Cracking is a persistent problem during the material development for the single material LPBF due to the extreme cooling rate that is inherent to the process. The causes of the cracking problem can be the different solidus and liquidus temperatures at solidification (e.g., liquidus cracking and solidus cracking) or the solid-state phase transformation, precipitation, and property change (e.g., ductility-dip cracking and strain-age cracking). On the other hand, cracking mechanisms like liquid metal embrittlement (LME), cracking at the intermetallic phases and cracking at the fusion zone due to the mismatched CTEs are unique to MMLPBF.

The recent review paper of Wei and Li (2021) nicely described the LME defect in MMLPBF using the Galvele’s atomic surface mobility model (Galvele, 1987). The LME induced cracking is susceptible when the interfacing solid and liquid phases have low miscibility. The larger difference between the melting points of the dissimilar alloys also accelerates the LME cracking. In MMLPBF, the LME cracking is a frequently reported problem for the Cu-Fe systems, which are desired for the fabrication of heat exchangers given the high thermal conductivity of copper alloys and the good high temperature strength and corrosion resistance of steels. At the interface of liquid copper and solid iron, the vacancies exchange and accumulate as the copper atoms diffuse into the iron grains, to form macroscopic defects at the interface that could result in crack initiation under the influence of residual stress. Several studies of MMLPBF Cu-Fe systems (Al-Jamal et al., 2008; Anstaett et al., 2017; Chen K. et al., 2020) reported the formation of microcracks at the interface as shown in Figure 2. As indicated by some of these studies, the formation of cracks could be complicated and is often attributed to multiple mechanisms which are difficult to deconvolute from each other. Both Chen K. et al. (2020) and Anstaett et al. (2017) suggest that the rapid solidification of the copper alloy could initiate cracks in the solidified steel region due to the different CTEs.

Similarly, some brittle intermetallic phases may form at certain composition ratios. The segregated intermetallic phases are susceptible crack initiation sites and preferred crack propagation paths under the cooling induced residual stress. Their CTEs and elastic moduli may be significantly different from those of the primary phase resulting in additional stress. In the graded material from Ti-6Al-4V to Invar 36 built by directed energy deposition (DED), Bobbio et al. (2017) experimentally identified the formation of the intermetallic phases (FeTi, Fe₂Ti, Ni₃Ti, and NiTi₂) in the transition region. The stress between the segregated phases, FeTi and Fe₂Ti, developed during rapid cooling exceeded 4 GPa and eventually led to cracking in the as-built specimens. Similar defects could occur in MMLPBF systems.

Process related defects like keyhole pores and lack-of-fusion in single material LPBF have been extensively studied (Martin et al., 2019; Mostafaei et al., 2022). Their formations are mainly driven by the suboptimal laser energy input. In MMLPBF, the additional materials bring another level of challenges to the optimization of the laser parameters because the optimal laser parameters will depend on the local compositions. That said, the ever-changing local thermophysical properties near the interface, which are often unknown, need to be considered in the optimization.

The high hardness and wear resistance of ceramics is of interest for high temperature and abrasive environment applications, e.g., TiB₂/Ti-6Al-4V. Chen et al. (2019) coupled finite element simulations with experiments in MMLPBF to show that the liquid lifetime, the maximum temperature, and the configurations of the melt pool are critical for the interfacial quality especially when the two materials have very different melting points. Adjusting laser parameters could either benefit the wettability at the interface or lead to the formation of defects due to the capillary instability at the extreme energy levels.

Interfacing materials with different thermal conductivities can also result in the formation of defects and unique microstructures. At the interface of maraging steel and copper, Tan et al. (2018) reported that the onset of lack-of-fusion (Figure 3) occurred at higher laser energy densities for the maraging steel due to the rapid heat dissipation through the more conductive copper base. The high thermal conductivity also caused grain refinement and resulted in stronger <111> texture along the largest thermal gradient as shown in Figure 3.

