behaviours and damage mechanisms
The two possible time-dependent cracking mechanisms, namely en-vironmentally assisted grain boundary attack and creep, are operative
Fig. 5. Fracture surfaces of (a) SLM HSA and (b) SLM 48HSA samples. (c) and (d) are the magnified SEM images showing intergranular fractures of HSA and 48 HSA conditions, respectively.
under the present test condition for IN718. Examining the crack paths, specifically the secondary cracks, is important to distinguish the oper-ative mechanism. The environmentally assisted grain boundary attack mechanism emphasizes the supply of oxygen from air, and therefore the secondary cracks, if present, have to be originated from the main crack. Differently, creep cracking is a thermal activation process, and oxygen is not necessary even though it has been reported that oxygen can accelerate creep strain rate [37,38]. Therefore, the secondary cracks resulting from creep damage usually do not connect to the main crack. The crack paths of forged, EBM and SLM materials are shown in Figs. 7,8and9, respectively. By the aforementioned guide, the envi-ronmental attack is identified as the predominant damage mechanism for the forged counterpart inFig. 7. And such a environmental attack mechanism is well agreed for conventional IN718 at the temperature of 550◦C [22,23,28,32,39,40]. In addition, 550◦C is generally too low to
activate creep damage for Ni-base superalloys.
For the EBM contour crack shown inFig. 8a and b, the secondary cracks are mostly interconnected with the main crack, which is sim-ilar to the forged case and suggests the environmental attack as the damage mechanism. For the EBM hatch case, the environmental attack mechanism is assumed, since (a) no secondary cracks found in the examined cross sections and (b) the overall cracking propagation rate is considerably low as shown in Fig. 6. Note that, all the columnar grain boundaries in the hatch region align perpendicular to the loading direction and bear the maximum load, and if creep damage is operative these grain boundaries would have been prone to cracked and have produced noticeable amount of secondary cracks. Further, coalescence of secondary cracks usually results in faster crack propagation rate, while the EBM hatch has the slowest crack propagation rate among all the tested samples. In fact, the EBM hatch microstructure is very similar to that produced from direction solidification (DS) process, and to certain extent it can be regarded as conventional microstructure. From these points of view, it is quite convinced that environmental attack is operative for the EBM hatch case.
Different to the forged and EBM cases, creep damage is operative in the SLM HSA microstructure. From the secondary crack point of view, as shown inFig. 9a and b, the secondary cracks can be a few grains beneath the main crack, and with the grain boundary 𝛿 precipitates crack propagates via initiation of micro-voids at 𝛿 and subsequent linkage of micro-voids. For the details of extrapolation of creep damage, please refer to the previous work [35].
For the combination of un-recrystallized (with sub-grain structures) and recrystallized microstructures in the SLM 48HSA condition, it is most likely that environmental attack and creep mechanisms are op-erative simultaneously. The recrystallized part probably undergoes the
Fig. 6. Crack propagation rate 𝑑𝑎∕𝑑𝑁 as a function of stress intensity factor range 𝛥𝐾 for SLM HSA, 48HSA, EBM SA Hatch, EBM SA Hatch+Contour and forged counterpart specimens.
Source:The SLM HSA, EBM SA Hatch and forged data is adapted from [34,35] and [25], respectively.
environmental attack with a slower crack propagation rate, which can by derived from the comparison made on forged IN718 in [26]. How-ever, the overall cracking resistance of SLM 48HSA is just comparable to the forged counterpart, which means the un-recrystallized part most likely undergoes creep damage and has inferior cracking resistance. This suggestion is based on (1) the similarity of non-recrystallized substructure shown in SLM HSA and 48HSA conditions, (2) cracking the un-recrystallized grain boundary similarly by linkage of discontinuous micro-voids or wedge cracks (seeFig. 9d).
