Numerical Simulation of the Hot Isostatic Pressing Densification Behavior of Ti6Al4V Powder for a Thin-Walled Tubular Component with Non-Axisymmetric Inner Ribs

Abstract

Hot isostatic pressing (HIP) technology is an efficient near-net-shape forming method to prepare complex-shaped structural components. However, for non-axisymmetric components with a complex shape, the powder flow and densification behaviors during HIP are still not clear, leading to a need for lots of experiments to optimize the process parameters. In the current work, a typical aerospace thin-walled tubular component with non-axisymmetric inner ribs was selected as the research object, and its instantaneous powder flow and relative density during the whole HIP process were investigated by a numerical simulation method, focusing on the influence of HIP process conditions on powder densification. The simulation results indicate that the upper end of the Ti6Al4V thin-walled tubular part is preferentially densified, and the lowest densification is observed at the inner rib of the cylinder wall. Moreover, the effect on densification of each HIP condition, including sintering temperature (900–970 °C), pressure (120–180 MPa), and holding time (3–4 h), was evaluated separately. The HIP sintering temperature contributes the most to the improvement of densification, followed by the pressure, while the holding time contributes the least. Investigating HIP densification behavior is beneficial to the structural and process optimization of metal near-net-shape forming applications.

1. Introduction

With the rapid development of aerospace in recent years, requirements for the performance of aerospace components have been increased significantly. Due to its low density, high specific strength, good corrosion resistance, and high temperature properties, titanium alloys are considered to be an ideal candidate for aerospace materials [1,2,3]. At present, there are many processing methods for titanium alloy aerospace components, such as casting [4,5], machining [6,7], hot isostatic pressing [8,9], etc. Among these processing methods, hot isostatic pressing (HIP) combines features of conventional powder metallurgy methods with enhanced material efficiency, fine grain size, and swift densification, which provide HIP with distinct benefits in creating high-performance and intricate structural components [10]. Rolls-Royce Plc adopted the HIP route developed by the University of Birmingham, and manufactured a HIPed 56 kg aeroengine frame by using only 60 kg of titanium alloy powder. The Vinci motor liquid hydrogen impeller on the Ariane space rocket of ESA was manufactured by HIP technology, which not only reduced the material loss, but also increased the service time of the impeller [11]. The titanium alloy support rod and fuselage pillar in the F-14 fighter, the TC4 alloy keel nose of the F-15, and the engine mounting support in the F-18 Hornet all showcase the extensive application of powder metallurgy through the HIP technique [12]. The Institute of Aerospace Materials and Technology has developed TC4/TA15 rudder wing frame products, which boast impressive attributes such as exceptional forming precision, high-quality surface finish, and effective management of internal defects. The material characteristics are on par with those found in forged components, and the overall dimensions of the wing extend to 2200 mm [13].

HIP of titanium alloys shows strong application prospects in aerospace component manufacturing, and has attracted much attention from researchers. Xu et al. [14] systematically investigated the effects of sintering temperature, pressure, and cooling rate on the density, microstructure, and mechanical properties of Ti6Al4V during HIP, and obtained optimized HIP parameters of a temperature range from 900 to 940 °C, a pressure of over 100 MPa, and a holding time of 3 h. Kim et al. [15] investigated the effects of sintering and heat treatment temperature on the microstructure of Ti6Al4V, especially on its phase composition and dynamic recrystallization. New HIP schemes for β Ti-5553 powder were designed, and the effects of these schemes on its tensile and high-cycle fatigue properties were evaluated, by Perevoshchikava et al. [16]. They found that the recrystallization processes in the vicinity of the powder boundaries were enhanced under the designed HIP schemes, and consequently, high tensile strength and excellent high-cycle fatigue properties were obtained. It can be seen that HIP processing conditions of titanium alloys have been widely investigated based on experimental methods.