The process parameters, the boundary orientation, and the thermophysical properties of the materials affect the width of the interface. To attain high compositional precision in components, it is important to understand the width of the interface. Bai et al. (2020) evaluated the interfacial characteristics of the dual interfaces between 316L steel and C52400 copper alloy. As shown in Figure 4, the 316L/C52400 interface (∼200 µm) is significantly wider than the C52400/316L interface. One reason is that the laser parameters of printing 316L are different from those for printing C52400. More importantly, the different thermal conductivities of the substrates magnified the effects from the laser parameters, which in turn changed the remelting behaviors at the interface. Some isolated alloy islands, as highlighted by the blue arrows in Figure 4, were found at locations away from the interface. High undercooling and non-uniform metal liquid flow could be responsible for the formation of these islands.

In the compositionally graded area, the variation of the liquidus/solidus temperatures are mostly non-linear and often raise challenges to the parameter selection. As shown in Figure 5, Bobbio et al. (2017) reported that materials overflowed at the transition region from Ti-6Al-4V to Invar 36 where the liquidus/solidus temperatures dropped at 12—18 vol% of Invar. Excessive remelting occurred even though the process parameters were constant. Note that this example of overflow was observed in a DED process; however, the MMLPBF is also susceptible to overflow at the transition. The outcomes of the varying liquidus/solidus temperatures may be different in the two processes as the MMLPBF parts are supported by the surrounding powder bed. It is expected to see changes in porosity population and surface texture.

As discussed in the previous section, to achieve three-dimensional material gradient/interface it is inevitable to incorporate non-contact powder deposition methods into the spreading process. The multi-material spreader of Wei et al. (2018) utilized a vacuum sucker and a ultrasonic dispenser array to deposit the secondary powder. A higher defect population was found in ultrasonic deposition area. They suggested that the low powder compression and the uneven powder layer were responsible for the defect formation.

Certain material interfaces are difficult to fabricate due to the formation of brittle intermetallic compounds and distinct solidification and shrinkage behaviors which can lead to fracture and delamination. An approach to join these dissimilar materials is to introduce a third material which can form defect-free interfaces with the two primary alloys. Tey et al. (2020) successfully demonstrated a steep material transition from 316L stainless steel to Ti-6Al-4V by adding an interlayer of copper alloy. Another approach is to turn a discrete material interface into a gradient material region to minimize the mismatches of the thermophysical properties of the different composition ratios. This approach has been extensively tested in DED process owing to its capability of pre-mixing powder before powder delivery.

Several studies have discussed the possibility of using simulation approach to improve the interfacial quality. Bobbio et al. (2017) performed thermodynamic calculation using CALPHAD to identify the possible phases in a Ti-6Al-4V/Invar interface. The simulation results can help with the selection of composition ratios at the interface to avoid the formation of detrimental phases. Simulation approaches can also benefit the optimization of laser parameters. Chen et al. (2019) used finite element thermal simulation to identify the process parameters that can result in the optimal melt pool configurations and temperature distributions at the TiB₂/Ti-6Al-4V interface.

The recent effort of Scaramuccia et al. (2020) attempted to optimize the laser parameters for fabricating Ti-6Al-4V and Inconel 718 by mixing different composition ratios using an integrated powder mixer and creating a series of compositional steps over the entire gradient (called a step interface). This is an intermediate strategy of using multiple discrete interfaces to replace a single discrete interface. The study found the optimal energy densities for interfaces with <20 wt% Inconel 718; severe cracking occurred as the Inconel 718 content was above 20 wt% due to the formation of the Ti₂Ni intermetallic phase. This is a great example of finding the optimal laser parameters for the interface of dissimilar metals and identifying the composition ratios that may result in brittle intermetallic phases. The aforementioned CALPHAD approach can certainly benefit the selection of composition ratios in this situation.

