1 Introduction

Wire arc additive manufacturing (WAAM) is a modern manufacturing technique, using the principles of fusion welding to deposit material. This process uses an electrical arc to melt a feedstock wire, depositing a component of the desired shape. The capability of this process to reduce lead times [1] and material waste [2] is advantageous for the production of high value components in industries such as defence, construction, maritime, energy and aerospace [3,4,5,6]. These components are often manufactured from materials such as titanium alloys [7], nickel-based superalloys [8, 9] and precipitation hardening (PH) stainless steels [10]. In addition, this process can reduce the environmental impact of manufacturing such components [11].

Alloy 15-5PH is a martensitic, PH stainless steel [12] widely used in the aerospace, marine and energy industries for its high strength and corrosion resistance [13]. Many industries are investing in the transition from conventional manufacturing, such as machining or forging, to AM (additive manufacturing) technologies including WAAM. To make this change viable, a set of process parameters must be identified to achieve bulk mechanical properties equal to or superior to those achieved by conventional manufacturing [14]. The key factor in the processing of PH stainless steels is the thermal history during manufacturing, particularly when high temperatures are experienced by the material such as during welding, casting, forging or AM techniques [15, 16]. Control of mechanical properties for PH stainless steels can also be achieved through heat treatment, where the material is typically supplied in a solution treated condition [13] and an aging treatment is performed to develop precipitates which increase strength [17].

The development of corrosion resistant alloys such as PH stainless steels as feedstocks for the WAAM process is not yet extensively investigated [18]. The microstructure and tensile strength have been previously investigated for thin plates joined using gas tungsten arc welding (GTAW) [19] which reports low elongation and reduced strength when the as-welded material is compared to equivalent wrought material. A related study has investigated the mechanical properties of PH stainless steel plates requiring multi-pass welding [20] showing that quasi-cleavage fracture is expected and leads to strength levels far below those specified by ASTM A693 [12]. Early research into the use of PH stainless steels for the WAAM process has been performed using cold metal transfer (CMT) deposition [10]. This variant of gas metal arc welding (GMAW) advances and retracts the filler wire to deposit small droplets of liquid metal many times a second. This results in a reduced weld heat input compared to conventional GMAW deposition and improves control over the weld bead profile [21].

The results of Caballero et al. [10] and Guo et al. [22] investigate the mechanical properties of 17-4PH produced using CMT-WAAM. They demonstrated low strength in the as-deposited condition due to the retention of interdendritic δ-ferrite. This is a body centred cubic phase [23] found in many stainless steel alloys in varying quantities [24] and must be controlled during welding [25]. In stainless steels, δ-ferrite evolves during the solidification of the alloy, followed by austenite and martensite through diffusion and shear transformation [26]. Due to the rapid solidification found during the WAAM process, little austenite is retained, leading to significant levels of δ-ferrite [27]. An increase in the δ-ferrite fraction of these alloys has been correlated with a reduction in toughness in stainless steels [28, 29].

The primary strengthening constituents of 15-5PH include niobium carbides (NbC), copper precipitates [22] and additional metal carbides of compositions M7C3 and M23C6. M23C6 carbides are formed during aging and are rich in Cr [15]. Changes in both the path strategy and weld heat input result in differences in bead geometry, microstructure [30] and thermal history [31] leading to improved mechanical properties. It is proposed herein that optimising the process parameters will improve the strength of this material in the as-deposited condition. This can be achieved by minimising the evolution of δ-ferrite and increasing the evolution of strengthening phases such as carbides and copper precipitates.

A previous study by Niu et al. [20] has developed impact toughness data at −20 °C for multi-pass GTAW welds following a range of post weld aging heat treatments (580 °C, 600 °C, 620 °C for 4 h). This shows an increase in impact toughness as the aging temperature is increased. The impact toughness of 15-5PH has not yet been studied for material produced by WAAM.

The strength of 15-5PH produced by WAAM, as reported to date [10, 22], is notably lower than that specified for wrought material. However, by controlling the δ-ferrite evolution, the present work demonstrates an improvement in strength for this material and process combination. In addition, data is presented on the resulting impact toughness, which is valuable for the consideration of this material and process in future industrial applications.

2 Materials and methods

The bulk material produced for this work was a wall of deposited material of dimensions 235 mm × 30 mm × 110 mm manufactured using an additive manufacturing cell. This cell was composed of an ABB 2400 L robot arm controlled by an ABB IRC5 control cabinet and teach-pendant in addition to a Fronius CMT fusion welding torch, power supply and wire feed.

The shielding gas was high-purity argon and the substrate was EN32B low carbon mild steel. Commercially available 15-5PH welding wire of 1.2 mm diameter was used as a feedstock with the chemical composition of the feedstocks which is shown in Table 1 as provided in the suppliers datasheets.