Although traditional experimental methods can reveal the microstructure and mechanical properties of alloys under specific process conditions, a high temperature and closed operating environment limit the real-time dynamic observation of the densification process, thus restricting the design and optimization of the hot isostatic pressing process. Numerical simulation technology can solve the drawbacks of the traditional experimental trial-and-error method to a large extent, and efficiently provide guidance for structure and process optimization; it is widely used in metal forming fields such as casting, welding, forging, and powder metallurgy [17,18]. In the HIP field, the yield criterion of porous materials in numerical simulation has been studied deeply. Considering the effect of stress on the plastic deformation of porous materials, Kuhn and Downey [19] and Shima and Oyane [20] proposed a yield criterion applicable to porous materials based on uniaxial experiments on powders. However, there were differences between the uniaxial loading and the densification of the powder materials, which reduced the accuracy of finite element simulation of HIP. Based on mechanical analysis and numerical calculation methods, Fleck et al. [21] and Mori et al. [22] proposed different constitutive models. By studying the yield behavior of metal powders connected by isolated contact points, Fleck et al. [21] proposed a macroscopic constitutive model applicable to a density range of 0.6–0.9; however, this model had poor matching ability for soft metal powder. Mori et al. [22] proposed a yield criterion based on the Cosserat continuum theory, which considered the microscopic rotation effect of powder particles through the finite element method, and could satisfy the balance equation of force and torque simultaneously. Doraivelu [23] and Park [24] revised the yield criterion proposed by Shima and Oyane by fitting test data. Lee and Kim [25] modified the Doraivelu yield criterion by considering the influence of relative density on the yield behavior of non-dense materials. Since then, there have been many studies on the mechanical behavior of porous metal materials, and many researchers have proposed yield criteria suitable for describing the powder forming process based on the constitutive model.

In terms of finite element model construction and numerical simulation of HIP, much research has been carried out to study the forming process and optimize the processing parameters. You et al. [10] established a numerical model utilizing a combined finite element approach of plasticity theory and thermodynamics to simulate the HIP process for atomizing and milling Ti6Al4V powders. The simulated results of the two kinds of compacts both agreed well with the experimental results. Considering the plasticity and creep behavior of dense materials under high temperature and pressure, Liu et al. [26] adopted a viscoelastic Perzyna model and combined this with the yield criterion of porous materials to establish a thermo-mechanical coupling finite element model of Ti6Al4V powder for HIP processing. Based on such a finite element model, the effects of temperature and pressure on the densification process of Ti6Al4V powder were investigated. In order to improve prediction by comparing a uniform and non-uniform initial powder distribution, Meng et al. [27] developed a modified Shima–Oyane model, considering the influence of initial powder distribution. After considering the uneven distribution of the initial powder, the maximum error was reduced to 3.16%, and the average error was also less than 2%. Liang et al. [28] investigated the HIP numerical simulation process for thin-walled aluminum alloy complex parts, and the characteristic structural error of the numerical simulation results was about 5%. Xu et al. [29] used MSC.MARC 2020 software to study the densification behavior of TC4 alloy internally reinforced cylindrical parts during hot isostatic pressing, and obtained optimized HIP process parameters of 940 °C/118.42 MPa/2.968 h. Ye [30] and others focused on HIP forming technology for non-axisymmetric complex structural parts used in aerospace, and prepared a series of non-axisymmetric complex structural parts by using atomized spherical titanium alloy powder to achieve near-net-shape forming of complex parts, with a forming accuracy of up to ±0.2 mm. Liu [31], of Huazhong University of Science and Technology, simulated the HIP forming of artificial bone stems for asymmetric complex parts, and found that the densification of non-axisymmetric complex parts is the sum of the changes in the combined forces caused by the geometrical structure that are taken into account on the basis of the law of densification of axisymmetric structures. For current numerical simulations of powder flow and densification behavior in HIP, the main research objects are axisymmetric components. However, for non-axisymmetric components, especially for a cylindrical component with non-axisymmetric inner ribs, there are few studies on their powder flow and densification behaviors. For present numerical simulations of powder flow and densification in HIP, the primary focus is on axisymmetric components, such as on the sizing optimization of inner ribs. Therefore, investigating the powder densification behavior at typical locations of non-axisymmetric components is very important for industrial applications.