To the authors’ knowledge, no printer is capable of adjusting the laser parameters dynamically to deliver the optimal laser energy input according to the material compositions at the target. Knowing the optimized local composition and the corresponding laser parameters are the prerequisites of achieving dynamic laser parameter control in MMLPBF. As suggested by Binder et al. (2018), there are technical barriers that need to be overcome in the pre-processing step. The current data format of the print file only includes the surface representation of the object, and it is not able to assign different parameters to the regions with different compositions. It is important to utilize different print file formats, such as AMF (Additive Manufacturing File Format) (ISO/ASTM 52915, 2020) during the pre-processing stage to enable location dependent parameters within the same part.

In addition to changing laser parameters, the scanning strategy can also be optimized for interfaces with different compositions. Beal et al. (2006) tested four scanning strategies for the H13 steel/Cu interface and showed that the interfaces with different composition ratios responded differently to the four scanning strategies as shown in Figure 6. Zhang et al. (2019) tested the remelting strategy on its ability to impact to the bonding strength of the CuSn/18Ni300 bimetallic interface.

Interface microstructure is an important research topic because it is critical to the performance of the multi-material AM components. Constantly varying local compositions may result in intermetallics and segregation; different thermophysical properties and interface orientations may affect the thermal history producing different grain sizes, morphology, and texture. All these present unique challenges of quantifying and comparing the microstructure in different material systems, which is compounded by the lack of universally agreed upon methods of evaluating the interface microstructure. Only a few comprehensive studies exist focusing on the interface microstructure and are discussed in this section. This presents a knowledge gap which is essential to the future development of multi-material additive manufacturing.

Chen K. et al. (2020) characterized the microstructure at the 316L/CuSn10 interface. The interface exhibits three different distinct regions, 1) 316L steel matrix zone, 2) interfacial region, and 3) CuSn10 matrix zone. The element distribution of the interfacial region is very uniform. All three regions display microscopic anisotropy, equiaxed grain morphology with grain size around 1 μm and various levels of recrystallization ratios. Wei et al. (2019) studied the 316L/Cu10Sn interface by characterizing the microstructure transitions along the step interfaces, i.e., multiple regions with constant ratios of the two alloys. They found that the microstructure started to separate to distinct phases as more Cu10Sn was added and segregation occurred at the interface of 25 vol% 316L/100 vol% Cu10Sn. The study detected phases like Cu_3.8Ni and Cu₉NiSn₃ at the interface. Bai et al. (2020) evaluated the interfaces between 316L stainless steel and C5240 copper alloy and found that only face-center cubic structures formed at the interface, i.e., γ-Fe and ε-Cu phases. The microstructure exhibits strong texture and grain size variation depending on the local compositions. The grain size at the copper-steel interface is significantly smaller than the grain sizes in the single material areas. They hypothesized that the pre-solidified 316L offered the nucleation sites for the subsequent grain growth; this phenomenon is unique to multi-material printing.

Different from the Fe/Cu interfaces, the Ti/Cu interface resulted in more phases such as L₂₁ ordered phase, amorphous phase, Ti₂Cu, and ε-Cu as reported by Tey et al. (2020). At the dissimilar interface of titanium alloys, Ti-5Al-2.5Sn/Ti-6Al-4V, Wei et al. (2020) showed that the epitaxial growth path of the prior-β grains crossed the interface and sustained for several millimeters. Martensitic transformation occurred because of the high cooling rate resulting the formation of the needle like α’. Microscopy indicated that the metastable α’ martensitic microstructure within the Ti-5Al-2.5Sn region was fully recrystallized; by contrast, the recrystallization process did not occur in the Ti-6Al-4V portion due to the existence of the β stabilizer element V.

At the interface of aluminum alloys, Al-12Si and Al-Cu-Mg-Si, Wang et al. (2020) observed a more gradual transition of grain size and morphology. The columnar grains grew from 7.9 μm at the interface to 23.3 μm within the Al-Cu-Mg-Si region. Additionally, the relative amount of <001> fiber texture along the building direction at the interface is significantly higher compared with that in the single material region. Intermetallics are common at the dissimilar metal interface. At the interface of W-Cu, Tan et al. (2018) reported a trend of increasing tungsten grain size when moving away from the interface due to the high thermal conductivity of Cu. They also revealed the random orientation of the columnar grains within the tungsten region which was attributed to the laser scanning strategy with 67° rotation.

