Journal Information
Vol. 8. Issue 6.
Pages 5813-5822 (November - December 2019)
Share
Share
Download PDF
More article options
Visits
...
Vol. 8. Issue 6.
Pages 5813-5822 (November - December 2019)
Original Article
DOI: 10.1016/j.jmrt.2019.09.050
Open Access
Wetting of liquid aluminum alloys on pure titanium at 873-973 K
Visits
...
Peng Jina,c, Yibo Liua,c,
Corresponding author
ybliu0701@hit.edu.cn

Corresponding authors at: State Key Laboratory of Advanced Welding and Joining, Harbin Institute of Technology, No. 92 West Dazhi Street, Harbin 150001, China.
, Qingjie Suna,c,
Corresponding author
qjsun@hit.edu.cn

Corresponding authors at: State Key Laboratory of Advanced Welding and Joining, Harbin Institute of Technology, No. 92 West Dazhi Street, Harbin 150001, China.
, Qiaoli Linb, Junzhao Lia,c, Kun Chenc, Jianfeng Wanga,c, Shaojun Houa, Fuxiang Lia,c, Jicai Fenga,c
a State Key Laboratory of Advanced Welding and Joining, Harbin Institute of Technology, No. 92 West Dazhi Street, Harbin 150001, China
b State Key Laboratory of Advanced Processing and Recycling of Non-ferrous Metal, Lanzhou University of Technology, Lanzhou 730050, China
c Shandong Provincial Key Laboratory of Special Welding Technology, Harbin Institute of Technology at Weihai, Weihai 264209, China
Article information
Abstract
Full Text
Bibliography
Download PDF
Statistics
Figures (9)
Show moreShow less
Tables (2)
Table 1. Chemical compositions of the used materials (wt. %).
Table 2. Physical parameters used to fitting the dynamic models.
Show moreShow less
Abstract

The wetting behaviour of Al alloys (i.e., 6061 Al and 4043 Al alloys) on the surface of pure titanium was studied by sessile drop method at temperatures of 873–973K under high vacuum with ˜10−4 Pa. All wetting processes were divided into two stages: in the nonlinear stage, the formation of a compact intermetallic layer (Ti7Al5Si12) blocked the mass transfer at solid/liquid (S/L) interface and deteriorated wetting, while the formation of loose phase (Al3Ti) at triple line region can removed the oxide film and improved wettability. The wetting activation energy (56kJ/mol for a 6061Al/Ti system, and 71kJ/mol for a 4043Al/Ti system) corresponded to the sum of the difference in surface energy per molar area between α-Ti and TiO2 (27.2kJ/mol), and the Gibbs energy of formation for Al3Ti (30kJ/mol) or the decomposition energy for Ti7Al5Si12 (48kJ/mol). In the near-linear stage, the droplet spread on the surface of intermetallic compound (IMC). The spreading dynamics for 6061 alloy satisfied the Jiang’s empirical equation, and for 4043 alloy satisfied the viscous dissipation-limited model. The interfacial reaction had little effected on a droplet spreading, while the viscous friction drove the spread of 4043 alloy on precursor film in the near-linear stage.

Keywords:
Interfacial mass transfer
Reactive wettability
Activation energy
Al-Ti
Full Text
1Introduction

Aluminum-titanium components have been widely used in petrochemical components, as biocompatible materials, and in aircraft engines due to their light-weight nature, oxidation resistance, and corrosion resistance [1,2]. Developing such aluminum-titanium components with excellent tensile strength allows their substitution for aluminum, as used in high-speed wind tunnels and propulsion exhaust systems to improve corrosion and oxidation resistance at high temperatures [3]; however, the joining process requires lower heat input to avoid residual stress because of their weakly metallurgical compatibility and the mismatch in chemical affinity [4]. Besides, the brittle Ti-Al IMCs are considered to be a source of cracking which causes joint fracture [5]. To decrease the heat input and avoid the formation of thick Ti-Al IMCs at Ti/Al joints, many researchers pointed out that the addition of trace elements can modify the phase compositions, especially for the Mg and Si [6,7]. Chen et al. [7] reported on the microstructure of Ti to Al joining by laser, and found that the heat input was decreased, the Ti-Al IMCs layer thinned and the strength of the joint increased; Si element segregation at solid-liquid interface affected the microstructure and tensile strength. Gao et al. [8] described how the hybrid heat sources was key to determining the joint strength rather than the type of groove in a Ti-Al joining process using laser-arc hybrid welding. Li et al. [1] revealed the addition of elemental Mg into Al filler wire can significantly improve the joint strength and further inhibit the precipitation of Al-Ti IMCs. Cheng et al. [9] studied the vacuum brazing of Ti and Al, and indicated that with the temperature increasesing, the wettability of solder was gradually improved. The Mg and Si elements can effectively reduce the thickness of Ti-Al IMCs, and improve joint strength. Besides, trace amounts of Mg can reduce the Ti2O and Al2O3 oxide film [10], and the addition of Si can contribute to a good flowability because of the surfactivity of Si in molten Al-Si alloy [11]. Therefore, suppressing the formation of Ti-Al IMCs and optimising the interfacial structure are of great concern to improving joint strength. However, heat input was not suited to being decreased to a certain value given the poor wetting behaviour and joint appearance of Ti/Al joints, limiting the further improvement of their joint properties. The intrinsic mechanism of interfacial microstructure regulation and joint formation is closely related to wetting and interfacial reaction behaviour of the Ti-Al metallurgical system with small amount of Mg and Si thus justifying this work.

