Co-Extrusion of Dissimilar Aluminum Alloys via Shear-Assisted Processing and Extrusion

Abstract

Bimetallic tubes are used when a component requires more than one performance requirement, for instance, strength/creep resistance and oxidation/corrosion resistance. Shear-assisted processing and extrusion (ShAPE) can be used to fabricate extrudates that are comparable or superior in performance relative to conventionally manufactured extrudates. For the first time, ShAPE has been successfully employed in producing bimetallic Al tubing consisting of 6061 and 7075 Al alloys. Both light and electron microscopy techniques were used to investigate the integrity of the tubes, especially the interface between the core and cladding of the bimetallic tubes. Void-free bimetallic tubes were produced using ShAPE. Quantification of tube integrity was carried out with tensile testing under as-extruded and post-aged conditions. All the bimetallic tubes in the as-extruded samples exhibited uniform elongation above 5% with good tensile strength. Key insights such as material flow during bimetallic tube extrusion were obtained from the characterization of remnant billet and simulation results from a smoothed particle hydrodynamics model.

1. Introduction

In many applications, more than one material property is required for a component to meet performance requirements, for instance, strength or creep resistance to provide structural integrity and corrosion/oxidation resistance to protect the components from severe environments in power plants. Frequently, bimetallic components are employed to achieve these distinct property requirements. The most important requirement of bimetallic structures is a strong interfacial metallurgical bond between the substrate and the coating to accommodate stresses resulting from differences in coefficients of thermal expansion and to avoid/reduce spallation of the coating. Various methods, such as roll bonding, explosive bonding, weld overlay, anodization treatment, and co-extrusion, are currently employed to obtain bimetallic/cladded components, and each method has its own unique advantages and disadvantages. Roll bonding is usually employed to produce clad/Al-clad structures with a corrosion-resistant high-purity Al layer on high-strength precipitation hardenable Al alloys [1], while anodization treatment is employed for forming a thin corrosion-resistant coating layer usually in the order of a few micrometers [2]. For applications in harsh environments, a thick corrosion-resistance coating is preferred, which is mostly fulfilled via co-extrusion. Conventional co-extrusion is also employed to produce cladded components such as Inconel 671 on alloy 800, austenitic stainless steel on carbon steel [3], 6063 on 7075 Al using porthole die extrusion [4], and copper clad aluminum [5]. Co-extrusion of 1100 Al with a ZK60 Mg plate was accomplished via conventional extrusion for the corrosion protection of Mg alloys [6]. Both single and multilayer co-extruded profiles were obtained with no visible defects along the interface. Furthermore, the total elongation of the multilayered structures was lower than that of the single co-extruded plates and was attributed to the increased fraction of intermetallic compounds. The bimetallic billet assembly for conventional extrusion is typically achieved through compound casting, where the core is cast first then the shell [7,8] or is fusion-welded along the core–shell interface [9] to promote good metallurgical bonding in the cladded tubes. For instance, Chen et al. [7] fabricated bimetallic billets through the compound-casting process using 7075 Al as the core and 6060 as the cladding, then completed hot co-extrusion of the cast billets. The extrudates exhibited a defect-free, metallurgically bonded interface between the core and cladding. The importance of fusion-welded bimetallic billets can be noted through the co-extrusion of Zircaloy-2 and Zr-Sn carried out by Jha et al. [9]. The absence of welding and degassing for the bimetallic billets led to the formation of voids along the core–cladding interface, while adding these steps during billet preparation led to a defect-free interface. Therefore, to obtain a metallurgically bonded interface via conventional co-extrusion, specific billet preparation steps are needed.

In this investigation, co-extrusion of 6061 Al with 7075 Al is examined using shear-assisted processing and extrusion (ShAPE) to investigate the feasibility of ShAPE in fabricating defect-free cladded tubes with good interfacial bonding [10,11]. ShAPE is a solid-phase extrusion process that exerts extreme plastic strain and elevated temperature (50%–80% of melting temperature) into extrudates, resulting in refined microstructures and excellent mechanical properties [11]. Initial research on the co-extrusion of 1100 Al and 7075 Al via the ShAPE process was successful and reported in a conference proceeding [12]. In this current manuscript, a detailed microstructural and mechanical characterization of the co-extruded 6061 and 7075 Al is presented. Furthermore, a process simulation model based on the meshfree smoothed particle hydrodynamics (SPH) method is used to understand material flow during ShAPE co-extrusion, which is validated by experimental analysis.