In this paper, a typical aerospace thin-walled tubular component with non-axisymmetric inner ribs, made of Ti6Al4V, is selected as the research object, and its HIP process is numerically simulated, in order to investigate the densification behavior and the effects of sintering temperature, pressure, and holding time on the relative density of the non-axisymmetric component.

2. Numerical Modeling

2.1. Geometric Modeling and Meshing

The geometric models of the Ti6Al4V thin-walled tubular component with non-axisymmetric inner ribs, can, and core were constructed with SolidWorks 2022 software, as shown in Figure 1a–c. The can was made of 304 stainless steel, with the wall thickness set as 4 mm. The core was made of high-strength graphite, which was set as a rigid body in the finite element simulation and was not assigned material properties, and consequently, the core was not meshed. Because the geometry of the Ti6Al4V thin-walled tubular component with inner ribs was not symmetrical, the whole model was used for meshing and finite element analysis. Ignoring the effects of degassing holes, the length of the assembly, and the weld joint connecting the end caps to the sidewalls of the can, the mesh was made up of two components: the can and the Ti6Al4V compact, as shown in Figure 1d,e. An eight-node hexahedral element and a six-node tetrahedral element were selected for can meshing and powder meshing, respectively. The model was divided into 139,081 elements, including 15,600 can elements and 123,481 powder elements.

2.2. Plasticity Theory of Titanium Alloy Powder HIP

During HIP processing, the dense 304 stainless steel can and non-dense Ti6Al4V compact both can be deformed under the high temperature and high stress. Therefore, according to the mechanics of continuous media, both the 304 stainless steel can and Ti6Al4V compact are considered as compressible continuous media. In this research, the von Mises yield criterion and Shima–Oyane yield criterion were selected to describe the yield and deformation phenomena during HIP for the dense 304 stainless steel can and the non-dense Ti6Al4V compact, respectively.

The von Mises yield criterion can accurately describe the yield characteristics of dense materials, and the yield criterion expression is defined as follows:

$${\sigma }_s={\displaystyle \frac{1}{\sqrt{2}}}{\left[{\left({\sigma }_1-{\sigma }_2\right)}^2+{\left({\sigma }_2-{\sigma }_3\right)}^2+{\left({\sigma }_3-{\sigma }_1\right)}^2\right]}^{{\displaystyle \frac{1}{2}}}$$

(1)

where σ₁, σ₂, and σ₃ are the principal stresses, and σ_s is the yield stress of the dense materials.

There are significant differences in the yield and deformation features between the dense and non-dense materials. The yield criterion expression of the non-dense materials is defined as follows:

$$F={\left(U{J}_2^{\prime }+W{J}_1^2\right)}^{{\displaystyle \frac{1}{2}}}={\delta }^{{\displaystyle \frac{1}{2}}}{\sigma }_F$$

(2)

where J₁ and $J_2^{\prime }$ are the stress tensor first invariant and the stress deviatoric tensor second invariant; U and W are the material constants; δ is the geometric reinforcement parameter; and σ_F is the yield stress of the powder. Shima and Oyane [20] introduced hydrostatic pressure and took into consideration its influencing factors and correction factors to update the yield criterion of metal powder; the updated yield criterion expression is defined as follows:

$$f^{\prime }{\sigma }_0={\left\{{\displaystyle \frac{1}{2}}\left[{\left({\sigma }_1-{\sigma }_2\right)}^2+{\left({\sigma }_2-{\sigma }_3\right)}^2+{\left({\sigma }_3-{\sigma }_1\right)}^2+{\left({\displaystyle \frac{{\sigma }_m}{f}}\right)}^2\right]\right\}}^{{\displaystyle \frac{1}{2}}}$$