To summarize, it can be assumed that the creep damage happened in the SLM materials is probably associated to the sub-grain structure. However, one can also find that, as shown in Fig. 8, the sub-grain structures are also present in the EBM microstructure but do not result in creep damage. Without such a sub-grain structure in the forged microstructure, environmental attack is operative but with inferior cracking resistance to the EBM hatch columnar grains, which is with sub-grain structures. In the following, we focus on the substructures,
Fig. 7. (a) Cross section of the dwell crack of the forged IN718 counterpart. (b) is the enlarged area indicated in (a).
Fig. 8. Cross sections of dwell crack of EBM (a) contour region and (c) hatch region. (b) is the enlarged secondary crack indicated in (a). (d) is the enlarged crack tip indicated in (c).
specifically the dislocation features, of forged, EBM hatch columnar grain and SLM HSA conditions to understand the their role on damage mechanism. The dislocation features of SLM 48HSA and EBM contour microstructures have also been examined, but for brevity will just be included in the supplemental materials.
4.2. The role of substructure on damage mechanism
The forged microstructure is shown in Fig. 10. There are consid-erable amount of dislocations randomly distributed within the matrix. Some of the dislocations tangle but no well-ordered dislocation walls or boundaries are visible. Dislocations are also present inside the twin grain. Similarly, considerable amount of dislocations are also present and randomly distributes in SLM HSA condition (seeFig. 11). However, sharp sub-boundaries are clearly visible within the grains, as indicated in Fig. 11a. Such sub-boundaries are essentially well-organized dislo-cation network (seeFig. 11b). Sub-boundaries are also present in the EBM hatch region (Fig. 12). The sub-boundary dislocations are on two {111}planes and arrange into tetragonal network, as imaged by two g vectors inFig. 12b and c. Generally, in the EBM hatch microstructure there is very few dislocations.
The dislocation features of EBM hatch, forged and SLM HSA mi-crostructures, which are representative of environmental attack and creep damage with low and high cracking rate respectively, are sum-marized and illustrated inFig. 13. It shows that:
• The dwell crack propagation rate increases with dislocation den-sity within the grains.
• The role of the dislocation network in the SLM HSA and EBM hatch microstructures is probably different, and is the critical factor for the susceptibility to creep or environmental damage. The dislocation network in the EBM hatch case accommodates the misorientation between the two adjacent sub-grains, and such a dislocation network boundary is more like an equilibrium low angle grain boundary; while the in the SLM HSA case the dislocation density might exceed that needs for accommodating the misorientation, and is associated more with the deformation process and is more appropriately described as a cell boundary. It is evident by the fact that the SLM dislocation network is more closely packed than the EBM one.
Particularly, since creep damage is associated with the cell bound-ary + tangled dislocations microstructure of the SLM HSA condition, it is of interest to compare such a dislocation configuration with crept or superplastic-deformed conventional IN718 reported in the literature and find the correlation. Such a dislocation cell substructure is only possible to develop during hot work (superplastic) above 900◦C [41,
42] for IN718, as a partially recovered state. Similar dislocation cell substructures have been reported for Alloy 617 at the late stage of creep-fatigue at 950◦C [43] and creep at 950◦C [44]. Lillo et al. [44] shows that such a dislocation cell substructure actually forms when the specimen is crept into tertiary creep regime, during but not before which regime grain boundary cavitation or cracking develops. Kuo et al. [15,16] has performed creep test on SLM IN718 with the typical and similar dislocation substructure at 650◦C, showing relatively high
minimum creep rate, short steady creep regime and rapid onset of tertiary creep regime comparing to a cast & wrought counterpart. Combining the results in [15,16,43,44], it is reasonable that, with such a dislocation substructure as in SLM HSA condition, the material can much easier be crept into the tertiary regime, or even the material itself is already in the tertiary regime, and accelerating the grain boundary cavitation and fracture in the following time-dependent deformation at elevated temperature is most likely.