Compared to the conventional extrusion process, ShAPE involves the rotation of the extrusion die, which results in additional strain components such as radial strain. This is expected to result in a metallurgically bonded interface in cladded tubes fabricated from mechanically interlocked bimetallic billets without extra steps needed to fabricate either welded or compound cast bimetallic billets. The current manuscript provides evidence that cladded Al tubes with a metallurgically bonded interface can successfully be fabricated from high-strength Al alloys (7075 Al and 6061 Al) via this process.

2. Experimental Methods

ShAPE experiments were carried out using a ShAPE machine manufactured by BOND Technologies, Inc. (Elkhart, IN, USA) and installed at Pacific Northwest National Laboratory (PNNL). The experimental setup to produce co-extruded tubes is presented in Figure 1. Figure 1a–c show the billet assembly for the co-extrusion process. Two designs were employed in this proof-of-concept ShAPE experiment to fabricate co-extruded Al tubes. In design 1, the billet volume fractions of 6061 and 7075 were the same (each had 50% of the total billet volume), and in design 2, the volume of 6061 was 66.7% of the total billet volume and 7075 made up the remaining billet volume. The schematic of the longitudinal cross section of the billet from design 1 is presented in Figure 1a and the transverse cross sections of designs 1 and 2 are shown in Figure 1b and Figure 1c, respectively. In both designs, 7075 formed the core while 6061 formed the cladding. The length and diameter of the billets were ~101.6 mm and 31.6 mm, respectively. A schematic of the ShAPE processing setup is presented in Figure 1d. Extrusion dies with an outer diameter of ~31.7 mm and an inner diameter of ~12 mm were fabricated from H13 tool steel. The diameter of the H13 tool steel mandrel was ~10 mm, thereby fabricating extruded tubes with an inner diameter of ~10 mm and outer diameter of ~12 mm. The extrusion ratio (ratio of the area of the billet to the tube cross section) was set to be 20.6.

During extrusion, the die was rotated and plunged into the billet. Consequently, the billet material experiences adiabatic and frictional heating and was plasticized in situ before flowing into the gap between the inner diameter of the extrusion die and outer diameter of the mandrel to form hollow tubes. The processing temperature was measured using a K-type thermocouple that was spot-welded onto the extrusion die face at about 10.2 mm from the center. The major processing variable in this study was the feed rate. A summary of the ShAPE processing conditions is presented in

Table 1. Summary of the ShAPE processing conditions employed to fabricate the cladded tubes.

Material Combination	Feed Rate (mm/min)	Die Rotational Speed (RPM)	Tube Length (mm)
Design 1–60	60	27	1956
Design 2–60		27	1981
Design 1–10	10	30	1803
Design 2–10		27	1820

. Designs 1 and 2 were processed at 60 and 10 mm/min die advance rates with minor changes in the die rotation speed to maintain a similar processing temperature between extrusions. The terminology to describe the two designs and two feed rates is presented in Table 1 and is used throughout this manuscript. The data acquisition for processing output such as temperature and force was kept at 100 Hz, and the raw data is reported in this manuscript. The fabricated tubes were sectioned and subjected to various microstructural and mechanical property characterizations. Billet remnants were also investigated to understand the material flow near the die face and orifice during the co-extrusion process.

After extrusion, transverse and longitudinal sections were taken from the extrudate tubes and billet remnants, mounted in epoxy, and ground with SiC abrasive papers, followed by final polishing with a 0.05 µm colloidal silica suspension. The samples were etched with Keller’s reagent for light microscopy analysis. An Olympus SZX16 stereo microscope was used to obtain a macro-overview of the transverse and longitudinal tube sections. The thickness of the core and cladding for all the conditions was measured using ImageJ 1.53e (image processing software, https://imagej.net/ij/, accessed on 20 December 2023). The samples were then repolished to a 0.05 m surface finish to observe sub-micron microstructural features using scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDS) techniques. SEM and EDS (area and line) analyses were performed using a JSM-IT500 scope (JEOL USA, Inc. Peabody, MA, USA). Light and electron microscopy analyses were mainly carried out to investigate the integrity of the tubes in as-extruded conditions.