(3)

where σ_m is the hydrostatic pressure, f is the hydrostatic pressure influencing factor, and f′ is the ratio of the actual stress applied to the powder system to the effective stress applied to the powder particles. f and f′ are expressed as follows:

$$\left\{\begin{array}{c}f={\left(b_1+b_2{\rho }^{b_3}\right)}^{b_4}\\ f^{\prime }={\left(q_1+q_2{\rho }^{q_3}\right)}^{q_4}\end{array}\right.$$

(4)

where b₁, b₂, b₃, b₄ and q₁, q₂, q₃, q₄ are the parameters that determine f and f′. According to the previous paper reported by Abondance et al. [32,33], the f and f′ functions of Ti6Al4V compact during HIP are expressed as follows:

$$\left\{\begin{array}{c}f\left(\rho \right)=0.274{\left[{\displaystyle \frac{1-\rho }{\rho -0.66}}\right]}^{0.864}\\ f^{\prime }\left(\rho \right)=1+0.6{\left[{\displaystyle \frac{1-\rho }{\rho -0.66}}\right]}^{0.87}\end{array}\right.$$

(5)

2.3. Boundary Conditions

The process of powder HIP densification is a coupled field of both temperature and stress fields; heat and pressure are applied directly to the outer surface of the can. Combined with the actual hot isostatic pressing process, the hot isostatic pressing route with synchronized loading temperature and pressure was determined, according to Figure 2. The detailed processing conditions are described as follows: firstly, the temperature and the pressure were increased from ambient temperature and pressure, to the target temperature (900–970 °C) and the target pressure (120–180 MPa), respectively. Then, the temperature and pressure were maintained for 180–240 min. After that, the temperature and pressure were gradually decreased to ambient temperature and pressure in 1.5 h. The initial relative density of the powder was assumed to be 65% and uniform everywhere. The initial temperature of the 304 stainless steel can and Ti6Al4V powder were 25 °C.

It was necessary to included nonlinearity in the contact modeling between the can, the powder, and the core during hot isostatic pressing, considering the large deformation that occurred. It was necessary to perform finite element analysis to ascertain the unit motion, contact status, and frictional behavior among the contact bodies. Both the powder and the can were defined as deformable contact bodies. The core was set as a rigid body. An accurate and adaptable direct constraint method was selected to apply penetration-free constraints to the contact bodies. Due to the obvious viscoplasticity of the powder at high temperature, the bilinear friction model was chosen for the hot isostatic pressing process [34,35], which associates the viscous friction and sliding friction with the relative displacement of elastic deformation and plastic deformation, respectively; its expression is as follows:

$$\varphi =‖f_t‖-\mu f_n$$

(6)

where ϕ is the criterion that determines whether the friction belongs to viscous friction or sliding friction, f_t is the shear force threshold of the powder, μ is the coefficient of friction, and f_n is the normal reaction force. When ϕ > 0, the sliding friction model was used for numerical simulation; otherwise, the viscous friction model was used.

2.4. Material Parameters

The can was made of 304 stainless steel, for which the material parameters were directly transferred from the material library of the Marc 2020 software. In the process of Ti6Al4V powder compaction, with increasing relative density, the mechanical and thermophysical properties of the Ti6Al4V compact varied, which were set as follows. The instantaneous elasticity modulus of the Ti6Al4V powder is expressed as follows:

$$E=E_m\left(0.15+0.85{\rho }^{12}\right)$$

(7)

where E_m is the elasticity modulus of the dense Ti6Al4V alloy, E is the instantaneous elasticity modulus of the non-dense Ti6Al4V, and ρ is the relative density of Ti6Al4V. The elasticity modulus E_m of the dense Ti6Al4V alloy at different temperatures is listed in

Table 1. Elasticity modulus of dense Ti6Al4V alloy at different temperatures.

