Experimental investigation of CMT discontinuous wire arc additive manufacturing of Inconel 625

Additive manufacturing (AM) is a progressive technology which holds promise for manufacturing of heat resistant super alloys. One of the most productive methods is wire arc additive manufacturing (WAAM). In this article, an alternative WAAM strategy is investigated. Experimental clads and material tests were performed to evaluate the material properties obtained through a cold metal transfer (CMT) discontinuous WAAM of Inconel 625 alloy. Using the modern terminology of Fronius Gmbh this method is called CMT cycle step. The difference is that it is automatically controlled by the welding source. CMT discontinuous WAAM has lower productivity and a higher consumption of shielding gas. However, it excels in low heat input and precise material cladding in comparison with a standard CMT continuous WAAM. It enables fabrication of finer details even on thin-walled components or in sections with problematic heat dissipation. Samples manufactured using this strategy were also compared with samples manufactured through a standard CMT continuous WAAM. Two sets of manufactured samples were thus tested. The following material tests were performed: (i) metallographic analysis, (ii) x-ray tomography, (iii) SEM analysis, (iv) hardness, (v) tensile strength (20 °C, 650 °C) and (vi) pin-on-disc (20 °C, 650 °C). The results show that the CMT discontinuous WAAM led to improved material properties in the Inconel 625 samples. Ultimate tensile strength improved by 15% at 20 °C and by 4% at 650 °C. Wear resistance at 650 °C was about two times higher. This paper concludes that the CMT discontinuous WAAM for Inconel 625 is definitely suitable for manufacturing of complex shapes, fine details and thin-walled components.