An early feasibility study of multi-material LPBF fabricated three 1 mm tall regions starting with a pure iron region, followed by a 55 vol% Fe, 45 vol% Al-12Si region, and then a 100 vol% Al-12Si region (Demir and Previtali, 2017). The multi-material region had much higher microhardness (450–550 HV0.5) compared to the single materials, pure Fe (150–160 HV0.5) and Al-12Si (90–100 HV0.3), due to the formation of the FeAl intermetallic Other ferrous alloys have also been studied, an iron-silicon alloy (Fe–3Si) and a quenched and tempered steel (34CrNiMo6) build had a 120—150 μm thick interface that exhibited good metallurgical bonding independent of what alloy was the base material or built first (Andreiev et al., 2021). The interface, characterized via microhardness, linearly increased and decreased from the Fe–3Si with an average hardness of 225 HV0.3 to 34CrNiMo with an average hardness of 495 HV0.3, depending on the build order. The UT specimens always broke in the Fe-3Si part in both build arrangements and exhibited properties similar to the individual Fe-3Si samples. High-cycle fatigue of the bimetallic samples after being annealed at 850 °C, quenched in oil, and then annealed/tempered at 550 °C had fatigue properties similar to the Fe-3Si samples. There are too few studies on ferrous interfaces to make a broad conclusion about the mechanical response of these materials in MMLPBF.

Stainless steel 316L is one of the most widely studied materials in multi-material LPBF processes, as shown in Table 2, Table 3. The face centered cubic (FCC) 316L alloy exhibits good ultimate tensile strength (630 MPa), Vickers microhardness (220 HV), and ductility (64% elongation to failure) as an as-built LPBF material (Sun et al., 2016; Wilson-Heid et al., 2019). In addition to the mechanical properties, 316L is also of interest in multi-material systems because of the materials corrosion resistance, which makes it a good pair for materials that do not exhibit the same properties. The existing process parameter knowledge base for manufacturing dense 316L samples is relatively extensive in the single-material LPBF community. This allows for studies to focus on the unique aspects of MMLPBF, such as hardness changes across an interface or build material order effects, after adding materials to 316L and not parameter development of the alloy.

The interface of maraging steel MS1 and 316L was characterized with Vickers microhardness across both discrete and gradient transitions, 12.5 wt% MS1 addition into 316L every five layers of until 100 wt% MS1. The microhardness in the gradient sample had a less abrupt change in hardness within 100 µm on either side of the interface compared to the discrete interface, whose microhardness increased in a similar trend to the wt% added of MS1 (Nadimpalli et al., 2019).

Inconel 718, also an FCC alloy, added to 316L has been studied because of the benefits of added high temperature mechanical properties, observed in both wrought and LPBF conditions (Sanchez et al., 2021). Using microhardness to evaluate the 100 µm fusion zone between the two materials, Yusuf et al. (2021) found an average hardness of 265 HV0.1 compared to 304 HV0.1 and 223 HV0.1 at the individual IN718 and 316L regions, respectively. The intermediate hardness in the fusion zone was attributed a lack of low-angle grain boundaries compared to those observed in the IN718 region, although both regions possessed a large fraction of high-angle grain boundaries. Other studies have also measured microhardness across the 316L-IN718 interface and found an intermediate hardness of the two primary constituents in the fusion zone (Mei et al., 2019; Duval-Chaneac et al., 2021). Duval-Chaneac et al. (2021) evaluated the fatigue properties of the 316L-IN718 multi-material system using notched three-point bend samples, where the interface boundary for the fatigue samples was perpendicular to the loading direction and found that in a sample with a single interface there was elemental partitioning that weakened the interdendritic areas and lead to decohesion and fast crack propagation. However, in a 4-layer fatigue bending sample, the multiple interfaces provided crack shielding that decreased crack growth rate in the additional interface leading to improved fatigue performance relative to the bi-layer specimen, which shows the importance of testing many configurations of the multi-material systems.