Temperature (°C)	25	100	200	300	400	500	600	700	800	900
Elasticity (GPa)	103	99	94	89	84	80	74	71	65	60

.

The instantaneous Poisson’s ratio of the Ti6Al4V compact is expressed as follows [36]:

$$\mu ={\mu }_m\sqrt{{\displaystyle \frac{\rho }{3-2\rho }}}$$

(8)

where µ_m is the Poisson’s ratio of the dense Ti6Al4V alloy, and µ is the instantaneous Poisson’s ratio of the non-dense Ti6Al4V.

In the process of HIP, the Ti6Al4V compact was subjected to the simultaneous action of heat and pressure; therefore, its yield strength varied with temperature and relative density. The yield strength, at different temperatures, of Ti6Al4V compacts with different relative densities is shown in Figure 3 [10].

Moreover, in the process of numerical simulation of HIP, the thermal physical property parameters of the Ti6Al4V compact, including its thermal conductivity, specific heat capacity, and thermal expansion coefficient, were defined as functions of temperature [10], which are shown in Figure 4.

3. Results and Discussion

3.1. Analysis of the Powder Densification Process

Due to the complex shape of the Ti6Al4V thin-walled tubular component with non-axisymmetric inner ribs, different locations of the Ti6Al4V compact could be expected to exhibit various densification behaviors during HIP process. Therefore, the finite element simulations were carried out using the advanced nonlinear finite element software MSC.MARC to analyze the densification process of the Ti6Al4V thin-walled tubular component. The target temperature, pressure, and holding time were set as 940 °C, 140 MPa, and 180 min, respectively. In order to facilitate observation, half of the Ti6Al4V thin-walled tubular component model (including the typical featured locations) was selected to display the displacement vector and relative density distribution simulation results.

Figure 5 shows the displacement vector distribution of the Ti6Al4V thin-walled tubular component with non-axisymmetric inner ribs at six typical sintering stages. Figure 5a shows the displacement vectors at 5540 s and at the midpoint of the heating and loading procedure. As can be seen in Figure 5a, the displacement vectors of the Ti6Al4V powder show a trend of expanding outward. At such a temperature, a densification process of the Ti6Al4V powder did not occur to a significant extent; meanwhile, outward expansion of the can and powder occurred to a significant extent as the temperature increased, due to thermal expansion. Due to the complex shape, there were obvious differences in thermal expansion displacement at various locations. The maximum displacement appeared on the upper surface of Region A, and the maximum displacement was 0.69 mm. Figure 5b shows the displacement vectors at 10,430 s and during the final stage of the heating and loading process. Generally, at such temperature, densification of the Ti6Al4V powder occurred to a significant extent. Therefore, the can and Ti6Al4V powder were deformed inwards; the maximum inward displacement was 2.61 mm, indicating that the powder flowed to the core in the last stage of the heating and loading process. Figure 5c,d show the displacement vectors at 17,490 s and 20,740 s, which are in the middle and end of the heat- and load-holding process, respectively. Because of the thicker wall thickness of the Ti6Al4V compact in Region A, more Ti6Al4V powder in Region A underwent radial shrinkage, which resulted in larger radial shrinkage displacement occurring in Region A. Thus, the can had non-uniform deformation between the A and B regions, as shown in Figure 5c,d. During the holding time, due to the decrement of deformation resistance between various locations, the relative density of most locations of the tubular sidewall (Region B) was rapidly increased. According to the principle of the lowest plastic potential, the powder in the high-density region could be expected to flow towards the low-density region, resulting in a decreased displacement difference between various locations (in Region B). Figure 5e,f show the displacement vectors at 23,690 s and 27,000 s, which are in the middle and end of the cooling and unloading process, respectively. The results show that the shape of the Ti6Al4V compact in the cooling and unloading stage was not notably distinct from its form during the holding process (Figure 5d). It is worth noting that, in the deep triangular ribs (Region D), because of the small loading area and large depth, the minimum displacement always occurred in this location. After the whole HIP process, the axial and radial average shrinkage displacements of the Ti6Al4V compact were 8.25 mm and 3.44 mm, and the axial and radial average shrinkage ratios were 1.63% and 48.61%. Therefore, the densification process of the Ti6Al4V thin-walled tubular component with inner ribs mainly resulted in radial shrinkage.