Introduction
Heat resistant super alloys are considered very important in the aerospace industry, mainly for the application of engine components. Approximately 50% of ''hot section'' gas turbine components are manufactured from nickel-based super alloys, as reported in [1,2]. Nickel-based super alloys are also logical choices in other key industrial sectors such as oil and gas industry components [3] and nuclear power plant components [4]. The main reason for using nickelbased super alloys is their exceptional material characteristics, such as mechanical strength, resistance to surface degradation and creep resistance at elevated temperatures. Many types of nickel-based super alloys have been developed over the years [5]. One commonly used nickel-based super alloy is Inconel 625. Because it contains niobium and molybdenum in a nickel chromium matrix, this super alloy is able to maintain its mechanical properties and corrosion resistance at temperatures up to 650 °C [6]. These excellent properties are retainable even without precipitation heat treatment. Components made from Inconel 625 are widely used in the aerospace industry in the ''hot section'' of gas 1 3 turbine engines and in nuclear reactors, the chemical industry and other high temperature and corrosion applications [7][8][9][10].
The most common manufacturing method for final products is machining. Ingeneral, nickel-based alloys are considered to be difficult-to-cut materials,mainly due to their low thermal conductivity, high hardness and low elasticmodulus [1]. This results in a high thermo-mechanical load of the component and the cuttingtool. Subsequently, progressive tool wear is observed in machining ofnickel-based alloys [11,12]. In response, new part manufacturing approaches are being explored with the aim of minimizing material removal through the cutting process, which can save both input material and tool costs. Considering current industrial and economical demands, additive manufacturing (AM) of Inconel 625 complex shaped components is one reasonable option [13,14].
Wire arc additive manufacturing (WAAM) is a subtype of AM that combines wire as a feedstock and electric arc as a heat source [16,17]. WAAM excels over other metal AM technologies in particular owing to its high deposition rates and low equipment costs [18]. On the other hand, it cannot fully match the geometrical complexity and precision of powder bed AM methods (e.g. selective laser melting or laser metal deposition) [15]. Hence, WAAM is suitable especially for manufacturing or repairing larger components which would typically be machined after an additive process [19][20][21]. The WAAM of nickel-based super alloys is an important research approach which is not satisfactorily investigated yet.
WAAM techniques can be further categorized by the type of arc weldingprocess. According to current studies [19,21,22], three basic techniques are being examined and used for research and development of prototypes. These welding techniques are gas tungsten arc welding (GTAW) [23][24][25], gas metal arc welding (GMAW) [26][27][28] and plasma arc welding (PAW) [22,29]. The advantage of the GMAW process lies in the fact that an electrode is a consumable wire which is standardly coaxial with the welding torch. In contrast, GTAW has the consumable wire added non-coaxially, rendering it more difficult for 5-axis control. While PAW can achieve very high productivity, the heat input is also quite high [30]. It is not beneficial for smaller, more precise parts. Moreover, the PAW process lead and tilt angles of the welding torch could be harmful for the manufacturing result. Thus, the PAW process is not suitable for 5-axis WAAM technology. For the purposes of WAAM, GMAW modification, known as cold metal transfer (CMT), is suitable [31,32], and it could be an effective approach to solution of manufacturing larger nickel-based super alloy components.
CMT is a modified GMAW process based on short-circuiting transfer, where movement of the wire feedstock at high frequencies (up to 80 Hz) combined with the welding current and voltage control improves the properties of the welding process [34]. The key improvements are elimination of spatter and considerably lower heat input [33]. As an alternative to a standard CMT continuous strategy, a CMT welding unit is also capable of a CMT discontinuous strategy. The process investigated in this paper stops the torch movement and creates a small "puddle weld" [35] by executing a defined  number of CMT cycles [34] than the torch moves to another "spot". Shielding gas flow continues through whole process which means the shielding gas consumption is higher as it blows even between welding "spots". Used CMT discontinuous process creates small "puddle welds" in a frequency circa 1 Hz (do not be confused with frequency of CMT droplets which is up to 80 Hz [34]), and it is approximately two times slower than standard CMT continuous process. This means the CMT discontinuous WAAM strategy has lower productivity and higher consumption of shielding gas. However, it excels in terms of very low heat input and very precise material cladding in comparison with a standard CMT continuous WAAM strategy. Fine details can be manufactured even on thin-walled components or in sections with problematic heat dissipation [34].
In relation to the current state of the art, numerous papers focus on WAAM manufacturing of Inconel alloys. Most papers address Inconel 718 (IN718) alloy, and only [26,28,38,[42][43][44] focus on Inconel 625. WAAM components from Inconel 718 require post-process heat treatment [9], which is no obstacle for manufacturing blank parts. According to [9,36], Inconel 625 (IN625) does not require post-process heat treatment. This property makes IN625 more suitable for hybrid WAAM [37] of a finished component in one machine. Analysis of material properties is a common step in the process of developing a technology strategy. Analysis enables an understanding of the elementary principles of the manufacturing process. To develop a reliable manufacturing method, it is necessary to study the material properties of samples manufactured using various complementary operations and methods.
Study [42] focused on comparison of the material properties of a standard CMT continuous WAAM process and a CMT WAAM process with pulsed droplet transfer during the manufacturing of IN625. Both techniques demonstrated YS and UTS results that were about 20% better than a casted sample. The CMT WAAM with pulsed droplet transfer manufactured sample was even slightly better (5 to 15%) than the standard CMT continuous WAAM samples in YS, UTS and hardness. There are similarities between CMT WAAM process with pulsed droplet transfer and CMT discontinuous WAAM. In pulsed droplet transfer, the welding variables oscillate in a defined range. Similarly, in CMT discontinuous WAAM, the welding variables oscillate from zeros to a defined values. However, the frequency is different. Investigation of the material properties of an IN625 manufactured  using CMT discontinuous WAAM can further confirm or contradict the results from [42]. For the CMT WAAM process of IN625, a special gas mixture designed for nickel based super alloys was chosen as a shielding gas. This gas mixture is recommended by the Messer Group GmbH [41]. Furthermore, the studies [39,40] are mentioning the importance of having a helium proportion of about 20-25% in the shielding gas mixture.
Standard analytical tools in contemporary research include optical microstructure and macrostructure analysis, hardness analysis and SEM analysis. Many studies also rely on tensile strength testing and EDS analysis. To keep the present research relevant to the current state of the art, all of these standard types of analyses were performed. It is clear that the properties of Inconel 625 alloy, which should retain its tensile strength at elevated temperatures, make a tensile strength test at 650 °C [6] very relevant for this research.
The main goal of this paper is to expand the state of the art in WAAM of Inconel 625. CMT discontinuous WAAM strategy which is known technology [35], but not yet used in WAAM of Inconel 625 according to current state of art.
CMT discontinuous WAAM will be tested and compared with a standard CMT WAAM. Complex material analyses will be performed, including metallography, SEM (scanning electron microscope), x-ray tomography, hardness, wear resistance and tensile strength. The last two material properties will be investigated at room temperature (20 °C) and also at an elevated temperature of 650 °C [45].
2 Materials, experimental procedure and methods