Rankouhi et al. (2022) tested Hastelloy X (HX), another FCC nickel superalloy, bonded to 316L through a combination of tensile and 3-point bend tests and determined a strong 240 µm fusion zone was formed. In the UT tests, failure occurred away from the interface in the 316L region and had similar strength to 316L alone, but lower average failure strain (43.8%) compared to HX (62.1%) and 316L (62.7%) as shown in Figure 8. In both the UT and 3-point bend tests no evidence of cracking or voids were found at the interfaces.

Copper and its alloys are of interest in the multi-material AM community because of potential applications to add highly conductive regions of components that would boost performance capabilities. Pairing copper with another material that possesses complimentary and improved mechanical properties such as 316L is an appealing application for multi-material LPBF.

In two studies by Chen et al., the 316L and CuSn10 tin-bronze system was explored (Chen J. et al., 2019; Chen et al., 2020 J.). They reported a 243 µm wide fusion zone without defects, a width that was in the middle of all recorded values after a broad sweep of process parameters, resulted in the best UT properties with the joint ultimate strength of 460 MPa and elongation to failure of 5%; the tests were conducted with the loading direction perpendicular to the interface (Chen J. et al., 2019; Chen et al., 2020 J.). Uniaxial tension tests parallel to the interface were lower performing than the two alloys alone. Nanoindentation characterization of the interface revealed a higher nano-hardness in the interface (2.97 GPa) compared to either of the constituents, which was attributed to fine grains in the interfacial region. However, Vickers microhardness does not show an increase at the fusion zone, but rather was an intermediate value between the two alloys. This is in contrast to another study of the same material interface where both microhardness and nano-hardness were found to be higher in the interface of 316L and CuSn10, where they attributed this finding to the formation a new intermetallic phase (Wang et al., 2021). Finally, 3-point bend tests revealed dendritic cracks near the boundary that resulted in lower maximum flexural strength in the multi-material samples. Evaluating the same system other authors found the flexural strength of 316L/CuSn10 was between 316L and CuSn10, regardless of which material was used as the base material (Chen K. et al., 2020).

The 316L/CuSn10 interface was the focus of another study that used 316L material as an intermediate step in the fabrication of tungsten W) to CuSn10 multi-material parts and concluded that the addition of the 316L region was necessary as they were unable to successfully manufacture W/CuSn10 bimetallic components without delamination (Wei et al., 2022). In the evaluation of the interfaces in the three-alloy component the authors concluded that microhardness testing was inadequate for the W/316L interface because the large difference in microhardness led to slipping of the indenter tip, instead nano-hardness is necessary in fusion zones of materials with large differences in hardness.

Another copper alloy, C52400, was also used with 316L as the base material, but an additional interface was added on top of the copper alloy by sandwiching the copper with 316L, resulting in a 9 mm tall build where each section was 3 mm (Bai et al., 2020). Both interfaces, the 316L/C52400 and C52400/316L, were evaluated with Vickers microhardness measurements. In the first interface the hardness was intermediate between the two alloys gradually falling in the interface region. In the second interface, there was a high peak hardness that was attributed to lack of intermixing, which was also observed in another study with a different copper alloy where bands of 316L remained in the interface (Tey et al., 2020). Tey et al. (2020) used a copper alloy (HOVADUR^® K220) as a thin interlayer to transition from 316L at the bottom to a Ti-6Al-4V region on the top of a build, as shown in Figure 9. In fractography of uniaxial tension tests the authors observed that in the copper alloy and Ti-6Al-4V interface region the α′-Ti phase that formed acted to arrest and deflect propagating cracks in the ductile copper alloy region improving the mechanical properties, where the sample with a higher concentration of α′-Ti phase had improved strength.