Recently, some researchers have conducted preliminary studies of the wetting behaviour of molten metal on the surface of Ti plate. Liu et al. [12] reported on the wettability of liquid Ti32.8Zr30.2Ni5.3Cu9Be22.7 on the surface of TC4, and indicated that as the temperature increases, the improvement of wettability was related to the reduced viscosity and surface tension of molten metal, besides the interfacial reaction could promote the wettability. Tashi et al. [13] found a temperature dependence in the wetting of AgCuZn to TC4. Lin et al. [14] suggested that the spreading dynamics of an Al drop on steel plates was independent of the interfacial reaction and determined by the characteristics of droplet itself. The wettability and spreadability were gradually improved with the increase of heat input and the capillary force was the factor limiting droplet spreading. Zhou et al. [15] investigated the wetting of 4043 Al alloy on galvanized steel, and found that the spreading of liquid Al was limited by the viscosity. Li et al. [16] claimed that the oxide film blocked the moving of triple line, while the interfacial reaction could remove the oxide film, making the surface of substrate to be more metallic, and enhance the affinity between droplet and metallic substrate. Eustathopoulos et al. [17] introduced a reaction product control model for a metal/ceramic system, and claimed that the production of intermetallic compounds can effectively promoted wetting. Satyanarayana et al. [18] proposed a reaction product obstruct wetting theory for a Sn-0.3Ag-0.7Cu/Al reactive wetting system, and indicated that the reaction products were unfavorable to interfacial mass transfer and inhibited wetting. Satoshi et al. [19] reported the wetting behaviour of molten brazing filler metallic alloys on metallic substrate: the production of intermetallic compounds almost had not effect on wettability. In conclusion, interface will be inevitably produce intermetallic compound in the reactive wetting of metal/metal system. However, it is still controversial and there is no unified conclusion that the intermetallic compound and morphology have effect on the wettability. Meanwhile, these researchers indicate that wettability is associated with temperature and intrinsic properties of liquid metals, they also cannot establish a relationship between wettability and the interfacial mass transfer. Besides, the effect of Mg, Si elements and interfacial microstructure evolution remains obscure. For these reasons, the wetting and interfacial microstructure evolution warrant further study.

In this study, re-melt 6061 and 4043 alloys were selected: 6061 and 4043 are filler materials used when joining Al and Ti alloys. The spreading dynamics, interfacial microstructures and wettability of these two alloys on the pure Ti substrate were studied. To further reveal the intrinsic mechanism of interfacial compound evolution, the microstructures of 6061 and 4043 Al/Ti interface were analyzed. Besides, the spreading dynamic models were adopted to reveal the limited factors of wetting in the two systems.

2Experimental procedure

TA2 pure titanium was selected as the base metal material with dimensions of 20×20×2mm, 6061 and 4043 filler wire were re-melted to cubes with 2×2×2mm for the later testing. The nominal chemical compositions of pure Ti, 6061 and 4043 alloys are summarised in Table 1. Prior to the experiment, the oxide film on the surface of Al alloy cubes were cleaned using an automatic polishing machine. The average roughness of the polished TA2 substrates were ˜40nm, as measured by DEKTAK 6M (Veeco Metrology Corp., Woodbury, NY, USA) over a distance of 2mm at a speed of 100mm/s. Then the TA2 substrate, 6061 and 4043 alloys cube block were immersed in a beaker of alcohol and ultrasonically cleaned for 10min.

Table 1.

Chemical compositions of the used materials (wt. %).

