3.1. The Microstructure
Φ16 mm tubes with different wall thicknesses were obtained by rolling Φ25 mm tube billets with ‘Q’ ratios of 0.65, 1.1, 1.6 and 2.0 in a single pass. The microstructure of tube billets, cold-rolled tube (CR) and annealing tube is shown in
Figure 2.
Figure 2a is the orientation image of the radial plane of the tube billet. Before rolling, the tube billet was dominated by equiaxed α grains of recrystallized, most of which <0001>∥RD, with an average grain size of about 11.5 μm.
Figure 2b is the diagram of grain boundary misorientation distribution of the tube billet. The high angle grain boundary (HAGBs) with grain boundary misorientation > 10° is the main distribution of the original tube billet, with HAGBs accounting for more than 95%, indicating that the original microstructure is completely recrystallized.
As shown in
Figure 2c–f, through a single pass cold rolling, the original tube grains along the axial elongated, forming tiny uniform deformation structure. As the ‘Q’ ratio increases, the tube wall thickness thinning increases, tube deformation increased from 53% to 70% (
Table 1), the length-width ratio of grain is gradually increasing also. When the ‘Q’ ratio is 2.0, the aspect ratio reaches 5–6. The tube after stress-relieved annealing (
Figure 2g–j) has basically the same microstructure and morphology as the cold-rolled tube.
Figure 2c–j shows the grain boundary misorientation distribution of the tube after rolling and annealing. For the tube after single pass cold rolling, the grain boundary misorientation is mainly distributed in low angle grain boundaries (LAGBs) of 0–10°, accounting for about 80%. By comparing the results of the cold rolled and annealed tube, it can be found that no recrystallization occurs in the grain of the tube after vacuum annealing at 490 °C/3 h. At this temperature, only the internal stress of the tube is eliminated. Therefore, annealing tube is only selected for subsequent texture research.
3.2. Texture Evolution
For Ti-3Al-2.5V alloy tube, different deformation processes will affect the C-axis orientation of the grain basal plane. When the C-axis is mainly concentrated along the radial direction of the tube, the tube has a strong radial texture; when the C-axis is mainly concentrated along the circumferential direction of the tube, the tube has a strong circumferential texture.
Figure 3 shows the pole figures of {0001}, {10
10} and {11
20} of the tube billet and rolled tube with different pass ‘Q’ ratios. The pole figures are drawn according to contour lines of the same level, and the contour lines are divided according to the intensity level of {0001} pole figure.
Figure 3a is the pole figure of the original tube billet. It can be seen that the pole density of the {0001} pole figure is mainly distributed in the radial (RD) and circumferential (TD) planes of the tube, and the strong pole density values is splitting distributed in the range of ±45° to 60° on both sides of the normal line, and the included angles between the positions of the two maximum points and the normal line are about −45° and 55°. The maximum value of the extreme density is not more than 4.0, indicating that the initial billet has not too strong radial texture. At the same time, it can be seen from the pole figure of {10
10} and {11
20} that the grain prismatic and pyramidal plane are centrally distributed along the axial direction of the tube. The density value of {11
20} pole is stronger reaching 3.15. The {11
20} plane is centrally distributed along the axial direction (AD) is the main characteristic of recrystallization texture of α type titanium alloy [
28].