Materials and shielding gas
To implement the experiment, it was necessary to define and obtain welding and substrate material along with shielding gas. The essential materials used in the experiments are specified in the tables below (Tables 1, 2 and 3). Thermanit welding wire Ø 1 mm from Voestalpine Böhler Welding GmbH was used as an input material. The chemical composition and elementary mechanical properties of the wire are shown in Table 1.
Steel plates were selected as a substrate material. It was found [40] that welding Inconel 625 on weldable carbon steel is suitable. Plates made of EN-S235 JRG 1 according to EN 10,025 were prepared for the following dimensions: steel plate 110 × 110 × 12 mm. The chemical composition and elementary mechanical properties of the plates are depicted in Table 2.
The shielding gas is a mixture designed by Messer Group GmbH for welding nickel-based alloys. The composition of the shielding gas mixture is provided in Table 3. The gas flow is controlled by the flow valve to a level of 13 or 15 l/min depending on the WAAM strategy. Although [38] presents testing of different gasses, the experiments implemented for the present paper rely on recommendations by Messer welding experts with industrial experience. The

Experimental procedure, strategy and samples
An innovative patented hybrid manufacturing method using WAAM was used for the metal deposition process. This method utilizes a welding unit as a part of a 3-axis vertical CNC milling centre. The welding torch is in the same workspace as the milling spindle. The hybrid manufacturing method is characterized by this sequence of operations [37]: (i) deposition of a metal layer using electric arc welding in a protective atmosphere, (ii) machining aimed at creating a geometrically defined surface for another weld deposit, (iii) mechanical surface cleaning and (iv) air cooling/cleaning. At the cost of slightly reduced productivity, WAAM on a defined machined surface greatly increases the manufactured part's precision and possible shape complexity. Inter-operational cleaning also reduces porosity and the probability of internal material defects [37]. The method was also used in a study for welding stainless steel [46] (Fig. 1).
The experiments were performed on a 3-axis CNC hybrid machine. A Bridgeport VMC 500 XP vertical machining centre with a Heidenhain iTNC 530 control system was combined with a Fronius CMT Advanced 4000R welding unit. A Fronius Robacta Drive CMT W pro welding torch was positioned parallel to the milling spindle. Other accessories required for hybrid WAAM technology (wire feeder, air cleaning/cooling, water cooled table) were incorporated into the machine setup (Fig. 2).
The experimental samples were manufactured using the welding parameters presented in Table 4. The layers were manufactured one after another in the manner shown in Fig. 3, where weld cladding starts and directions are described. The size of the shape is approximately 110 × 110 mm with a height of 50 mm from the 12-mm-thick substrate. For both of the welding strategies described in Tables 4 and 5, two identical samples were manufactured, resulting in a total of four samples. Wall structure detail of these samples is shown in the Fig. 4. The welding parameters were optimized during the preceding research through visual inspection only. The material properties were unknown, and therefore, performing material tests is an overall advancement in WAAM technology research at the CTU.
The specimens for the material tests were extracted from samples using an EDM (electrical discharge machining) process. The layout of all specimens is shown in Figs. 5 and 6. There are two types of tensile specimens: (i) horizontal and (ii) vertical.