Table 1 Chemical composition of material feedstocks
Full size table

Two sets of process parameters were tested to vary the weld heat input. These values are shown in Table 2 and subsequently referred to as HH and LH. The resulting parameters were obtained from settings within the robot and welding control systems and the interpass temperature measured with a digital thermometer probe. Weld heat input was calculated using Eq. 1 from BE EN 1011-1 [32]. In this equation, \(Q\) is the weld heat input, \(k\) is the thermal efficiency (0.8 for MIG welding), \(V\) and \(I\) are the voltage and current of the arc and \(v\) is the arc travel speed.

Table 2 Process parameters for WAAM production
Full size table
$$Q=k\frac{V\cdot I}v\cdot10^{-3}\;\mathrm{in}\;\mathrm{kJ}/\mathrm{mm}$$
(1)

Samples were extracted from the bulk wall and prepared by standard metallographic grinding and polishing techniques followed by etching with Kalling’s no. 2 reagent. Examination of the microstructure was performed using an Olympus GX51 inverted microscope. Microhardness measurements were taken on these same samples using a Q-Ness 60 A + automated hardness tester with a 0.05 kg load.

The point count method from ASTM E562-19 [33] was used to quantify the δ-ferrite volume fraction with 3 fields of 100 points on each of the HH and LH material and compared to the microstructural images from Guo et al. [22].

Tensile samples and Charpy impact samples were produced from the bulk material as indicated in Fig. 1. The tensile and impact testing samples were extracted from the bulk material by water-jet cutting before milling to shape. The dimensions of these samples were derived from standards ASTM-E8 [34] and BS EN ISO 148 [35], respectively. Tensile testing was performed using an Instron 8802 servo-hydraulic universal testing machine with a capacity of 250 kN in accordance with ASTM-E8. Charpy impact testing was performed using a Losenhausenwerk 14,590 Charpy impact apparatus with a capacity of 290 J at 20 °C.

Fig. 1
figure 1

Schematic of samples extracted from WAAM deposited material (a), dimensions of tensile sample (b) and Charpy sample (c)

Full size image

3 Results and discussion

3.1 Microstructure

Figure 2 shows micrographs of the samples manufactured using both HH and LH parameters at mid-height of the deposition, being representative of the bulk microstructure in larger components. The resulting material is characterised by a dendritic microstructure with a martensitic matrix and interdendritic δ-ferrite (Fig. 2(i)). A directional columnar microstructure is noted due to solidification of dendrites in the direction of cooling (Fig. 2a) at low magnification.

Fig. 2
figure 2

Micrographs of WAAM-produced 15-5PH with high heat input during deposition (1a) and (1b) (HH) and low heat input during deposition (2a) and (2b) (LH)

Full size image

Differences can be noted when comparing the microstructures which evolve with these different parameters. When analysed using the method presented in ASTM E562 [33], the material produced using high weld heat input (HH) displays a δ-ferrite fraction of 12%, while the material produced with a low weld heat input (LH) reduces the δ-ferrite fraction to 6.7%. When this method is applied to the micrographs provided in Guo et al. [22], a δ-ferrite fraction of 22% is found. The number of carbide precipitates in each micrograph (Fig. 2(ii)) was also totalled within equivalent areas. It is noted that the carbide count increases by 100% as the weld heat input is reduced. A comparison of these with 95% confidence intervals is shown in Fig. 3. An examination of these constituents under SEM and using an EDS sensor will allow the chemical composition to be determined, for example the locations indicated in Fig. 4. Locations (i) and (ii) were determined to have composition matching that of the alloy, with their morphology indicating martensite and δ-ferrite are present. The precipitates, such as location (iii), were identified to have a composition rich in carbon suggesting the presence of large carbides within the matrix. Copper precipitates and niobium carbides are expected within this alloy but were not identified due to their small size (3 and 300 nm, respectively) [16, 20]

Fig. 3
figure 3

δ-ferrite volume fraction % and carbide count comparison for WAAM-produced 15-5PH using high and low weld heat input during deposition compared with data from Guo et al. [22]

Full size image
Fig. 4
figure 4

SEM micrograph of WAAM-produced 15-5PH identifying the phases present

Full size image

3.2 Mechanical property testing

Figure 5 shows the results for Vickers microhardness testing with 250 indents. The results show greater hardness correlates with lower weld heat input during deposition (LH). When compared to the material standard A693 [12], it was found that the results for HH material closely correlated with the specified hardness of heat treatment H1025. It was also found that the LH material met the hardness requirements for H900.

Fig. 5
figure 5

Microhardness values from WAAM-produced 15-5PH using high and low weld heat input during deposition compared with values from A693 [12]

Full size image

Tensile testing displayed that the greater weld heat input deposition (HH) resulted in greater yield, UTS and elongation when compared to reduced weld heat input as shown in Fig. 6. In both conditions, all properties meet the requirements for material in the H1150 condition specified in A693 [12].

Fig. 6
figure 6

Results from tensile testing of WAAM-produced 15-5PH using high and low weld heat input during deposition compared with values from A693 [12]

Full size image

Finally, the impact testing results shown in Fig. 7 illustrate that low weld heat input (LH) resulted in a significantly higher impact toughness, meeting the requirements for H1025 and H1150, while the high weld heat input material (HH) did not meet any requirements for impact toughness specified by ASTM A693 [12].