The extruded tubes were solutionized at 753 K (450 °C) for 1 h and stored in a freezer at 204 K (−69 °C) to avoid natural aging. Various artificial aging heat treatments were selected to increase the strength of the 6061 and 7075 Al alloys: (1) 448 K (175 °C) for 8 h, (2) 423 K (150 °C) for 2 h, (3) 394 K (121 °C) for 24 h, (4) 394 K (121 °C) for 3 h + 438 K (165 °C) for 4 h, and (5) 378 K (105 °C) for 6 h + 448 K (175 °C) for 8 h. These heat treatment conditions were selected from the ASM Handbook, Volume 28, Properties and Selection of Aluminum Alloys [13]. All the samples were mounted and polished for Vickers microhardness measurements. Hardness measurements were carried out using a Clark Microhardness tester CM-700AT (Sun-Tec Corporation, Novi, MI, USA) with a testing load of 300 g and a dwell time of 12 s. A total 10 indents per condition was measured with 0.5 mm spacing between indents and the average value with standard deviation is reported. Based on the hardness analysis, the tensile samples underwent the aging heat treatment condition that resulted in the highest hardness in both 6061 and 7075 Al. In accordance with ASTM-B557 [14], tensile testing of 127 mm long tubes in as-extruded and solutionized and aged conditions was carried out using a 22 kN capacity MTS servohydraulic testing machine at an initial strain rate of 5.1 mm/min.

3. SPH Modeling

Numerical modeling based on the meshfree SPH method was carried out [15,16] to better elucidate the material flow, temperature distribution, and stress–strain state during the ShAPE processing of the bimetallic tubes. The major benefit of SPH is its meshfree nature, which enables easy handling of the severe plastic deformation during the ShAPE process [17]. SPH also has a Lagrangian description and can track field variables accurately and explicitly at various time steps. More importantly, the particle representation of SPH can describe free surfaces and multimaterial interfaces in a natural way, which is a challenging task for conventional mesh-based finite element methods. These features make the SPH model ideal for simulating the ShAPE process for bimetallic systems. SPH has been recently used for modeling the friction stir extrusion of 1100 wires [18] and the ShAPE of 7075 tubes [19]. In this study, we used SPH for the first time for modeling the process of bimetallic co-extrusion using ShAPE.

The SPH model setup is shown in Figure 2a, which consists of a rigid billet container, a rigid four-scroll die (Figure 2b), and a deformable SPH billet composed of 7075 as the core and 6061 as the cladding. The ShAPE processing condition of design 2–60 shown in Table 1 was simulated with the billet configuration, as indicated in Figure 2. A billet length of 10 mm was used. The SPH particle size was 0.125 mm, leading to a total of 3,590,720 SPH particles in the domain. The billet container was discretized into 327,432 hexahedral bilinear elements while the die was discretized into 260,405 tetrahedral bilinear elements. Ghosted SPH particles were used for the rigid container and die, enabling heat transfer between the SPH billet and the neighboring rigid parts [18]. A simplified billet container geometry and a shortened billet were used in the simulation to reduce the computational cost because applying exact boundary conditions is nearly impossible. Thus, a thermal convection boundary condition with a convective coefficient of 1000 W/(m²·$°$C) was prescribed on the outer surface of the container to represent the active water cooling that stabilizes the steady-state temperature in the experiment. This convective coefficient is effective and accurate, as demonstrated in our previous study [19].

The Johnson–Cook (J-C) constitutive model [20] is used to describe the thermo-visco-plastic behavior of the billet materials, which is expressed as follows:

$${\sigma }_f=\left[A+B{\epsilon }_e^{p^n}\right]\left[1+C\ln \left(\frac{{\dot{\epsilon }}_e^p}{{\dot{\epsilon }}_0^p}\right)\right]\left[1-{\left(\frac{T-T_r}{T_m-T_r}\right)}^m\right]$$