Analytical methods
To analyse the material properties of the manufactured samples, the material tests shown in Table 6 were performed. These tests are divided into structural tests and mechanical tests: The microstructure of the specimens was revealed by electrolytic etching in an electrolyte consisting of a 10% aqueous solution of oxalic acid with a Struers Polipower DC source with the following setup: 2.5 V, 20 s. The microstructure was then observed using a Carl Zeiss Jena Neophot 32 light optical microscope.
A Jeol JSM-7600F SEM microscope equipped with an EDS detector and EBSD detector was used to make observations. The acceleration voltage was 15 kV. Images were taken with a LEI detector (secondary electrons) and LABE detector (back scattered electrons). The secondary electrons provide information about the structure and morphology of the observed sample. The back scattered electrons slightly reflect the chemical composition of the observed sample.
X-ray tomography was conducted using a Zeiss Metrotom 1500 system with a 225 kV and 2 K detector with a pixel size of 0.2 mm. Samples with a square crosssection of 3 × 3 mm were selected to obtain high-resolution images of the scanned point cloud. The CT system settings for the tested samples were acceleration voltage of 225 kV, with a current of 460 μA, integration time of 600 ms with activated image averaging from 3 pictures in each angular position for noise reduction and a 2 mm cooper filter to adjust the X-ray spectrum to eliminate beam hardening artefacts to obtain the best imaging. Magnification of 8.8 × led to a final voxel size of 0.0227 mm. The porosity analysis was evaluated using VG Max software with an EasyPore algorithm; all pores bigger than 8 voxels were considered in the statistics, which is a common rule to  (iv) Hardness (specimens 29A, 29B).
The Vickers hardness (load 10 kg, HV10) of the specimens was measured using a Struers Duramin 40-AC3 universal hardness testing machine.
Dry sliding tribological tests were carried out on the specimens (ground with water-cooled 800 grit SiC paper) at 20 °C and 650 °C using an Anton Paar THT-S-CE 0000 pin-on-disc tribometer. The static friction partner was an Al 2 O 3 ball with a diameter of 6 mm. The test parameters were: wear track radius 3 mm, normal force 5 N and linear speed of the revolving partner 5 cm/s with a stop condition of 5000 laps. The relative humidity in the laboratory was 37%. The results of the pin-on-disc testing (i.e. profiles of wear track on both the ball and the revolving friction partner) were evaluated using an Olympus DSX1000 light optical microscope.
An Instron 3369 machine was used for the measurements. The crosshead speed was 2.4 mm/min. All the tests were performed according to the EN ISO 6892-1:2009 standard. Five measured specimens were used for each set.

Analysis of metallographic specimen
Micrographs of the etched 33A, 33B, 34A and 34B specimens obtained by LOM showed the presence of the layers characteristic for the additive manufacturing technologies as well as molten pool boundaries (MPBs). The microstructure is coarser closer to the boundaries of these layers, where it can act as a crack nucleation site. The observed areas are significant with the dendritic structure in the same direction. However, the microstructure is not cellular as in the case of 3D printing technologies, but is more reminiscent of the welded/clad microstructure. This can be seen in Fig. 7. The boundaries of the dendrites are rich with dark particles (Fig. 8); depending on the chemical composition of the filler wire, these may be Nb-, Mo-or Ti-rich carbides. Further investigation was conducted using SEM.
In the CMT discontinuous WAAM samples (B), there are areas with polyhedric grains, which imply partial recrystallization (see Fig. 8). This leads to the conclusion that, in the B setup, the energy input was insufficient for partial recrystallization. The structure of the CMT discontinuous WAAM samples (B) is more inhomogeneous but also more finegrained than the structure of the CMT continuous WAAM samples (A). A more fine-grained structure could indicate better mechanical properties. These mechanical properties will be further investigated.
There was no observable structural difference between specimens 33A and 34A or between specimens 33B and 34B. Thus, no structural difference was observed along the layer building direction for either of the WAAM strategies.

Analysis of morphology and element composition
Scanning electron microscopy (SEM) did not show any notable differences between WAAM strategies A and B or any notable differences along the layer building direction. The scanning electron microscopy observations were mainly focused on documenting the particles located at the recrystallized grain boundaries and dendrite boundaries.
The chemical composition analysis of the particles was performed via EDXS analysis. The two micrographs shown below (Figs. 9 and 10) document the microstructure of partially recrystallized grains and dendritic structure in specimen 33B. These two microstructural states are regularly repeated with respect of the clad 3D printed layers and their thermal influence on each other. The micrographs below document the morphology and chemical composition of the clad layers (Fig. 11). Oriented dendritic grains, which grow from partially recrystallized grains with particles at their grain boundaries, are visible. This type of microstructure corresponds with layer cladding and heat involving the previous layer.
Chemical composition map measurements (Fig. 12) confirmed the homogeneity of the 3D printed samples, with segregation of niobium and precipitation of niobiumrich phases. The chemical maps at lower magnification are listed below.
Chemical composition measurements were taken at the same measurement spot. Spectrum 4 represents the global chemical composition, and spectrum 2 and 3 correspond with the niobium base particles. The results of the EDXS measurements are shown in Fig. 13.
The global chemical composition corresponds with the filler metal used for 3D printing. The measured chemical composition in spectrum 2 and 3 confirms the higher niobium content, but the particles are small. The surrounding base material was involved in the measurements and affected the results.
EDXS measurements of the chemical composition for the precipitated particles were taken using higher magnification. The chemical composition maps are summarized in Fig. 14.
The measurement of the chemical composition of the niobium part is shown in Fig. 15. The chemical composition led to the result that the particle is niobium carbide with a niobium content of over 24 wt %.