From the crack paths shown inFig. 9and in our previous study [35], we suggest that the creep void at the grain boundary probably forms via firstly diffusion of vacancies and then accumulation or growth to voids. This suggestion is based on: (1) the rapid solidification of SLM process
inevitably brings in large amount of vacancies in non-equilibrium state and (2) vacancies are also created when dislocation climbs under exter-nal stress to knit the dislocation network [45] . Considerable dislocation climb is required to knit a close-packed dislocation network as shown for the SLM HSA microstructure, and therefore is possibly associated with excessive vacancies. On the other hand, vacancy can pin the dislo-cation, and pinning force increases with the content of vacancies [45], which means that with increasing the vacancies associated with dis-location networks, recover might be more difficult since disdis-locations are more strongly pinned by vacancies. This can be supported by the relatively slow recovery and recrystallization kinetics shown by the comparison between SLM HSA and 48HSA microstructures in Figs. 3 and11.
Under the assumption that there are excessive vacancies within the dislocation network in the cell boundaries, if the external stress is high enough at elevated temperature, dynamic recover or breaking down of the dislocation network in the cell boundary is possible (seeFig. 13), by either pushing the dislocations from the cell interior to interact with cell boundary or encouraging the dislocation climbs/interactions within the cell boundary. This process would probably free the vacancies that associated with the broken-down part of dislocation network. After that, these freed vacancies, particularly those near grain boundaries, would diffuse to random high angle grain boundaries for further void nucleation and growth. The degradation or dissolution of cell boundary might be evident inFig. 9, as close to the grain boundary the cell substructure is no longer visible. One may also question if dislocation interaction and vacancies diffusion are thermally favoured at 550◦C for
creep fracture in the present experiments. If we consider that (1) tensile yield strength for the SLM HSA specimen is even higher than the forged at 550◦C, (2) stress is largely concentrated at the crack tip and (3)
the considerable amount of dislocations are able to promote diffusion, 550◦C is possible for creep damage for this specific microstructure. This is a possible mechanism for the typical association of dislocation cell substructure to rapid creep void formation, but further studies are require to better understand this phenomena.
Without such a dislocation cell substructure in EBM hatch and forged microstructures, it is less likely to have creep damage, since the microstructures are in a less creep-deformed state and the domi-nated environmental attack damage would not allow enough time for developing creep damage.
The dislocation densities in the EBM hatch and forged microstruc-tures surely influence their environmental cracking resistances, prob-ably by influencing the diffusion and oxidation process. In [46] it is suggested that the enhanced diffusion of Cr, by dislocation pipe path from grain interior to grain boundary, to form protective Cr2O3oxide is a key factor to reduced the grain boundary embrittlement at the crack tip. However, as clearly shown inFigs. 7b and8d, the forged grain boundaries are enriched of Nb and 𝛿-Ni3Nb comparing to EBM Hatch
grain boundaries. And the Nb enrichment probably weights out the effect of dislocation on the oxidation process and results in the inferior environmental cracking resistance for the forged counterpart, since Nb is more reactive than Cr [47,48].
On the other hand, the effects of dislocation on EBM and forged counterparts can also be interpreted from with the point of crack tip blunting. With such a large amount of tangled or individual disloca-tions, the forged microstructure has been strengthened and has higher strength than the EBM hatch microstructure. Due to that, ahead the crack tip of the EBM hatch grains, plastic deformation or plastic zone is bigger than that of the forged grains (seeFig. 13). At a micro-scale, the crack tip of EBM hatch grains has probably been more blunted, which reduces the stress concentration and correspondingly the stress related oxygen diffusion process.
Fig. 10. (a) Dislocation substructure of forged IN718, showing the random and tangled dislocations within grains. (b) Diffraction pattern of the twinning area marked in (a).
Fig. 11. (a) A general picture showing both dislocations in the sub-grain interiors and cell/sub-grain boundaries of SLM HSA specimen. (b) Dislocations are well-arranged to form cell/sub-grain boundary dislocation network.