Figure 6 shows the relative density distribution of the Ti6Al4V thin-walled tubular component with non-axisymmetric inner ribs at six typical stages. Figure 6a shows the relative density distribution at 5540 s and at the midpoint of the heating and loading procedure. As can be seen in Figure 6a, the relative density of the Ti6Al4V powder was approximately 64.53%, which is even lower than the initial 65%. The lower relative density is in accordance with the viewpoint that thermal expansion took the lead at this stage. Figure 6b shows the relative density at 10,430 s and in the last stage of heating and loading procedure. The relative density of most of locations on the Ti6Al4V compact was higher than 69%, indicating that the Ti6Al4V compact began to densify in this stage, which is in agreement with the displacement results. Figure 6c shows the densification outcomes in the middle of the heat- and load-holding process (17,490 s). Although non-uniform densification occurred at different locations, the relative density at all locations increased significantly. As the most representative location, the relative density of the tubular sidewall (Region B) reached 82.53%. The maximum relative density occurred in Region A, and the maximum relative density was 99.33%, which is close to fully dense. With the holding time increasing to 20,740 s, the relative density of Region B and Region C increased to 83.40% and 87.95%, respectively. It should be noted that the most difficult-to-densify locations with a remarkably low density were found in Region D, with a relative density as low as approximately 71.23%. In Region D, a particularly small loading area and large shrinkage distance were the dominant reasons for the lowest relative density. Then, in the cooling and unloading process, as shown in Figure 6d,e, the relative density of the Ti6Al4V compact could not continue to obviously increase. The final relative density distribution was close to the end of the holding process (Figure 6d). After the whole HIP process, the Ti6Al4V thin-walled tubular component exhibited four featured locations, including an easy-to-densify location (upper circular ribs, Region A), a moderately difficult-to-densify location (tubular sidewall, Region B, and large square inner rib, Region C), a difficult-to-densify location (junctions between Region B and Regions A/C/D), and the most difficult-to-densify location (deep triangular ribs, Region D), respectively. In summary, under typical HIP conditions (940 °C × 140 MPa × 180 min), besides the easy-to-densify location, most of the other locations of the Ti6Al4V thin-walled tubular component with non-axisymmetric inner ribs are not dense enough. Therefore, the HIP process conditions of the Ti6Al4V thin-walled tubular component with a complex shape need to be further optimized.

3.2. Effects of Processing Parameters on Densification

For the HIP process, sintering temperature, pressure, and holding time are critical processing variables. In order to further improve the relative density of the Ti6Al4V thin-walled tubular component with non-axisymmetric inner ribs, the effects of sintering temperature, pressure, and holding time on the relative density of the as-HIPed Ti6Al4V thin-walled tubular component were investigated as follows.

Various sintering temperatures, of 900 °C, 920 °C, 940 °C, 950 °C, and 970 °C, were employed to replicate the HIP process for the Ti6Al4V thin-walled tubular component. The other parameters were consistent with those described above (Section 3.1). Figure 7 shows the relative density distribution of the Ti6Al4V thin-walled tubular component after HIP at various sintering temperatures. It can be seen clearly that, with increasing sintering temperature, the relative density of the Ti6Al4V thin-walled tubular component remarkably increased. When the sintering temperatures were 900 °C and 920 °C, the relative densities at the most representative location of the tubular sidewall location (Region B) were only approximately 77.28% and 81.74% (Figure 7a,b), much lower than that of 88.15% HIPed at 940 °C (Figure 7c). As the sintering temperature increased to 950 °C and 970 °C, the relative densities at Region B increased to approximately 94.19% and 99.99% (Figure 7d,e), 29.38% higher than the densities of the same location at 900 °C. It should be noted that, although the relative density of most locations was higher than 99% after HIP at 970 °C, there were still some non-dense locations in the as-HIPed component, including the deep triangular ribs (Region D, where the relative density was approximately 71.86%), and the junctions between the tubular sidewall and its upper, as well as ribs (junctions between Region B and Regions A/C/D, where the relative density was approximately 90.37%). According to the simulation results, increasing sintering temperature is beneficial for improving the relative density of the Ti6Al4V thin-walled tubular component.