Further expanding the number of interfaces evaluated in a multi-material LPBF build, Wang et al. (2020) manufactured a four-alloy component consisting of base CoCrMo alloy, then 18Ni300, followed by CuSn10 and finally 316L. The familiar CuSn10/316L interface was the only interface where microcracks formed, and each interface had different fusion zone trends with respect to microhardness and nano-hardness. The 230 µm wide CoCrMo/18Ni300 interface had lower hardness that the constituents, the broader 345 µm wide 18Ni300/CuSn10 interface exhibited an intermediate hardness, and the narrowest 135 µm CuSn10/316L interface had a higher hardness than either alloy. The 18Ni300/CuSn10 interface was also examined in a different study that fabricated cylindrical uniaxial tension and lattice compression specimens (Zhang et al., 2019). The tensile properties of the multi-material samples were weaker than LPBF CuSn10 alone, 144.1 ± 41.59 MPa and 441.0 ± 7.048 MPa, respectively. However, the compression tests showed improved energy absorption properties in the bimetallic lattice structure relative to the LPBF CuSn10 fabricated in the same geometry. Testing bulk properties versus those in lattice geometries reveals that multi-material samples may be best used in a variety of different applications and should be evaluated across a wide array of test geometries to determine the proper application space.

Other alloys have also been explored for bonding with copper alloys with specific applications in mind such as in heat exchangers and electrical connectors (Sing et al., 2015; Marques et al., 2022).

In the study of components with a discrete gradient between UNS C18400 copper alloy and AlSi10Mg the bimetallic UT specimens, with the loading direction parallel to the interface, exhibited higher tensile strength (176 ± 31 MPa) than that of copper but lower than that of AlSi10Mg (Sing et al., 2015). The authors completed 3-point bend tests that showed a large difference in flexural strength depending on sample orientation, the flexural strength when the Cu alloy was on the bottom was 200 MPa and it was 500 MPa when the aluminum alloy was on the bottom. The variation in properties was hypothesized to be caused by the higher amount of porosity in the copper region that when in tension at the bottom of the samples led to earlier crack propagation and failure. This finding shows the potential to use materials that are more prone to porosity formation from LPBF processing in the component locations that are only designed to remain in compressive loading, while using more printable, dense alloys in tension loaded regions of the same component.

In an XY gradient structure of IN718 and pure Cu, that is the two materials are used on the same build layer instead of consecutive layers like a majority of the multi-material LPBF parts are, the Vickers microhardness measurements were unable to detect an independent hardness value of the thin 25 μm fusion zone and instead the parts had two distinct hardness values, 344 HV in the IN718 regions and 126 HV in the pure Cu regions (Marques et al., 2022). The XY multi-material components present an interesting opportunity to learn about interfacial bonding and as the field matures, further mechanical evaluations should be performed.

Titanium and its alloys, primarily Ti-6Al-4V, are used in a wide variety of applications ranging from use in aerospace, for their combination of the high strength and low density, to biomedical applications due to properties that promote biocompatibility (Peters et al., 2003; ASTM Standard F1108-21, 2021). In LPBF literature, Ti-6Al-4V has been extensively studied (Martin et al., 2019; Furton et al., 2021; Luo et al., 2022), which makes it an easier starting point for manufacturing multi-material components, similar to the usage of stainless steel 316L.