4.3. Other effects on crack propagation rate
It is believed that the dislocation substructure is the key factor responsible for the damage mechanisms as shown above. Though other microstructure features, such as grain size, grain boundary precipitates, chemical segregation, grain alignment and texture, can affect simulta-neously the crack propagation rate under the same damage mechanism context. They are probably not as determinant for the damage mecha-nism as the dislocation substructure. In addition, it is also difficult to distinguish the exact contribution from each aforementioned features to the overall crack propagation rate. For example, the tortuous crack path of the fine-grained forged counterpart theoretically may introduce more crack deflection and hence shielding effects [49], comparing to the straight intergranular path in the EBM Hatch sample. However, this positive effect seems to be counterbalanced by other microstructure features, and resulting to faster crack propagation rate in the forged material.
For the environmental and creep damage mechanisms simultane-ously operative in SLM 48HSA case, the overall crack propagation rate
cannot simply be treated as a linear addition of the recrystallized and un-recrystallized (∼9.3 vol.%) parts. Xu et al. [50] reported that both heat-treated un-recrystallized and recrystallized SLM IN738LC show intergranular tensile fractures even at room temperature. This indicates that recrystallization does not significantly affect the grain boundary ductility in the SLM cases. Therefore, the environmental cracking re-sistance from the recrystallized part is open to further studies, and is of interested to compare with conventional counterparts with similar microstructure.
5. Conclusion
The dwell-fatigue crack propagation behaviours of SLM HSA, SLM 48HSA, EBM Hatch and EBM Hatch+contour microstructures have been investigated and compared with a forged counterpart at 550◦C with
2610s dwell hold. From the detailed microstructure characterizations, the effects of dislocation substructure on the damage mechanisms were discussed. The main conclusions can be summarized as
Fig. 12. (a) The sub-boundary dislocation network in EBM hatch microstructure. (b) and (c) are imaged with two g vectors to render the dislocation segments in the dislocation network.
Fig. 13. Illustrations of damage mechanisms and crack propagation rates with the regards of dislocation structures in EBM Hatch, forged and SLM HSA grains. • EBM material generally has better dwell fatigue cracking
resis-tance than SLM material.
• The dwell damage mechanism is largely dependent on the dis-location substructure. The disdis-location cell substructure in the SLM microstructure is susceptible to creep damage, while without the dislocation cell substructure forged and EBM materials are susceptible to environmental attack damage.
• The possible reason for creep susceptibility of dislocation cell substructure is that the dislocation configuration is correlated to the tertiary creep regime, even though the starting material has not been physically crept after the manufacturing process. • The environmental cracking resistance decreases with the
disloca-tion density that is not in the form of dislocadisloca-tion network, and the superiority of the less dislocation dense microstructure is related to plasticity-induced crack tip blunting.
• The recovery and recrystallization of the SLM dislocation cell sub-structure change the damage mode from creep to environmental attack and slow down the crack propagation rate, though the recovery and recrystallization kinetics are sluggish at 1080◦C.
CRediT authorship contribution statement
Dunyong Deng: Investigation, Formal analysis, Writing - original draft.Ru Lin Peng: Formal analysis, Writing - review & editing, Su-pervision.Johan Moverare: Conceptualization, Methodology, Formal analysis, Writing - review & editing, Supervision, Funding acquisition.
Declaration of competing interest
The authors declare that they have no known competing finan-cial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
This research is supported by Siemens AG, Germany and Sandvik, Sweden for providing test materials for this research. Faculty grant SFO-MAT-LiU#2009-00971 from Linköping University, Swedish Gov-ernmental Agency for Innovation Systems (Vinnova grant 2016-05175), Chinese Scholarship Council and Agora Materiae are also acknowledged for financial support.
Appendix A. Supplementary data
Supplementary material related to this article can be found online athttps://doi.org/10.1016/j.msea.2020.140072.
References
[1] Q. Jia, D. Gu, Selective laser melting additive manufacturing of Inconel 718 superalloy parts: Densification, microstructure and properties, J. Alloys Compd. 585 (2014) 713–721.