Figure 8 shows the relative density distribution of the Ti6Al4V thin-walled tubular component after HIP under various pressures (120 MPa, 140 MPa, 160 MPa, 180 MPa). The other parameters are consistent with those discussed in Section 3.1. As can be seen in Figure 8a,b, as the pressure increased from 120 MPa to 140 MPa, the relative density of Ti6Al4V thin-walled tubular component at the typical location of tubular sidewall (Region B) significantly increased from 78.57% to 88.15% with relative density growth rates of 12.19%. Moreover, for the easy-to-densify location (Region A), the relative density also obviously increased from 93.71% to 99.65%, with a relative density growth rate of 6.32%. When the pressure further increased to 160 MPa and 180 MPa, the relative density of most locations continued to increase, but this increase was not as significant as under relatively low pressure (120–140 MPa). Especially for the most difficult-to-densify locations, the relative density did not obviously increase with an increase in pressure from 140 MPa to 160 and 180 MPa. In summary, under such a sintering temperature and holding time, the densification improvement caused by increased pressure is significant in the range of 120 MPa to 140 MPa, but not significant in the range of 140 MPa to 180 MPa.

Figure 9 shows the relative density distribution of the Ti6Al4V thin-walled tubular component after HIP with various holding times (3 h, 3.5 h, and 4 h). The other parameters are consistent with those discussed in Section 3.1. As can be seen in Figure 9, with the holding time increasing from 3 h to 3.5 h and 4 h, the relative density at the tubular sidewall location (Region B) increased from 88.15% to 90.86% and 92.38%, with average relative density growth rates of 3.07% and 4.80%, respectively. However, for the difficult-to-densify locations (junctions between Region B and A/C/D), the densification improvement was not significant. Therefore, compared to sintering temperature and pressure, holding time makes a relatively small contribution to the relative density of the Ti6Al4V thin-walled tubular component.

4. Conclusions

In this paper, the powder flows and densification behavior at various locations of the Ti6Al4V thin-walled tubular component with non-axisymmetric inner ribs were studied by finite element simulation. Moreover, the influence of the processing parameters on densification was also studied. The primary conclusions are presented below.

(1) The displacement vector and densification behavior at various locations of the thin-walled tubular component with non-axisymmetric inner ribs are distinctly different. Densification preferentially occurs at the upper locations, followed by the tubular sidewall locations, and finally at the inner rib locations. Under the typical HIP conditions of 940 °C × 140 MPa × 180 min, the highest relative density appeared at the upper location, achieving higher than 99%; meanwhile, the lowest relative density appeared at the triangular rib locations, equaling only 71%.

(2) Increasing the sintering temperature, pressure, and holding time can the promote densification process of the Ti6Al4V thin-walled tubular component. Among these processing parameters, sintering temperature contributes the most to densification improvement, followed by pressure, and the holding time contributes the least. A high-dense thin-walled tubular component with non-axisymmetric inner ribs made of Ti6Al4V alloys can be obtained by HIP at 970 °C. The relative density at most locations is higher than 99%.

(3) On the basis of the existing steel can and graphite core, the inner triangular ribs and the junctions between the tubular sidewall and bottom/ribs were difficult to densify, indicating that more densification strategies need to be further explored to achieve complete densification of the whole component at every location. Combined with the powder flow and densification behavior results, a reverse design of the can and core structure should be further investigated in the future.