The pole figures of {0001}, {10
10} and {11
20} of tubes rolled with different pass ‘Q’ ratios are shown in
Figure 3b–e. It can be seen that for the tube rolled with pass ‘Q’ ratio of 0.65, the pole density of {0001} pole figure is mainly distributed in the range of ±30–90°, and the pole density with strength over 4.0 is mainly distributed in the vicinity of −75° and 35–60°, and the extreme point is located at the Angle of −75° and 50° with RD. It shows that the grain C-axis of tube is mainly distributed in a circumferential direction with circumferential texture. When the tube is rolled with ‘Q’ ratio of 1.1, the part of pole density over 4.0 in {0001} pole figure is mainly distributed in the range of ±15° to 70°, and the angle between the pole point and RD is about ±45°, which is similar to the radial and circumferential uniform distribution. When the tube is rolled with ‘Q’ ratio of 1.6, the pole density of {0001} pole figure is mainly distributed in the range of ±20° to 60°, and the poles deviate from RD about ±45°, indicating that the grain orientation of the tube is gradually concentrated in radial direction, with radial texture characteristics. When the ‘Q’ ratio of the tube is further increased to 2.0, the main distribution range of the pole density of the {0001} pole figure is ±0–60°, and the angle between the pole point and RD is ±40°, and the radial texture is enhanced. At the same time, the {10
10} and {11
20} pole figures have similar pole density distribution characteristics after rolling with different ‘Q’ ratios. The pole density of{11
20} pole figure are distributed around the pole figure in circular shape, while the pole density points of {10
10} pole figure are concentrated along the axial direction, and the pole density values are very high. This is the main characteristic of the deformation texture of α titanium alloy with close-packed hexagonal structure. By comparing the {10
10} and {11
20} pole figure of the original tube billet (
Figure 3a), it can be seen that the {11
20} oriented grains in this part may have rotated 30° along the C-axis, making the {10
10} plane aligned with the axial direction.
In order to more clearly and intuitively represent and analyze the variation characteristics of the pole density along the radial-circumferential section of the tube,
Figure 3f presents the pole density distribution on the 0° RD-TD section of the {0001} pole figure of the rolled tube with four Q values. It can be seen that the pole density distribution of the RD-TD section of the pole figure presents two peak values, which are distributed on both sides of the RD direction, respectively, representing the concentration points of the two pole density values. The strength of {0001} texture has a certain relationship with the included angle of the peak point of the pole density deviating from the RD direction. Generally, the peak points of pole density are symmetrically distributed on both sides of RD. The degree to which the peak point of pole density deviates from RD can be expressed by the angle between the two peaks. After measurement, the angle between the {0001} peak pole density of tube billet is 100°, and the radial texture of the initial tube is not significant. Similarly, after the tube rolled with ‘Q’ ratio of 0.65, 1.1, 1.6 and 2.0, the included angles between the peak values of {0001} pole density are 122.5°, 90°, 90° and 80°, respectively. It can be seen that, with the increase of ‘Q’ ratio, the peak values of the {0001} pole density of the tube are closer to the RD axis. In addition, when the Q < 1, the angle between the {0001} pole density peak and RD axis of the tube increases, which is not favorable to obtain the radial texture required.
Figure 4 shows the evolution characteristics for {0001}{10
10}{11
20} pole figures and RD-TD section pole density distribution of rolled tube with different ‘Q’ ratios.
The radial texture is very important for Ti-3Al-2.5V alloy tube. The tube with strong radial texture has strong resistance to radial deformation. The radial inverse pole figure can show the texture distribution along the radial directions.
Figure 5 shows the radial direction (RD) inverse pole figure of the experimental tube billet and the rolled tubes with four ‘Q’ ratios. It can be seen from
Figure 5a that the original tube billet has a strong {0001} texture along the RD direction of the sample coordinate system, while it cannot be seen that {10
10} and {11
20} planes are almost concentrated in the radial direction.
Figure 5b–e are RD inverse pole figures of tube rolled with four ‘Q’ ratios. The pole density distribution is still dominated by {0001} orientation. Different from the tube billet, the rolled tube with ‘Q’ ratios, except that {0001} plane is oriented in radial direction, There is also a scattered {11
2X} pole density distribution on the RD inverse pole figure from {0001} to {11
20} section, and its concentration decreases with the increase of ‘Q’ ratio.