Material  Fe  Si  Mn  Mg  Zn  Ti  Cu  Al 
6061 Al  –  0.70  0.60  0.15  0.90  0.25  0.15  0.15  Balance 
4043 Al  –  0.80  5.00  0.05  0.05  0.10  –  0.305  Balance 
TA2 Ti  0.01  0.30  0.15  –  –  –  Balance  –  – 

A modified sessile drop method [20] (when plate is heated to the desired experimental temperature, a magnetic push-rod moves the metal to the surface of the plate) was used to study wettability of an Al/Ti system at high temperatures, which can avoid the reaction between the droplet and plate in the heating process compared to the traditional sessile drop method. The wetting experiment was performed using an advanced LWAT/H-1 wetting system, and the measurement accuracy of contact angles reached±0.1 degree. Prior to the wetting experiment, TA2 substrate was placed at the horizontal ceramic heating zone in the chamber of a wetting furnace. The chamber was evacuated to a high vacuum state, and the chamber was heated at a rate of 20K/min to the experimental temperature. When the experimental temperature was stable, an Al cube was dropped onto the surface of Ti substrate. During the experiment, the clear profiles of the molten drop were obtained by CCD camera (at a laser backlight wavelength of 650nm).

After the experiment, ultra-high-purity argon atmosphere (99.999%) was pumped into the chamber as a shielding gas until the specimens were cooled to ˜303K. The wetting samples were cross-sectioned perpendicular to the radial direction for metallographic analysis. Backscattered electron image observation and quantitative compositions identification of the interfacial microstructure were conducted by scanning electron microscope (SEM, MERLIN Compact) equipped with an energy-dispersive X-ray spectrometer (EDS). The phases were identified by X-ray diffraction (XRD, D8 Advance).

3Results3.1Wetting phenomena

Fig. 1 shows the variation of contact angles and normalized contact radii R/R0 (where R is the dynamic contact radius, and R0 is initial contact radius) with time for the molten Al alloys on the pure Ti plate. The contact angles decreased with time from 46±0.1° for 6061Al and 121±0.1° for 4043Al to stable equilibrium contact angles. Liquid 4043Al alloy was non-wetting the Ti plate at 873K, while the 6061 Al alloy had excellent wettability (Fig. 1(a)). R/R0 varied with time monotonously, and there were two stages to the spreading process of the whole Al alloys (Fig. 1(b)). Non-linear spreading was presented in the first stage, and then the near-linear spreading was observed. The variation of R/R0 to the non-linear spreading can be fitted by R/R0=R/Ri-mexp[−(t/k)], where Ri is initial contact radius, m and k are fitting parameters [21]. The precursor film was produced in a near-linear stage, and the spreading rate in a 6061Al/Ti system was much better than a 4043Al/Ti system. The different initial wettability and spreading dynamics in these two systems caused by the Mg and Si elements in Al alloy.

Fig. 1.

The variations with time in the (a) dynamic contact angles and (b) normalized contact radius R/R0 for molten 6061 and 4043 alloys on pure Ti surface at 873K, 923K, and 973K.

(0.38MB).
3.2Macro-morphologies

The macro-morphologies of the 6061 and 4043 alloys at pure Ti plates were showed in Fig. 2. 6061Al alloys exhibited good wettability and the whole substrate was covered by the precursor film at 973K (Fig. 2(c)). At the same experimental temperature, the final wettability for 6061 alloy was much better than 4043 alloy, and the width of precursor film was larger. At the triple line region, the precursor film clearly showed two layers. The extended rate of precursor film was much faster than the movement of triple line, and then the Al droplet spreading on the surface of precursor film (reaction layer, i.e. inert surface) rather than the pure Ti. Therefore, the spreading dynamics was different from the “chemical reaction control” model in near-linear stage.

Fig. 2.

Macro-morphologies after isothermal wetting at different temperatures, (a)-(c) 6061Al/Ti and (d)-(f) 4043Al/Ti at 873K, 923K and 973K, respectively.

(0.6MB).
3.3Microstructures

Fig. 3 shows the cross-sectional view and top-view of the microstructures at the triple line region, and the interfacial microstructure at the central position for 6061Al/Ti after isothermal wetting at 873K. Precursor film exceeded the triple line with respective characteristics can be observed (Fig. 3(a)). The spreading rate of the precursor film was much faster than a molten droplet, and two different morphologies could be observed: the under-layer produced a continuous intermetallic compound layer, and the top-layer was residual liquid Al. Actually, the precursor film was the extended reaction layer. To further observe the morphology of the precursor film, top-view of triple line were selected, as the details showed in Fig. 3(c). Some acicular phase was segregated at the surface of precursor film after solidification. The spreading rate of precursor film for under-layer was slightly higher than a top-layer (Fig. 3(d)). Seen in combination with EDS and elemental surface distribution analysis, the under-layer was mainly consisted of Al3Ti, and the top-layer was residual liquid Al (Fig. 3(e)-(j)). The cross-sectional view of the interfacial structure at the central position was selected, as be showed in Fig. 3(b), and the continuous reaction layer beneath the drop can be observed. Some granular phase evenly dispersed into the droplet, points EDS analysis shown that the chemical composition of the continuous reaction layer was Al3Ti, and small amount of granular Al3Ti was distributed in the droplet near the interface zone. The results were showed in Fig. 3(h)-(j), which is corresponding to the positions of yellow and red crosses in Fig. 3(b). It was noteworthy that Mg was not found either in the precursor film or in the interface region, while Si segregation in the reaction layer was obvious. The concentration of Si far exceeded the original value (0.6wt. % in Table 1) at reaction layer. Also, the element line distribution suggested the Si segregation in the reaction layer but no Mg track (Fig. 3(b)).