Defect volume and pores inspection
The basic statistics and visualizations generated for all samples and results from the porosity analysis are summed up in Fig. 16. Figure 16 indicates that there is a significant difference between the pore count and size on samples made by CMT continuous WAAM and those made CMT discontinuous WAAM. The largest pore detected among all of the samples was recorded on sample 31B (horizontal sample made by spot welding), which had a volume of 7.3 × 10 −3 mm3, which is 619 voxels in total (Fig. 16). The smallest pore size and the only pore detected in the entirety of specimen 32A had a pore volume of 9.0 × 10 −5 mm 3 in a total of 8 voxels. This is the smallest pore size that may be practically detected by the given CT system and parameter settings for the specific material and sample size.
Porosity was evaluated as a defect volume ratio; in other words, the ratio between total material volume and total defect volume, both calculated in mm 3 . The porosity analysis results are described in Table 7.  Both of the CMT continuous WAAM samples had significantly lower porosity than the CMT discontinuous WAAM samples. The largest pore detected on a CMT continuous WAAM sample was detected on sample 31A with a volume of 1.3 × 10 −4 mm 3 and 11 voxels. The largest and the only pore detected on sample 32A was a pore with a volume of 9.0 × 10 −5 mm 3 in a total of 8 voxels. The largest pore detected from all of the samples was recorded on sample 31B (horizontal sample made by CMT discontinuous WAAM) with a volume of 7.3 × 10 −3 mm 3 which is 619 voxels in total. The largest pore detected on vertical sample 32B made by CMT discontinuous WAAM had a volume of 4.7 × 10 −4 mm 3 with 40 voxels in total.

Hardness of the machined surface
The hardness measurement was carried out on specimens 29A and 29B. The measurement layout along with the results is presented by the graph in Fig. 17. The average hardness of specimen 29A is 252 ± 14 HV10, and the average hardness of specimen 29B is 251 ± 13 HV10. Based on these results, there is no hardness variance between sets A and B. The scatter of the hardness values along the x-axis corresponds with the heterogeneous microstructure after the manufacturing process.

Wear resistance of the machined surface
The results of the pin-on-disc test were evaluated as the volume loss in cubic millimetres for both the pin and the disc. The wear profile of the pin and disc was measured using optical microscopy profiling in the Z-axis and then calculated by approximative equations given in the standard ASTM G99-5 [47]. The profile of the wear tracks and loss area were then calculated as the volume loss (see Fig. 18). The dynamic coefficient of friction was recorded during the test. The significant difference between the 20 and 650 °C tests can be seen (Fig. 19) in the dynamic friction coefficient and also in the volume loss of the disc. The ball wear was approximately the same in all four cases, while the volume loss at 20 °C and 650 °C on the side of the disc is two to three times higher than by the static friction partner.
The dry sliding dynamic friction coefficient (see records in Fig. 19) in the room temperature test fluctuated around an average of 0.75. Significant peaks up to 0.9 can be seen in the middle of the total time. This may be caused by an uneven microstructure with the particles causing a ploughing effect. In the case of the elevated temperature test at 650 °C, the friction coefficient was stable at 0.15-0.2.
This can lead to the conclusion that this alloy performs better at higher temperatures. Nevertheless, the wear track and wear mechanisms are different (oxidation and adhesion at elevated temperature) and must be considered in future product design. The cross-section of the wear track and its   Fig. 20. The wear track after the high temperature test was shallow with a stacked volume in the middle.