(1)

where ${\sigma }_f$ is the flow stress, ${\epsilon }_e^p$ is the effective plastic strain, and ${\dot{\epsilon }}_e^p$ and ${\dot{\epsilon }}_0^p$ denote the effective plastic strain rate and reference total strain rate, respectively. $A$, $B$, $C$, $n$, and $m$ are material parameters: $A$ is the initial yield stress, $B$ and $n$ are the strain-hardening constants, while $C$ denotes the material sensitivity to the plastic strain rate. $T_r$ is the reference temperature at which the material parameters are evaluated, $T_m$ is the material solidus temperature, and $T$ is the instantaneous temperature. As discussed in [18], a main drawback of the J-C model for solid-phase processing simulation is its nonsaturating strain-hardening effect under conditions of high strain, leading to an incorrect high flow stress even at 90% of the material melting point. To address this issue, we used the J-C model parameters for 6061-T6 [21] and 7075-T6 [22] with modified $A$, and set the $B$ and $n$ values to zero to remove the strain-hardening effect of the model. The detailed J-C model parameters and thermal properties used in the SPH model at the steady-state temperature of 380 $°$C are presented in

Table 2. Johnson–Cook parameters and thermal properties for 6061-T6 [21] and 7075-T6 [22].

Parameter	Description	Unit	6061-T6	7075-T6
			Values
$A$	J-C material model parameter	MPa	36.86	51.69
$B$		MPa	0	0
$n$		-	0	0
$C$		-	0.002	0.024
$m$		-	1.34	1.56
$T_r$		$°\mathrm{C}$	24	24
$T_m$		$°\mathrm{C}$	640	628
${\dot{\epsilon }}_0^p$		${\mathrm{s}}^{-1}$	1	1
$\rho$	Density	g/cm3	2.7	2.81
$E$	Young’s modulus	GPa	48	31.5
$\nu$	Poisson’s ratio	-	0.33	0.33
$C_p$	Heat capacity	J/g·°C	1.11	1.28
$T_c$	Thermal conductivity	W/m·°C	181	168

.

A partial stick–slip contact condition was assumed at the die–billet interface by using the Coulomb–Tresca friction law with a frictional coefficient of 0.6 [16]. Ninety percent of the material plastic deformation energy was converted to heat, while all of the die–billet frictional sliding energy was converted to heat. An initial temperature gradient was assigned to the model to avoid the transient state based on previous SPH modeling efforts. To this end, initial temperatures of 380 °C (steady-state temperature measured from the experiment) and 200 °C were assigned to the die and billet container, respectively. The SPH billet was equally sectioned into five layers in the longitudinal direction (Figure 2c). The top layer in contact with the die face was assigned an initial temperature of 440 °C due to the expected higher temperature in the processing volume, and a temperature difference of 20 °C was used for the other four layers (Figure 2c).

Because excessive material deformation and complex contact conditions are involved in ShAPE co-extrusion, the SPH model was solved using a mixed implicit/explicit analysis in a staggered manner. That is, the mechanical analysis was first carried out using explicit analysis with a conditionally stable time step of $\mathsf{\Delta }t$ = 3.33 $\times$ 10⁻⁴ to determine particle velocity, location, stress–strain state, and heat generation in the current time step. Then, the thermal analysis was carried out using implicit analysis with an unconditionally stable time step of $\mathsf{\Delta }t$ = 3.33 $\times$ 10⁻⁶ to determine temperature distribution in the current time step. The SPH model was implemented in commercial software LS-DYNA and all the simulations were run on 24 cores parallelly and solved using the LS-DYNA R12.0 Massively Parallel Processing (MPP) solver with double precision. It took around 72 h for a single simulation to finish on a dual Intel Haswell E5-2670 24-core 2.3 GHz CPU.

4. Results and Discussion

4.1. Process Data

The process temperature of the thermocouple attached to the extrusion die is presented in Figure 3. Due to the nature of the process, the temperature quickly increased at the start of the extrusion. The temperature rise was due mainly to adiabatic and frictional heating components from the interaction between the extrusion die and material. After about a 10 mm plunge depth, the temperature reached a steady state, signifying a balance between heat generation and heat dissipation rates. Overall, for all the conditions, more than 87% of the extrusion occurred under steady-state deformation conditions. As noted in the table in the inset of Figure 3, the steady-state temperatures varied based on the feed rate. The rotational speed of the die was the same among the four extrusions and the temperature variations arose only from the feed rate differences. Lower steady-state temperatures were noted for extrusions carried out at a 60 mm/min feed rate, while relatively higher steady-state temperatures were noted for extrusions performed at a 10 mm/min feed rate. Furthermore, a similar processing temperature was noted between the two billet designs processed at the same feed rate. A summary of all the average steady-state temperatures is presented as an inset table in Figure 3. Samples for microstructural and mechanical property characterization were taken from the steady-state extrusion region.