[2] Y. Lu, S. Wu, Y. Gan, T. Huang, C. Yang, L. Junjie, J. Lin, Study on the microstructure, mechanical property and residual stress of SLM Inconel-718 alloy manufactured by differing island scanning strategy, Opt. Laser Technol. 75 (2015) 197–206.
[3] D. Zhang, W. Niu, X. Cao, Z. Liu, Effect of standard heat treatment on the microstructure and mechanical properties of selective laser melting manufactured Inconel 718 superalloy, Mater. Sci. Eng. A 644 (2015) 32–40.
[4] E. Chlebus, K. Gruber, B. Kuźnicka, J. Kurzac, T. Kurzynowski, Effect of heat treatment on the microstructure and mechanical properties of Inconel 718 processed by selective laser melting, Mater. Sci. Eng. A 639 (2015) 647–655.
[5] T. Trosch, J. Strößner, R. Völkl, U. Glatzel, Microstructure and mechanical properties of selective laser melted Inconel 718 compared to forging and casting, Mater. Lett. 164 (2016) 428–431.
[6] D. Deng, R.L. Peng, H. Brodin, J. Moverare, Microstructure and mechanical properties of Inconel 718 produced by selective laser melting: Sample orientation dependence and effects of post heat treatments, Mater. Sci. Eng. A 713 (2018) 294–306.
[7] A. Strondl, M. Palm, J. Gnauk, G. Frommeyer, Microstructure and mechanical properties of nickel based superalloy IN718 produced by rapid prototyping with electron beam melting (EBM), Mater. Sci. Technol. 27 (5) (2011) 876–883.
[8] W.J. Sames, K.A. Unocic, R.R. Dehoff, T. Lolla, S.S. Babu, Thermal effects on microstructural heterogeneity of Inconel 718 materials fabricated by electron beam melting, J. Mater. Res. 29 (17) (2014) 1920–1930.
[9] D. Deng, J. Moverare, R.L. Peng, H. Söderberg, Microstructure and anisotropic mechanical properties of EBM manufactured Inconel 718 and effects of post heat treatments, Mater. Sci. Eng. A 693 (2017) 151–163.
[10] M. Kirka, K. Unocic, N. Raghavan, F. Medina, R. Dehoff, S. Babu, Microstruc-ture development in electron beam-melted Inconel 718 and associated tensile properties, Jom 68 (3) (2016) 1012–1020.
[11] M.M. Kirka, F. Medina, R. Dehoff, A. Okello, Mechanical behavior of post-processed Inconel 718 manufactured through the electron beam melting process, Mater. Sci. Eng. A 680 (2017) 338–346.
[12] A.R. Balachandramurthi, J. Moverare, N. Dixit, R. Pederson, Influence of defects and as-built surface roughness on fatigue properties of additively manufactured Alloy 718, Mater. Sci. Eng. A 735 (2018) 463–474.
[13] M.M. Kirka, D.A. Greeley, C. Hawkins, R.R. Dehoff, Effect of anisotropy and texture on the low cycle fatigue behavior of Inconel 718 processed via electron beam melting, Int. J. Fatigue 105 (2017) 235–243.
[14] Y.-L. Kuo, S. Horikawa, K. Kakehi, Effects of build direction and heat treatment on creep properties of Ni-base superalloy built up by additive manufacturing, Scr. Mater. 129 (2017) 74–78.
[15] Y.-L. Kuo, A. Kamigaichi, K. Kakehi, Characterization of Ni-based superalloy built by selective laser melting and electron beam melting, Metall. Mater. Trans. A 49 (9) (2018) 3831–3837.
[16] Y.-L. Kuo, T. Nagahari, K. Kakehi, The effect of post-processes on the mi-crostructure and creep properties of Alloy718 built up by selective laser melting, Materials 11 (6) (2018) 996.