Fig. 3.

SEM images of the 6061Al/Ti system after isothermal wetting at 873K, (a) cross-sectional view of the microstructures near the triple line region, (b) at central position of the interfacial microstructures with the elemental line distribution results, (c) top-view near the triple line region, (d) the details for the black rectangle in (c), (e)-(g) the elemental surface distribution results for the red rectangle in (c); (h)-(g) EDS results for the corresponding yellow, red and blue crosses in (b) and (d), respectively.

(0.95MB).

Fig. 4 shows the cross-sectional view and top-view microstructures for 4043Al/Ti interface and triple line region after isothermal wetting at 873K and 925K. At 873K, the reaction layer was continuous, and denser in Ti side (bottom layer) with thickness of ˜ 3μm, the virgate and granular phase were uniform distribution in the upper layer near the droplet (Fig. 4(a)-(b)). At 925K, the bottom layer and virgate phase disappeared at central position, while a small amount of virgate phase also remained in the continuous interface layer near the triple line region (Fig. 4(c)-(d)). Based on the element line distribution and point analysis by EDS in Fig. 4(b)-(d) and (i)-(k), the colors in Fig. 4(i)-(k) were corresponding to the color of crosses in Fig. 4(b)-(d), suggested that granular phase was Al3Ti with Si-rich, virgate phase and bottom layer were the Ti7Al5Si12, the concentration of which was surprisingly high in the reaction layer almost 8–9 times than the original concentration in 4043 alloy which suggested the Si was an interfacial active element. The precursor film also extended reaction layer with some residual liquid Al, the reaction layer was not so compact that some residual liquid Al permeated into the layer (Fig. 4(e)-(h)). The precursor film also had two layers, the bottom layer was Al3Ti and the top layer was residual liquid Al with some granular phase, elemental surface distribution clearly showed the two layers (Fig. 4(h)). The wetting temperatures were marked by the red lines in Fig. 5. The solubility of Ti in Al was limited by the experimental temperature. Therefore, the intermetallic compound could quickly produce, and then further resisted interfacial reaction.

Fig. 4.

SEM images of the 4043Al/Ti system after isothermal wetting at 873K and 923K, (a) cross-sectional view of the microstructures near the triple line region at 873K, (b) at central position of the interfacial microstructures with the elemental line distribution results at 873K; (c) cross-sectional view of the microstructures at close of triple line at 923K; (d) at central position of the interfacial microstructures with the elemental line distribution results at 923K, (e) cross-sectional view of the microstructures at triple line region at 923K, (f) top-view at the triple line region, (g) the details for the black rectangle in (f), (h) the elemental surface distribution results for the red rectangle in (f); (i)-(k) EDS results for the corresponding red, blue and yellow cross in (a)-(d), respectively.

(0.87MB).
Fig. 5.

Binary phase diagram of Al-Ti system [8].

(0.18MB).

Fig. 6 shows the macro-morphologies and SEM images after corrosion (the solidified Al was removed by 0.5mol/L sodium hydroxide aqueous solution) for 6061 and 4043Al/Ti samples at 923K. The microstructures of interfacial production for 6061Al/Ti samples were different from that 4043Al/Ti, and the microstructures for beneath the drop were different from that at close to the triple line. A lot of loose granular phase and residual Al at beneath the 6061 alloy drop, while the granular phase were smaller and denser at triple line region, besides the virgate phase disappeared in all range of solid-liquid interface (Fig. 6(b)-(c)). At beneath the 4043 alloy drop, the interface distributed regular granular phase and the size was about 3 times than that of 6061 alloy sample; the triple line region also distributed dense granular phase with small amount of acicular phase, as shown in Fig. 6(b)-(c), (e)-(f). By comparing to the XRD spectrum of original Ti surface, the main phases in reaction layer were Al3Ti and some residual Al, and the precursor film was constituted of Al3Ti and residual liquid Al for 6061Al/Ti sample. However, the main phase in the precursor film were Al3Ti and some Ti7Al5Si121) for 4043Al/Ti sample, were showed in Fig. 7 (corresponding from a to d, respectively), which were in according with the EDS results.

Fig. 6.