Tensile strength
An example set of ruptured specimen is shown on the Fig. 21. The results of tensile tests at a temperature of 20 °C are shown in Table 8 and Fig. 22. The standard deviations of YS and UTS are small, which corresponds to the excellent homogenous structure of the samples. The standard deviation of elongation values is quite noticeable, which is probably caused by the different position of rupture on the specimen during the tensile test.
For both the CMT continuous WAAM and CMT discontinuous WAAM strategies, the UTS in the horizontal direction is slightly higher than in the vertical direction (2% continuous, 8% discontinuous). This phenomenon does not apply to YS. Overall, the CMT discontinuous WAAM strategy specimen has a YS that is 13% higher in the horizontal direction and 1% lower in the vertical direction than the CMT continuous WAAM strategy specimen. Similarly, the UTS of the CMT discontinuous WAAM strategy specimen is 20% higher in the horizontal direction and 13% higher in the vertical direction than of the CMT continuous WAAM strategy specimen. Table 9 and Fig. 23 show the results of tensile tests of WAAM samples of Inconel 625 at a temperature of 650 °C. Deviations for YS and UTS are similarly small, which corresponds to the excellent homogenous structure of the samples. Deviations for elongation values are also quite noticeable and are caused by the different position of rupture on the specimen during the tensile test.
Even at 650 °C, the WAAM Inconel 625 retains most of its mechanical properties. However, in comparison with an annealed bar of Inconel 625 [6], the decrease of YS and UTS is more significant. While annealed Inconel 625 at 650 °C has approximately 15% YS and lower UTS compared to its 20 °C values, the CMT discontinuous WAAM specimens show a decrease of approx. 26%, and the CMT continuous WAAM specimens show a decrease of approx. 30% compared to the 20 °C values. The comparison is described in Tables 10 and 11.
The last comparison pertains to the published tensile strength results of WAAM Inconel 625; see Table 12. Both the CMT continuous WAAM and CMT discontinuous WAAM specimens have the highest YS of the published results. The UTS of the CMT discontinuous WAAM specimen is also the highest among the published results, while the UTS of the CMT continuous WAAM specimen is slightly below the average.

Conclusions
The aim of this study was to analyse the material properties of an innovative CMT discontinuous WAAM strategy used for Inconel 625 material. The CMT discontinuous WAAM strategy is suitable for building complex shapes, fine details and thin-walled components, mainly due to the lower heat input. The results of this innovative WAAM strategy were compared to the standard CMT continuous WAAM strategy performed by the authors of the study and also to published results pertaining to WAAM of Inconel 625. The material analysis demonstrated that the CMT discontinuous WAAM strategy has the following benefits and differences.
1. Metallography showed that the structure of the samples is quite similar to the welded/clad structure. This was expected given the character of WAAM technology. The CMT discontinuous WAAM strategy samples were affected by partial grain recrystallization, and their structure is more fine-grained. Thus, the microstructure of the CMT discontinuous WAAM strategy samples is more inhomogeneous than the CMT continuous WAAM strategy samples.
2. SEM was used to analyse the details of the microstructure. No element composition difference was observed between the CMT continuous (A) and CMT discontinuous (B) WAAM strategies. The element composition corresponds to the Thermanit 625 filler wire that was used. The presence of precipitated niobium particles was confirmed. These particles were quite small (2-5 µm). The chemical composition led to the result that the particle is niobium carbide with a niobium content of over 24 wt %.
3. X-ray tomography proved that there were no significant internal defects in the manufactured structure of either the CMT continuous WAAM strategy or the CMT discontinuous WAAM strategy samples. It is reasonable to say that the pores detected on the CMT continuous WAAM strategy samples were fewer in total and much smaller in size than the pores on the CMT discontinuous WAAM strategy samples. On the other hand, all defect volume ratios are on a very small scale compared to the total volume of material. Even the largest porosity of 0.00584% is a very good result compared to, for example, PBF-LB technology, where the lowest porosity detected with the same technology was in the range of 0.02%, where other researchers achieve the lowest levels of porosity in the range of 0.8% with PBF using metallography analysis [48].
5. Wear resistance measurements lead to the conclusion that this alloy performs better at higher temperatures. Nevertheless, the wear track and wear mechanisms are different (oxidation and adhesion at elevated temperature) and must be considered in future product design. The CMT discontinuous WAAM strategy samples performed even better at an elevated temperature of 650 °C, as the volume of its wear was about half of that of the CMT continuous WAAM strategy samples.
6. Tensile strength test results clearly show that the tensile mechanical properties of the CMT discontinuous WAAM strategy specimens are better than those of the CMT continuous WAAM strategy specimens. The YS is similar for both strategies, but the UTS of the CMT discontinuous WAAM specimens is approx. 15% higher at 20 °C, and the retention of the mechanical properties (YS and UTS) at a temperature of 650 °C is approx. 4% better. The CMT discontinuous WAAM strategy achieved the second highest UTS amongst the published results of Inconel 625 clads.