4.2. Microstructural Characterization

Macro-overviews of longitudinal and transverse sections of the tubes (designs 1–60, 1–10, 2–60, and 2–10) are presented in Figure 4. In all four extrusions, the outer layer (cladding) is 6061 and the inner layer (core) is 7075. Continuous core and cladding were present on all of the analyzed cross sections. More importantly, defect-free cross sections were observed in all four extrusions. The average and standard deviation of the core and cladding thicknesses for various extrusions are presented in

Table 3. Summary of the core and cladding thicknesses for all four extrusions.

Material Combination	Average Thickness (mm)
	Core	Cladding
Design 1–60	0.5 ± 0.02	0.45 ± 0.04
Design 1–10	0.54 ± 0.07	0.47 ± 0.1
Design 2–60	0.34 ± 0.03	0.62 ± 0.03
Design 2–10	0.38 ± 0.03	0.65 ± 0.05

. As stated above, billet design 1 contained equal volumes of core and cladding, while billet design 2 consisted of 75% cladding and 25% core. The measured thicknesses of the cladded tubes reflect the billet designs. For instance, the thicknesses of the core and cladding measured along multiple locations in design 1–60 were 0.5 ± 0.02 and 0.45 ± 0.04 mm, respectively. And in the case of design 2–60, the thicknesses of the core and cladding were 0.34 ± 0.03 and 0.62 ± 0.03 mm, respectively. Therefore, co-extrusion via ShAPE can fabricate cladded tubes with a designed cladding-to-core ratio. Comparable average core and cladding thickness values were noted between designs 1–10 and 2–10 as well. Notably, the extrusions fabricated at a 10 mm/min feed rate exhibited a higher average thickness standard deviation and visibly wavy interfaces compared to the extrusions at 60 mm/min. This is mainly due to the increased revolutions the material goes through in a millimeter. For instance, at a 60 mm/min feed rate, the material goes through about 0.45 revolutions in a millimeter, while at 10 mm/min, the material goes through 2.7 revolutions.

The results of the SEM-BSE analysis of all four extrusions are presented in Figure 5. The first column contains low-magnification SEM-BSE images of the longitudinal cross section. The middle and right-hand columns of the images are high magnifications focused on the interface between 6061 and 7075. The interface regions between 6061 and 7075 are marked by solid arrows. The interface of design 2–60 exhibited a few micro-cracks, as marked by dotted arrows in Figure 5(c₂). The interfaces of the remaining three extrusions did not exhibit any discontinuities. The interface is continuous, as observed in the high-magnification images, and is presented in the right-hand column (Figure 5(a₃,b₃,c₃,d₃)); hence, the core and cladding are metallurgically bonded. The residual darker regions near the interface are not porosity/voids but Mg- and Si-rich particles whose presence will become clear with the SEM-EDS discussion presented in the next subsection.

SEM-EDS analysis of the design 1–60 cladded tube is presented in Figure 6. EDS elemental maps are displayed in Figure 6b–e and line EDS scan results across the interface are shown in Figure 6f. Interface regions are marked with solid arrows. The regions above and below the interface are 6061 and 7075, respectively. As expected, the presence of Zn and Si was noted in the 7075 and 6061 alloys, respectively. In the 6061 region, just above the interface, spotted regions of darker contrast were observed and are marked with dotted arrows. These correspond to Mg- and Si-rich particles as seen in Figure 6d,e. Furthermore, based on the line scan EDS analysis, Zn content increased and Si decreased across the cladded interface from 6061 to 7075. Both SEM-BSE and SEM-EDS confirm the presence of a metallurgically bonded interface between the 6061 and 7075 cladded tubes processed via the ShAPE process. Metallurgical bonding is an important factor with respect to the structural integrity of the cladded tubes and is presented in the next subsection.