Macro-morphologies and SEM images of the 6061 and 4043 wetting samples after Al was removed by NaOH aqueous solution at 923K, (a) macro-morphology for a 6061Al/Ti sample; (b) and (c) are the corresponding details at beneath the drop and triple line region, respectively; (d) macro-morphology for a 4043Al/Ti sample; (e) and (f) are the corresponding details at beneath the drop and triple line region, respectively.

(0.52MB).
Fig. 7.

XRD patterns of the phases at the precursor films for (a) 6061 and (b) 4043 samples after isothermal wetting at 923K; XRD patterns of the phases for the exposed interface for (c) 6061Al/Ti and (d) 4043Al/Ti after isothermal wetting at 923K through removing of the solidified Al drop, and (e) for the original surface of TA2.

(0.14MB).
4Discussion4.1Wetting mechanism and spreading dynamics for nonlinear stage

In this Al/Ti reactive wetting system, the wettability and spreadability are determined by many factors, such as the presence of an oxide film, interfacial reaction, a small amount of solute elements, etc. The oxide film as one of main factor, the polished surface of Ti plate was covered by the TiO2 oxide film of about ˜2nm [22], blocking triple line movement. In this 6061Al/Ti system, Mg vapor which around the drop can reduce the thickness of this TiO2 oxide film according to thermodynamic data (2Mg(g)+TiO2(s)=2MgO(s)+Ti(s), ΔrGm0=−485300+208.3T, J·mol−1), exposing the Ti atoms to the substrate surface, thus increasing the surface energy of Ti substrate at triple line region, then the liquid Al would infiltrate the crack of oxide film and formation a precursor film, triggering the triple line movement and promote interfacial reaction. Besides, the formation of loose phase (Al3Ti) at triple line region can also remove the oxide film, and then improve wettability. In 4043Al/Ti system, 4043 alloy had weak wettability on the pure Ti surface at 873K, but the interface produced intermetallic compound with the continuous compact Ti7Al5Si12 layer in the bottom (Fig. 4). The solubility of Ti in liquid Al is rather small and the continuous compact reaction layer (Ti7Al5Si12) would block the mass transfer of S/L interface at triple line region, and then restrain wetting and spread. The primary reaction product Ti7Al5Si12 decomposed at 852K [23]. However, high concentration of Si inhibited the decomposition of Ti7Al5Si12 at 873K. Once the continuous compact Ti7Al5Si12 layer was formed, the mass transfer at close triple line would further slowdown, and deteriorate wettability. On the contrary, 4043 alloy had a good wettability on the surface of Ti plate at 923K, the bottom layer and virgate phase had disappeared, the loosened layer was consisted of the granular Al3Ti phase. The decomposition of Ti7Al5Si12 and production of Al3Ti (loosened structure) removed the oxide film and promoted mass transfer at triple line region, improving the wettability. Based on the above considerations, the formation of Ti7Al5Si12 at L/S interface, as a compact intermetallic layer block the mass transfer and deteriorate wettability. The formation of a loose phase (Al3Ti) at triple line zone would remove the oxide film and promote wettability.

As mentioned above that Si segregation at L/S interface can be found but no Mg track, the active element segregation at L/S interface is a prerequisite for the precipitation of reaction product when the concentration of the active elements is below the saturation precipitation concentration. Therefore, adsorption energy of Si and Mg at Al-Ti interface can be used for the prediction of Al-Ti interfacial structure before experiment, which can be using the following equations [24,25],

where  ESi(Al)∞SL and EMg(Al)∞SL are adsorption energy of Si and Mg at Al-Ti interface, m1 is a positive interfacial structure factor, λijis the mole exchange energy for i and j, △H¯i(j)∞ and △H¯j(i)∞ represent the mixing enthalpy of i in j and j in i. The more negative the value of adsorption energy for Si and Mg, which is more inclined to segregate at Al-Ti interface.  ESi(Al)∞SL/m and EMg(Al)∞SL/m are -73.4kJ/mol and 221.4kJ/mol for Al-Si/Ti system and Al-Mg/Ti system, respectively. Therefore, Si is segregated at L/S interface.

In nonlinear stage, the IMCs were produced at L/S interface, and the reaction mechanism influence the spreading dynamics. The spreading dynamics of this system have typical the characteristics of reaction product control (RPC) limited model [17,24]. The wetting activation energy can be obtained based on RPC fitting equation,