[17] J. Xu, H. Gruber, D. Deng, R.L. Peng, J.J. Moverare, Short-term creep behavior of an additive manufactured non-weldable Nickel-base superalloy evaluated by slow strain rate testing, Acta Mater. 179 (2019) 142–157.
[18] J.P. Pedron, A. Pineau, The effect of microstructure and environment on the crack growth behaviour of Inconel 718 alloy at 650◦C under fatigue, creep and combined loading, Mater. Sci. Eng. 56 (2) (1982) 143–156.
[19] M. Bache, W. Evans, M. Hardy, The effects of environment and loading waveform on fatigue crack growth in Inconel 718, Int. J. Fatigue 21 (1999) S69–S77.
[20] F. Antunes, J. Ferreira, C. Branco, High temperature fatigue crack growth in Inconel 718, Mater. High Temp. 17 (4) (2000) 439–448.
[21] H. Andersson, C. Persson, T. Hansson, Crack growth in IN718 at high temperature, Int. J. Fatigue 23 (9) (2001) 817–827.
[22] D. Gustafsson, J. Moverare, S. Johansson, K. Simonsson, M. Hörnqvist, T. Månsson, S. Sjöström, Influence of high temperature hold times on the fatigue crack propagation in Inconel 718, Int. J. Fatigue 33 (11) (2011) 1461–1469.
[23] D. Gustafsson, E. Lundström, High temperature fatigue crack growth behaviour of Inconel 718 under hold time and overload conditions, Int. J. Fatigue 48 (2013) 178–186.
[24] D. Gustafsson, E. Lundström, K. Simonsson, Modelling of high temperature fatigue crack growth in Inconel 718 under hold time conditions, Int. J. Fatigue 52 (2013) 124–130.
[25] J. Saarimäki, J. Moverare, R. Eriksson, S. Johansson, Influence of overloads on dwell time fatigue crack growth in Inconel 718, Mater. Sci. Eng. A 612 (2014) 398–405.
[26] J. Saarimäki, M.H. Colliander, J.J. Moverare, Anisotropy effects during dwell-fatigue caused by 𝛿-phase orientation in forged Inconel 718, Mater. Sci. Eng. A 692 (2017) 174–181.
[27] G. Osinkolu, G. Onofrio, M. Marchionni, Fatigue crack growth in polycrystalline IN 718 superalloy, Mater. Sci. Eng. A 356 (1–2) (2003) 425–433.
[28] U. Krupp, C. McMahon Jr, Dynamic embrittlement–time-dependent brittle fracture, J. Alloys Compd. 378 (1–2) (2004) 79–84.
[29] R. Bricknell, D. Woodford, Grain boundary embrittlement of the iron-base superalloy IN903a, Metall. Trans. A 12 (9) (1981) 1673–1680.
[30] T.P. Gabb, J. Gayda, J. Telesman, P.T. Kantzos, Thermal and mechanical property characterization of the advanced disk alloy LSHR, 2005.
[31] H. Li, J. Sun, M. Hardy, H. Evans, S. Williams, T. Doel, P. Bowen, Effects of microstructure on high temperature dwell fatigue crack growth in a coarse grain PM nickel based superalloy, Acta Mater. 90 (2015) 355–369.
[32] U. Krupp, W.M. Kane, X. Liu, O. Dueber, C. Laird, C.J. McMahon Jr., The effect of grain-boundary-engineering-type processing on oxygen-induced cracking of IN718, Mater. Sci. Eng. A 349 (1–2) (2003) 213–217.
[33] U. Krupp, P.-G. Wagenhuber, W. Kane, C. McMahon, Improving resistance to dynamic embrittlement and intergranular oxidation of nickel based superalloys by grain boundary engineering type processing, Mater. Sci. Technol. 21 (11) (2005) 1247–1254.
[34] On the dwell-fatigue crack propagation behavior of a high strength superalloy manufactured by electron beam melting, Mater. Sci. Eng. A 760 (2019) 448–457.