4.3. Mechanical Property Characterization

Both hardness and tensile property measurements were carried out on the cladded tubes and the results are presented in Figure 7 and Figure 8. First, the cladded tube from design 1–60 was subjected to a solutionizing heat treatment at 450 °C for 1 h. The temperature was selected to produce a solid solution in both alloys without introducing deleterious phases or incipient melting in the 7075 alloy. As mentioned in the Experimental Methods section, various aging heat treatments were completed to simultaneously increase the strength of the 6061 and 7075 alloys. All the aging heat treatments were taken from ASM Handbook, Volume 28, Properties and Selection of Aluminum Alloys [13]. The peak aging heat treatments for 6061 (175 °C for 8 h) and 7075 Al (121 °C for 24 h) were also selected. Hardnesses for 6061-T6 (rods, bars, and tubes) and 7075-T6 are generally 107 HV and 175 HV, respectively. After aging treatment at 121 °C for 24 h, the hardness of the 7075 region was 183 ± 4.3 HV, while the 6061 region exhibited only 70 ± 0.7 HV. On the other hand, after aging treatment at 175 °C for 8 h, the hardness of 6061 increased to 92 ± 1.3 HV, but the hardness of 7075 was 164 ± 2.0 HV compared to the previous heat treatment. As observed in Figure 6, the 7075 Al region exhibited hardness in the range of 164–183 HV under various heat treatments, and 6061 Al registered hardness ranging from 51 ± 0.7 HV to 100 ± 1.0 HV. The aging condition, 105 °C for 6 h + 175 °C for 8 h, resulted in the highest hardness of 100 ± 1.0 for the 6061 region, and the 7075 region also exhibited a hardness of about 177 ± 2.4 HV, which is similar to 7075-T6. Therefore, this post-aging heat treatment was selected as the optimum aging condition for tensile testing.

Engineering stress–engineering strain curves for both the as-extruded and the solution-treated and aged conditions are presented in Figure 8a and Figure 8b, respectively, and a summary of the tensile properties is presented in

Table 4. Summary of the tensile properties of all four cladded tubes in the as-extruded and the solution-treated and aged conditions.

Material Combination	YS (MPa)	UTS (MPa)	Total Elongation
As-extruded
Design 1–60	180 ± 2	254 ± 2	12.4 ± 0.1
Design 1–10	218 ± 3	349 ± 3	18.0 ± 1.0
Design 2–60	167 ± 2	223 ± 1.0	7.0 ± 1.0
Design 2–10	145 ± 1	241 ± 0	7.1 ± 0.1
Solution heat treated and aged
Design 1–60	270 ± 0	339 ± 2	5.4 ± 1.5
Design 1–10	281 ± 2	347 ± 5	2.8 ± 0.1
Design 2–60	208 ± 0	270 ± 3	2.5 ± 0.1
Design 2–10	202 ± 4	272 ± 9	4.7 ± 1.0

. As mentioned above, the design 1 extruded tubes have equal volume fractions of 7075 and 6061, while the design 2 extruded tubes exhibit ~67% 6061 and ~33% 7075. With respect to tensile properties in the as-extruded condition, design 1 exhibited higher strength and elongation compared to design 2. For instance, design 1–10 exhibited a yield strength (YS) of 220 MPa, an ultimate tensile strength (UTS) of 351 MPa, and a total elongation of 18.3% compared to design 2–10, which exhibited a YS of 146 MPa, a UTS of 241 MPa, and total elongation of 7.1%. A reason for design 1 exhibiting a higher YS could be the higher fraction of 7075 in the extruded tubes. And the reduced elongation in design 2 could stem from the micro-cracks noted at the interface between the 7075 and 6061 regions. Similar strength trends could be noted in the solutionized and aged condition as well. For example, design 1–10 exhibited a YS of 281 MPa, a UTS of 350 MPa, and total elongation of 2.7%, while design 2–10 exhibited a YS of 199 MPa, a UTS of 265 MPa, and total elongation of 4.7%. All the conditions showed significant work hardening and no sudden failure of the cladded tubes formed via the ShAPE process. However, the solution-treated and aged samples exhibited lower elongation than the as-extruded tubes. For instance, design 1–10 exhibited total strains of 18.3 and 2.7 in the as-extruded and aged conditions, respectively. This limited the UTS of the design 1–10 tubes in the aged condition to 350 MPa, which is similar to the extruded condition. The origin of reduced elongation in the aged tubes was not investigated in this study, but the increased flow stress of the aged tubes might have intensified the strain incompatibility and subsequently the stress concentration at the interface.