where the θe, θ, and θ0 are the equilibrium, dynamic and initial contact angles, respectively. k1 is a kinetic constant, and can be deduced from the fitting results based on the variations of θ with t in Fig. 1(a). The wetting activation energy of 6061 and 4043Al/Ti systems was further deduced from Arrhenius plots of k1 and 1/T (Fig. 8), with values of 56kJ/mol and 71kJ/mol. For a 6061Al/Ti system, the activation energy (56kJ/mol) is associate with the Al3Ti compound formation (30kJ/mol) and the surface energies in difference per molar area between α-Ti and TiO2 (27.2kJ/mol). The per molar area surface energies are the 1650mJ/m2[26] and 850mJ/m2[27] for Ti and TiO2, respectively, molar surface area is 33975.4m2/mol for α-Ti. Further, we note that the Al3Ti compound formation can reduce the TiO2 oxide film, which may be why the values of the wetting activation energy was slightly smaller. Therefore, the removing of the oxide film and interfacial reaction mainly contributed mainly to the spreading forces. For a 4043Al/Ti system, Ti7Al5Si12 decomposition is key factor to promote wetting. Besides, production of Al3Ti (loosened structure) removed the oxide film and promoted mass transfer at triple line region, improving the wettability, therefore the activation energy (71kJ/mol) is corresponding to the Ti7Al5Si12 decomposition energy (48kJ/mol) and the surface energies in difference per molar area between α-Ti and TiO2 (27.2kJ/mol).

Fig. 8.

Arrhenius plots of the kinetic constant k1 for nonlinear spreading stage.

(0.14MB).
4.2spreading dynamics for near-linear stage

Significantly, the precursor film occurred at the end of nonlinear stage and the spreading rate of the precursor film was much greater than the droplet. Besides, the precursor film was consisted of the granular phase Al3Ti and residual liquid Al at the surface of former (Fig.7), proved that the continuous granular phase prevents the mass transfer. When the precursor film precipitated, actually the droplet further spread on the precursor film rather than pure Ti surface. The spreading dynamic may be different from the RPC model at triple line region in nonlinear stage, and the interfacial reaction is weak. Therefore, inert forces should be considered as the factor of limited spreading. In order to confirm the limited factors of control the droplet spreading, some other factors should be taken into consideration, such as adsorption/desorption in the triple line region, density, viscosity, surface/interfacial tension etc., and then three typical inert wetting dynamic models are introduced.

The adsorption/desorption of fluid elements in the triple line region maybe affect the droplet spreading for the molecular dynamic model. The relationship of U (spreading rate) and  θ (dynamic contact angle) as following [28,29],

Where Ke is fitting parameter, k is distance between two adsorption or desorption sites, KB is the Boltzman constant, σlv is L/V surface energy. The spreading rate U as showed in Fig. 2. The fitting of U and θ to this model shows a poor fitness, with R square are the 0.479-0.581, 0.427-0.615 for 6061 and 4043 at 873K–973K, respectively. Therefore, adsorption/desorption of fluid elements in the triple line region is not the limiting factor affecting spreading in this system.

Jiang’s empirical formula incorporate both the viscous dissipation-limited and molecular dynamic effects, the factor of affect spreading maybe associate with density, viscosity, surface/interfacial tension etc., and it can be expressed as following [30,31],

where σlv and μ are the surface tension and liquid viscosity, μ/σlv is a fitting parameter. The fitting parameter μ/σlv are 1.61×10−3, 1.49×10−3 and 1.26×10−3 for the 6061Al/Ti, at 873K, 923K and 973K, respectively. Physical parameters were shown in Table 2, substituting this value into the above fitting parameter, μ/σlv can be calculated based on the physical parameters which are 1.5×10−3, 1.4×10−3 and 1.3×10−3, respectively. Results show an excellent fitness for the 6061Al/Ti system, the fitting value is close to the actual value. However, 4043Al/Ti wetting systems shows a rather poor fitness (Fig. 9(b)).

Table 2.

Physical parameters used to fitting the dynamic models.

T, K  σlv, mN/m [26] 6061 Al  σlv, mN/m [33] 4043 Al  μ, mN s/m2[26]  ρ, kg/m3[26] 
873  925  912  1.41  2471 
923  914  903  1.30  2385 
973  900  891  1.18  2245 
Fig. 9.

Plots of θ (dynamic contact angle) versus U (velocity) and (cosθe-cosθ)/(cosθe+1) versus U fitted by (a) hydrodynamic model and (b), Jiang’s empirical formula respectively.

(0.41MB).

In the hydrodynamic (namely, viscous dissipation-limited) model, the viscosity is the main factor for limitting the spreading. The relationship of dynamic contact angle (θ) and spreading rate (U) as following [32],

where L and Ls are the capillary length and slip lengsth, respectively. σlv, μ, ρ, g are the surface tension, liquid viscosity, density and the gravitational constant, respectively. For convenience of fitting, we defined 9(μ/σlv)ln(L/Ls) as a fitting parameter. The fitted results show an excellent fitness for the 4043Al/Ti wetting system, while the 6061Al/Ti wetting system show a poor fitness (Fig. 9(a)). The fitting paramaters (9(η/σlv)ln(L/Ls)) are 1.27×104 and 1.18×104 for the 4043Al/Ti at 923K and 973K, respectively. Physical parameters were shown in Table 2. Ls can be calculated based on the 9(μ/σlv)ln(L/Ls) and in the order of ˜ 10−9 m. Ls is the order of atomic or molecular dimension, ˜ 10−9–10−8, implied the spreading was more inclined to the dependence of liquid viscous [32]. Therefore, in the near-linear stage, the spreading of 4043 alloy on precursor film was mainly determined by the viscous friction.