[35] D. Deng, R. Eriksson, R.L. Peng, J. Moverare, On the dwell-fatigue crack propagation behavior of a high-strength Ni-base superalloy manufactured by selective laser melting, Metall. Mater. Trans. A (2019) 1–11.
[36] A.R. Balachandramurthi, J. Moverare, N. Dixit, D. Deng, R. Pederson, Microstruc-tural influence on fatigue crack propagation during high cycle fatigue testing of additively manufactured Alloy 718, Mater. Charact. 149 (2019) 82–94.
[37] S. Srinivas, M. Pandey, D. Taplin, Air-environment-creep interaction in a nickel base superalloy, Eng. Fail. Anal. 2 (3) (1995) 191–196.
[38] K. Maciejewski, J. Dahal, Y. Sun, H. Ghonem, Creep–environment interactions in dwell-fatigue crack growth of nickel based superalloys, Metall. Mater. Trans. A 45 (5) (2014) 2508–2521.
[39] D. Gustafsson, J. Moverare, S. Johansson, M. Hörnqvist, K. Simonsson, S. Sjöström, B. Sharifimajda, Fatigue crack growth behaviour of Inconel 718 with high temperature hold times, Procedia Eng. 2 (1) (2010) 1095–1104.
[40] J. Pfaendtner, C. McMahon Jr, Oxygen-induced intergranular cracking of a Ni-base alloy at elevated temperatures—an example of dynamic embrittlement, Acta Mater. 49 (16) (2001) 3369–3377.
[41] L. Huang, F. Qi, P. Hua, L. Yu, F. Liu, W. Sun, Z. Hu, Discontinuous dynamic recrystallization of Inconel 718 superalloy during the superplastic deformation, Metall. Mater. Trans. A 46 (9) (2015) 4276–4285.
[42] M. Weis, M. Mataya, S. Thompson, D. Matlock, The hot deformation behavior of an as-cast alloy 718 ingot, in: Alloy 718—Metallurgy and Applications, 1989, pp. 135–154.
[43] M. Carroll, L. Carroll, Developing dislocation subgrain structures and cyclic softening during high-temperature creep–fatigue of a nickel alloy, Metall. Mater. Trans. A 44 (8) (2013) 3592–3607.
[44] T.M. Lillo, R.N. Wright, Microstructural characterization of Alloy 617 crept into the tertiary regime, in: ASME 2015 Pressure Vessels and Piping Conference, American Society of Mechanical Engineers Digital Collection, 2015.
[45] V. Lindroos, H. Miekk-Oja, Knitting of dislocation networks by means of stress-induced climb in an aluminium-magnesium alloy, Phil. Mag.: J. Theor. Exp. Appl. Phys. 17 (145) (1968) 119–133.
[46] H. Ghonem, D. Zheng, Oxidation-assisted fatigue crack growth behavior in alloy 718-Part I. Quantitative modeling, Fatigue Fract. Eng. Mater. Struct. 14 (7) (1991) 749–760.
[47] C.F. Miller, G.W. Simmons, R.P. Wei, High temperature oxidation of Nb, NbC and Ni 3 Nb and oxygen enhanced crack growth, Scr. Mater. 42 (3) (2000) 227–232.
[48] C.F. Miller, G.W. Simmons, R.P. Wei, Evidence for internal oxidation during oxygen enhanced crack growth in P/M Ni-based superalloys, Scr. Mater. 48 (1) (2003) 103–108.
[49] R. Ritchie, Mechanisms of fatigue crack propagation in metals, ceramics and composites: role of crack tip shielding, Mater. Sci. Eng. A 103 (1) (1988) 15–28.
[50] J. Xu, H. Gruber, R. Boyd, S. Jiang, R.L. Peng, J.J. Moverare, On the strength-ening and embrittlement mechanisms of an additively manufactured Nickel-base superalloy, Materialia (2020) 100657.