Fig. 7
figure 7

Results from Charpy impact testing of WAAM-produced 15-5PH using high and low weld heat input during deposition compared with values from A693 [12]

Full size image

3.3 Fractography

The fracture surfaces were examined to determine the failure modes and identify differences in the mechanism leading to failure between weld heat input conditions during deposition. The fracture surfaces produced during tensile testing are shown in Fig. 8; they show that failure is dominated by cleavage in the HH material, while a combination of ductile rupture and quasi-cleavage is prominent in the LH material.

Fig. 8
figure 8

SEM imaging of the fracture surfaces of tensile samples produced by HH (a) and LH parameters (b)

Full size image

When the fracture surfaces of the samples tested for impact toughness are examined, a similar trend is noted in Fig. 9, with HH material showing cleavage features across the fracture surface. When LH material is considered, quasi-cleavage is only noted near the notch, with most of the fracture surface displaying dimple rupture with shear wings developing at the edges of the sample.

Fig. 9
figure 9

SEM imaging of the fracture surfaces of Charpy impact samples produced by HH (a) and LH parameters (b)

Full size image

3.4 Discussion

The weld heat input during deposition is shown to have a significant impact on the mechanical properties of PH stainless steel. The as-deposited condition was compared against standard A693 [12] in addition to existing literature using the same process and material with variations in weld heat input. This comparison is shown in Table 3 with the HH material displaying superior mechanical properties to the results from both Caballero et al. [10] and Guo et al. [22].

Table 3 A comparison of testing results from WAAM-produced 15-5PH against material standard ASTM A693 [12]
Full size table

When the LH material is considered, yield strength is shown to be superior to the existing literature while UTS and elongation are comparable to results previously presented. When compared against standard A693 [12], the results for LH material in the as-deposited condition meet the requirements for wrought 15-5PH in the H1150 condition for all properties except hardness. The latter is higher than specified due to a higher concentration of carbides compared to the wrought alloy.

An examination of the fracture surfaces indicates that the HH material experienced brittle fracture, while the dominant mode of failure for the LH material was ductile rupture during tensile testing. This corresponds with the results presented in Niu et al. [20] where post weld heat treatment results in ductile failure and an increase in impact toughness.

There is an inconsistency between the elongation measured during tensile testing and the recorded impact toughness. In many metallic materials, increased elongation will also be associated with increased impact toughness [36]. However, this inconsistency has also been observed in other steel alloys containing δ-ferrite. For example, studies by Wang et al. [37] on stainless steel 13-4 and Rosenauer et al. [38] on 13-8PH noted a decrease in impact toughness when δ-ferrite concentration is increased, which corresponds with the results presented in Table 3.

The HH material displays a larger fraction of δ-ferrite which leads to the embrittlement of the WAAM deposited material [39]. The reduced weld heat input in LH material results in a reduction in the δ-ferrite fraction and improved impact toughness, while the prevalence of carbides within this material corresponds with the increase in hardness over the HH material. It is proposed that δ-ferrite present within WAAM deposited 15-5PH alloy leads to a decrease in hardness, resulting in differences in the microstructure and mechanical properties.

The alloy 15-5PH is frequently used in the manufacture of aerospace components such as landing gear, actuators and fasteners. These applications require a high corrosion resistance [40] and high strength to resist the stresses and harsh environment of aerospace service [17]. In addition, such components must be designed to reduce weight, taking advantage of “design for additive manufacturing” techniques. Therefore, the material must respond well to additive manufacturing processes to produce components with suitable properties [1, 41, 42]. As such, the LH process parameters are recommended, due to their reduced δ-ferrite volume fraction, improving corrosion resistance in addition to high impact toughness and relatively high strength.

4 Conclusions

In this study, the impact of weld heat input on the microstructure and mechanical properties of as-deposited 15-5PH stainless steel manufactured through WAAM was investigated. The conclusions can be summarised as follows.

  • The process parameters presented in this study demonstrate the higher mechanical properties than previously published in the literature for this material and process. The low weld heat input (LH) parameters meet the requirements for all mechanical properties except hardness for wrought alloy in the H1150 condition in ASTM A693. The excess hardness is attributed to an increase in carbides within the microstructure.

  • Increased weld heat input (HH) results in an increased δ-ferrite volume fraction within the microstructure due to a reduction in the cooling rate during the WAAM process. This results in lower impact toughness compared to LH material, which displays a lower δ-ferrite volume fraction and a corresponding higher impact toughness and hardness.

  • The difference in microstructure between HH and LH results in changes in the fracture behaviour in both impact and tensile testing, with HH samples exhibiting brittle fracture, while LH samples display evidence of ductile rupture. Again, this is a consequence of the variation in δ-ferrite volume fraction.

  • An inconsistency has been observed between the elongation and impact results; this difference can be attributed to the variation in volume fraction of δ-ferrite. In this WAAM-produced material, it is noted that the increased δ-ferrite fraction in HH leads to an increase in elongation, but a reduction in impact toughness, making the LH condition more favourable.