5Conclusions

The wetting of pure Ti by Al alloys (6061 and 4043 Al) was investigated by modified sessile drop method at 873–973K. The main conclusions are listed as follows:

  • (1)

    The wetting dynamic was determined by the reaction products at liquid/solid interface in the nonlinear stage. The formation of the Ti7Al5Si12 at solid-liquid interface, as a compact intermetallic layer blocked the mass transfer and deteriorate wettability. The formation of a loose phase (Al3Ti) at triple line zone could remove the oxide film and promote wettability.

  • (2)

    In nonlinear stage, the spreading dynamics can be described by RPC model, with the wetting activation energies are 56kJ/mol and 71kJ/mol for 6061 and 4043Al/Ti system. The activation energy (56kJ/mol and 71kJ/mol) was associated with the Al3Ti compound formation (30kJ/mol) or Ti7Al5Si12 decomposition (48kJ/mol) and the surface energies in difference per molar area between Ti and TiO2 (27.2kJ/mol).

  • (3)

    In near-linear stage, the droplets spread on the surface of intermetallic compounds. The spreading dynamic of 6061 alloy satisfies the Jiang’s empirical equation and 4043 alloy satisfies the viscous dissipation-limited model.

Conflicts of interest

There is no conflict of interest in this paper.

Acknowledgement

This project is supported by the National Key Research and Development Programme of China (2016YFB0300602) and the National Natural Science Foundation of China (Grant No. U1960102, 51705103, 51475104).