Microscopy and mechanical property characterization were carried out in the tube sections extracted from the steady-state regime. And, to understand the co-extrusion of the design 1 bimetallic billet from the start to the end, SPH modeling was used, and the results are presented in Figure 9. Figure 9a shows the SPH-predicted material flow where, at the start of extrusion, the core of the billet (7075) deforms and forms an initial tube. This is in accordance with conventional extrusion because the center of the billet deforms first and flows between the die and the mandrel. Then, a transition occurs when the extruded 7075 tube is gradually cladded by 6061. The predicted steady-state temperature distribution in Figure 9b reveals that the billet material temperature at the thermocouple location is around 380 °C, which is consistent with the experimental measurement. A temperature of around 430 °C (higher than 380 °C die temperature) was found in the billet material when it travelled through the gap between the mandrel and die throat. This is because the material will undergo additional plastic deformation at this location and generate more heat. Moreover, frictional sliding energy between the billet and mandrel can contribute to this temperature increase. Figure 9c shows the effective stress (von Mises) distribution in the billet and tube. The effective stress is in the range of 10 to ~30 MPa due to the thermally softened flow stresses, as described by the J-C model (Equation (1)). By looking at the effective stress distributions on three transverse cross sections (as indicated in Figure 9d), stress incompatibility exists between the tube core (7075) and cladding (6061) at Sections A and B. However, because the flow stresses of 7075-T6 and 6061-T6 at high temperature (380–450 °C) are similar, the stress difference between the two materials is about 10 MPa. Therefore, the observed stress difference will not cause defects, such as cracking or surface tearing, in the extruded tube. At the start of the extrusion, the von Mises stress is uniform across the cross section because of extrusion of a single material, as shown in Section C (Figure 9d).

4.4. Material Flow Analysis during Co-Extrusion via ShAPE

This is the first time a co-extrusion of bimetallic tubes has been implemented through a rotating die; therefore, an understanding of the material flow is essential to not only learn how the process works but also to use the knowledge when designing further co-extrusion experiments. The understanding of the material flow has been achieved by analyzing the remnant billet’s (unprocessed billet remaining at the end of the extrusion process) cross section and using the SPH model to predict and visualize the material flow. The pertinent results are presented in Figure 10 and Figure 11. A detailed schematic of the ShAPE process is presented in Figure 10a, showing all the key components—the rotating extrusion die, mandrel, bimetallic billet, and extruded cladded tube. An optical overview of the remnant billet cross section is presented in Figure 10b and shows the bimetallic billet deforming and thinning down to form the tube sections. The width of the 7075 part of the undeformed billet is 5 mm (displayed on the left side of Figure 10b), and has been reduced to about 0.35 mm at the start of the die throat and is shown on the right side of Figure 10b marked as the core of the tube. These measurements were made on the polished remnant puck using ImageJ. The initial billet was press-fitted and was not welded along the interface, as can be seen from the crack along the interface as marked by the solid arrow. Therefore, even though a continuous defect-free tube was produced, the cladding was broken off when the rotating extrusion die was extracted; hence, the expected material flow of the outer 6061 was drawn with the dotted line. The SPH model presents the visualization of two materials deforming and entering the die throat, as shown in Figure 9c. Based on the measurements, similar to the experiments, the 5 mm 7075 billet deformed and formed the core of the tube with a width of about 0.35–0.37 mm. Furthermore, a tapered billet material morphology during steady state is predicted by the model—similar to the experimental observations. A longitudinal section of the cladded tube is presented in Figure 10d for comparison with the SPH modeling results of the steady-state material morphology that is displayed in Figure 10e. The thickness of the core measured in the cladded tube was 0.34 ± 0.03, which is similar to the measurement at the die throat. Furthermore, the two extruding materials join in the die throat and form a defect-free cladded tube. The SPH model result along the transverse cross section, as marked in Figure 10c, is presented in Figure 10f, in which results similar to the experimental results can be observed. These results validate the SPH model and the accuracy of its predictions of the ShAPE of bimetallic tubes. With confidence that the SPH model predictions are similar to the experimental observations, the SPH model was used to further enhance understanding of the material flow.