The effect of varying discrete interface configurations was evaluated in pure Ti and Ti-6Al-4V bimetallic components. Four variations of interfaces were fabricated, as shown in Figures 7C–F, the first three evaluated the interfaces parallel to the loading direction: a single interface down the length of the UT sample, a Ti/Ti-6Al-4V/Ti sandwich through the width of the flat UT sample, a Ti/Ti-6Al-4V/Ti sandwich through the thickness of the flat UT sample, and in the last variation the loading direction was perpendicular to two Ti-6Al-4V regions in the gauge section surrounded by Ti (Borisov et al., 2021). The authors found that the Ti-6Al-4V sandwich through the thickness of the flat UT sample resulted in the best mechanical behavior, exhibiting the highest UTS (839 ± 14 MPa) and elongation to failure (7% ± 3%) compared to the other bimetallic configurations, the strength was between the pure Ti (700 ± 12 MPa) and pure Ti-6Al-4V (998 ± 21 MPa) samples while the ductility was lower than both the Ti (16% ± 4%) and Ti-6Al-4V (10% ± 3%) only samples. The sample with worst UT properties were the bimetallic samples that had interfaces perpendicular to the loading direction, which highlighted the importance of understanding the loading orientations with respect to interfaces for multi-material parts built via LPBF.

In another study that primarily used UT tests to evaluate mechanical bonding, a Ti-6Al-4V and Ti-5Al-2.5Sn multi-material component was reported to have tensile strength and elongation to failure significantly lower than that of LPBF Ti-6Al-4V, but relatively close to that of LPBF Ti-5Al-2.5Sn (1,074 MPa and 7.7%, respectively) (Wei et al., 2020). The authors concluded that the softer Ti-5Al-2.5Sn determined the mechanical properties of the multi-material and not the 70 µm wide interface, built perpendicular to the loading direction, that was formed between the alloys.

In addition to using other Ti-based alloys, Ti-6Al-4V was graded to IN718 by Scaramuccia et al. (2020) in 10 wt% increments up to 40 wt% IN718 and then was transitioned to 100 wt% IN718. The gradual introduction of IN718 led to a measured microhardness in the 10 wt% IN718 region (381 ± 21 HV) to be lower than Ti-6Al-4V (402 ± 7 HV) due to the addition of beta phase stabilizers from the IN718 alloy. However, at 20–40 wt% the hardness increased dramatically to 477 ± 16 HV for 20 wt%, 684 ± 48 HV for 30 wt%, and 582 ± 27 HV for 40 wt% IN718 content, due to the formation of hard intermetallic phases that led to significant cracking in the 30 and 40 wt% samples, and then dropped back down to the expected lower hardness for 100 wt% IN718 (255 ± 13 HV). The formation of intermetallic phases can be detrimental to other mechanical properties and printability due to their brittle nature.

Shear bond tests were using to quantify the bonding strength of a NiTi and Ti-6Al-4V multi-material lattice structure with applications for biomedical hip implants (Bartolomeu et al., 2020). The shear bond test consisted of a custom-made sliding shear set-up where the shear load was applied to the transition region of the samples to determine the shear bonding strength of the fusion zone. The NiTi/Ti-6Al-4V samples had a shear strength of 33.2 ± 5.3 MPa, which was between the shear strength, of the two mono-materials evaluated with the same sample geometry, 25.5 ± 4.8 MPa and 47.1 ± 4.0 MPa for NiTi and Ti-6Al-4V, respectively. In a different porous cellular geometry, the Schoen Gyroid lattice structure, TiB₂ was added in increments of 1 wt%, up to 3 wt%, to a Ti-6Al-4V initial region (Zhang et al., 2020). Vickers microhardness was recorded in each region and the material exhibited an increase in hardness as a function of TiB₂ content going from 370.3 HV in the Ti-6Al-4V region to 428 HV with 3 wt% TiB₂. Compression tests of the gyroid structures revealed that compression stress-strain behavior of TiB/Ti-6Al-4V samples had a smaller elastoplastic stage and fracture strain than Ti-6Al-4V samples no matter the relative density of gyroid.

Titanium based MMLPBF studies have shown that build order and fusion zone orientation with respect to the loading direction, specifically in uniaxial tension tests, are important design criteria when evaluating mechanical performance of MMLPBF parts. The testing of AM specific geometries fabricated with multi-materials, presents both the challenges of correctly testing the geometry and interfaces simultaneously, although compression and shear tests have proven to be useful tests based on the expected applications of the structures.