References
[1]
Q.J. Sun, J.Z. Li, Y.B. Liu, et al.
Microstructural characterization and mechanical properties of Al/Ti joint welded by CMT method-assisted hybrid magnetic field.
Mater Des, 116 (2017), pp. 316-324
[2]
F. Appel, U. Brossmann, U. Christoph, et al.
Recent progress in the development of gamma titanium aluminide alloys.
Adv Eng Mater, 2 (2010), pp. 699-720
[3]
F. Appel, M. Oehring, J.D.H. Paul, et al.
Physical aspects of hot-working gamma-based titanium aluminides.
Intermetallics, 12 (2004), pp. 791-802
[4]
H. Okamoto. Al-Ti (aluminum-titanium) Journal of Phase Equilibria, 14, 1993, pp. 764-764.
[5]
Y. Chen, S. Chen, L. Li.
Effects of heat input on microstructure and mechanical property of Al/Ti joints by rectangular spot laser welding-brazing method.
Int J Adv Manuf Technol, 44 (2009), pp. 265-273
[6]
W.V. Vaidya, M. Horstmann, V. Ventzke, et al.
Improving interfacial properties of a laser beam welded dissimilar joint of aluminium AA6056 and titanium Ti6Al4V for aeronautical applications.
J Mater Sci, 45 (2010), pp. 6242-6254
[7]
S.H. Chen, L. Qun, Y.B. Chen, et al.
Si diffusion behavior during laser welding-brazing of Al alloy and Ti alloy with Al-12Si filler wire.
Trans Nonferrous Met Soc China, 20 (2010), pp. 64-70
[8]
M. Gao, C. Chen, Y. Gu, et al.
Microstructure and tensile behavior of laser arc hybrid welded dissimilar Al and Ti Alloys.
Materials, 7 (2014), pp. 1590-1602
[9]
S.Y. Chang, L.C. Tsao, T.Y. Li, et al.
Joining 6061 aluminum alloy with Al-Si-Cu filler metal.
J Alloys Compd, 488 (2009), pp. 174-180
[10]
M. Gelbstein, N. Froumin, N. Frage.
Wetting behavior of liquid Fe-C-Ti alloys on sapphire.
Mater Sci Eng A, 495 (2008), pp. 159-163
[11]
Y. Hyogo, H. Momosaki, K. Tohma.
Fillability of Al-Si brazing sheet using fluoride flux.
Japan Inst Light Metals, 47 (1997), pp. 655-660
[12]
C.C. Liu, C.L. Ou, R.K. Shiue.
The microstructural observation and wettability study of brazing Ti-6Al-4V and 304 stainless steel using three braze alloys.
J Mater Sci, 37 (2002), pp. 2225-2235
[13]
R.S. Tashi, S.A.A.A. Mousavi, M.M. Atabaki.
Diffusion brazing of Ti-6Al-4V and austenitic stainless steel using silver-based interlayer.
Mater Des, 54 (2014), pp. 161-167
[14]
Q. Lin, C. Zeng, R. Cao, et al.
The spreading simulation of molten Al alloy on Q235 steel in the first cycle of cold metal transfer process.
Int J Heat Mass Transf, 96 (2016), pp. 118-124
[15]
Y. Zou, Q. Lin.
Wetting of galvanized steel by Al 4043 alloys in the first cycle of CMT process.
J Alloys Compd, 589 (2014), pp. 307-313
[16]
Y. Li, W. Liu, Dusan, et al.
Reactive wetting of AgCuTi filler metal on the TiAl-based alloy substrate.
Appl Surf Sci, 259 (2012), pp. 343-348
[17]
O. Dezellus, F. Hodaj, N. Eustathopoulos.
Progress in modelling of chemical-reaction limited wetting.
J Eur Ceram Soc, 23 (2003), pp. 2797-2803
[18]
K.N. Satyanarayana, Prabhu.
Study of reactive wetting of Sn-0.7Cu and Sn-0.3Ag-0.7Cu lead free solders during solidification on nickel coated Al substrates.
International Journal of Chemical, Molecular, Nuclear, Materials and Metallurgical Engineering, 7 (2013), pp. 25-28
[19]
K. Satoshi, K. Tetsurou, Yukiakira.
Wetting and spreading behavior of molten brazing filler metallic alloys on metallic substrate.
Mater Sci Eng, 61 (2014), pp. 012-017
[20]
P. Shen, Q. Lin, Q.C. Jiang, et al.
Reactive wetting of polycrystalline TiC by molten Zr55Cu30Al10Ni5 metallic glass alloy.
J Mater Res, 24 (2009), pp. 2420-2427
[21]
L. Gremillard, E. Saiz, V.R. Radmilovic, et al.
Role of titanium on the reactive spreading of leading-free solders on alumina.
J Mater Res, 21 (2006), pp. 3222-3233
[22]
Q. Sun, P. Jin, Y.B. Liu, et al.
Wetting of liquid copper on TC4 titanium alloy and 304 stainless steel at 1273-1433 K.
Mater Des, 169 (2019), pp. 107667
[23]
R. Voytovych, F. Robaut, N. Eustathopoulos.
The relation between wetting and interfacial chemistry in the CuAgTi/alumina system.
Acta Mater, 54 (2006), pp. 2205-2214
[24]
N. Eustathopoulos, M.G. Nicholas, B. Drevet.
Wettability at high temperatures.
Elsevier, (1999),
[25]
R.F. Zhang, S.H. Sheng, B.X. Liu.
Predicting the formation enthalpies of binary intermetallic compounds.
Chem Phys Lett, 442 (2007), pp. 511-514
[26]
W.F. Gale, T.C. Totemeier.
Smithells metals reference book.
seventh ed., Elsevier, (1998),
[27]
J. Humenik, M. Parikh.
Metal-Ceramic Interactions: III, Surface tension and wettability of metal-ceramic systems.
J Am Ceram Soc, 37 (1954), pp. 18-23
[28]
Y. Zhou, Q. Lin.
Wetting of galvanized steel by Al 4043 alloys in the first cycle of CMT process.
J Alloys Compd, 589 (2014), pp. 307-313
[29]
T. D. Blake, J. Haynes. M. Kinetics of [formula omitted] displacement Journal of Colloid and Interface Science, 30(3), pp. 421-423.
[30]
Z.K. Li, G.F. Ma, H.M. Fu, et al.
The spreading kinetics and precursor film characteristics of Zr-based alloy melt on W substrate.
Mater Lett, 98 (2013), pp. 98-101
[31]
T.S. Jiang, O.H. Soo-Gun, J.C. Slattery.
Correlation for dynamic contact Angle.
J Colloid Interface Sci, 69 (1979), pp. 74-77
[32]
S.F. Kistler.
Hydrodynamics of wetting.
Wettability, pp. 311-429
[33]
P. Shen, H. Fujii, K. Nogi.
Wettability of R (011-2) single crystalline and polycrystalline α-Al2O3 substrates by Al-Si alloys over wide composition and temperature ranges.
J Mater Sci, 41 (2006), pp. 7159-7164
Copyright © 2019. The Authors
Journal of Materials Research and Technology

Subscribe to our newsletter

Article options
Tools
Cookies policy
To improve our services and products, we use cookies (own or third parties authorized) to show advertising related to client preferences through the analyses of navigation customer behavior. Continuing navigation will be considered as acceptance of this use. You can change the settings or obtain more information by clicking here.