To visualize the material flow during the ShAPE process, the model-predicted equivalent plastic strain distribution on the central slice (~1 mm thick) of the billet and tube is given in Figure 11. Simultaneously, material flow trajectories of the core (7075) and cladding (6061) in the deforming billet are presented in Figure 11a,b, respectively. Key points noted in this analysis are as follows: The billet material at the billet–die interface was plasticized more than it was in other locations because of the shear forces induced directly by die–material interaction. The tube cladding material originating from the 6061 billet cladding traveled a longer distance in the tangential direction before being extruded, while the tube core originating from the 7075 billet experienced less tangential deformation before being extruded. As a result, the tube cladding material 6061 has more plastic strain accumulation than the tube core material 7075, as indicated by the colors in the extruded tube depicted in Figure 11. Zhang et al. [23] investigated the material flow during friction extrusion by tracking particles present at various radial locations, r = 0, 3, 6, and 11 mm, from the center. The particle at the outer locations (e.g., r = 11) experienced more revolutions than the particles near the center before being extruded. The current investigation provides similar insights into material deformation and subsequent flow but yields further understanding of bimetallic extrusion.

Co-extrusion has been successful via conventional extrusion, and bimetallic tube combinations such as Al and Cu, carbon steel and stainless steel, and Al and steel have been produced [3,24]. In conventional extrusion, the heated billet assembly is pushed against the die and the bimetallic extrudate is formed as a result. The extrusion die does not rotate but remains stationary. Recently, various severe plastic deformation (SPD) techniques, such as high-pressure torsion extrusion [25,26,27], KOBO extrusion [28,29], equal-channel angular extrusion [30,31], and friction stir back extrusion [32,33], have been modified to extrude either rods or tubes. However, the co-extrusion of tubes has never been implemented using these SPD techniques, while bimetallic rods and plates have been demonstrated [34,35]. This study represents the first time co-extrusion has been implemented through a process wherein the die makes a complete rotation. Therefore, the extruding material volume undergoes rotational, radial, and longitudinal strain due to die rotation and advance into the billet. In multimaterial extrusion, the relative flow stress differences between various alloys determine the integrity of the cladded tubes. If the two materials have similar strengths at the process temperature, then both materials will flow into the die orifice and create a planar interface. However, if the outer material is weaker than the inner material, then it will not have enough effective stress to push the inner material and enter the die throat. In such cases, the co-extruded tube will predominantly be composed of the inner material. This was not the case in the current investigation. In both designs (design 1 with 50% 6061 and 7075 and design 2 with 66.7% 6061 and 33.3% 7075), the extrudates had fractions of the inner and outer materials similar to the fraction of the billet area. Overall, the initial co-extrusion via the ShAPE process was largely successful and can likely be used to fabricate other metal combinations, like Ni and Cu, as well.

5. Summary

Co-extrusion of 6061 and 7075 Al alloys was successfully implemented via the ShAPE process and the following are the key conclusions:

• The co-extruded tubes exhibited defect-free surfaces and cross sections under all four extrusion conditions. The processes were in steady state for more than 87% of the extrusion length, which demonstrates the stability of the co-extrusion process. Furthermore, the designed core-to-cladding ratio was realized in all four cladded tubes, which shows the robustness of co-extrusion despite having a rotating extrusion die.
• The interface between 6061 and 7075 was metallurgically bonded in all the extrudates; more importantly, the extrudates exhibited defect-free interfacial regions. Obtaining a defect-free tube is a first step toward fabricating cladded tubes that have structural integrity. The cladded extrudates exhibited a strong work-hardening regime, which indicates the ability of these tubes to sustain significant loads before failure. An optimized heat treatment of 105 °C for 6 h + 175 °C for 8 h was found to lead to increased hardness and tensile strength due to age hardening of the 6061 and 7075 alloy regions.
• Material flow analysis based on the SPH model revealed that the cladding material travels a longer distance than the core material before being extruded. Furthermore, no material mixing other than the metallurgically bonded interface across the core and cladding was noted in both experiments and SPH modeling.

Overall, based on current observations ShAPE can successfully produce cladded tubes using various material combinations including high-temperature materials.