The change of the radial texture with the rolling ‘Q’ ratio is mainly affected by the orientation distribution of the grain c-axis inside the tube. The more the grains of c-axis deviated to the radial direction of the tube and the smaller the angle with RD, the stronger the radial texture will be. In the RD inverse pole figures of the rolled tube with four ‘Q’ ratios, it is observed that in addition to the strong {0001} texture along the radial plane of the tube, there is also a changing sub-strength texture, whose extreme density intensity is about 1.77–2.02. When the ‘Q’ ratios are 0.65, 1.1 and 1.6, The orientation plane of sub-strong texture is {11
22}, {11
23}, {11
25} in order. When the ‘Q’ ratio is 2.0, no obvious sub-strong texture is found. It indicates that, in the radial plane (AD-TD), except that the {0001} basal plane of most grains is parallel to it, the direction plane of sub-strong texture is {11
22}, {11
23}, {11
25} plane of some grains are parallel to the radial plane. This relationship can be graphically described in
Figure 5f. As shown in the figure, when the orientation plane of sub-strong texture is {11
22}, the c-axis deviation to RD is the largest, and the deviation angle can be calculated to be about 57.8°; when the orientation plane is {11
23}, the c-axis deviation to RD is about 46.6°; when the orientation plane is {11
25}, the c-axis deviation to RD is about 32.4°; when Q is 2.0, the sub-strong texture disappears. This is because the X value of {11
2X} increases sharply with the decrease of the angle between grain c-axis and RD, and the texture changes to the dominant {0001} texture. With the increase of ‘Q’ ratio, more c-axis of grain tends to radial distribution, so the radial texture of tube is continuously enhanced.
3.3. Discussion
It is found that the multi-stroke periodic Pilger cold rolling process is a composite deformation process which includes the reduction of tube diameter and wall thickness, and a non-steady and non-uniform plastic deformation process which combines geometry, material and boundary conditions. The stress and strain state in the rolling process dominates the plastic deformation mechanism of tube which is the direct cause of different texture types of tubes. According to the concept of true strain. When the size of tube before and after deformation is known, the total strain of a single pass deformation and the strain components along radial, circumferential and axial directions can be calculated.
We calculate the total true strain components in different directions of the rolled tubes with four ‘Q’ ratios in this experiment (
Table 3). It can be seen that the rolled tubes with different ‘Q’ ratios have different strain components. Controlling the ‘Q’ ratio and the deformation amount of the rolled passes essentially dominates the distribution of the strain variables along different directions. The radial, circumferential and axial components of the strain can be represented by a vector in a three-axis plane coordinate system with an angle of 120° to each other. This diagram is called the strain diagram, and any change in tube size can be represented by a vector on the strain diagram, called the strain vector. The magnitude and angle of strain vector are directly corresponding to the rolling deformation degree and deformation mode, and the projection length along the three axes represents the deformation assign weights along the radial, circumferential and axial directions [
10,
27]. The strain diagram of this experiment is shown in
Figure 6. Compared with the experimental results, when the strain vector is in the −ε
r—ε
a quadrant, the rolled tube has radial texture, and the more the strain vector is deviated to the −ε
r axis, the stronger the radial texture, and the closer to the ε
a axis, the weaker the radial texture. Accordingly, in the −ε
t—ε
a quadrant, the rolled tube has circumferential texture, and the strength of texture is related to the degree of strain vector deviation from −ε
t axis. It can be seen that the texture evolution of single pass rolled tube is determined by the distribution of the strain.
In the production of Ti-3Al-2.5V tube, ‘Q’ ratio was taken as the main process control parameter, and ‘Q’ ratio was often used to control the degree of circumferential and radial pass deformation in the cold rolling process [
7]. ‘Q’ ratio of tube rolling refers to the logarithmic ratio of wall reduction rate and diameter reduction rate.
is the wall thickness of finished tube (mm); is wall thickness of tube at last intermediate annealing (mm); is middle diameter of finished tube (mm); is middle diameter of tube at last intermediate annealing (mm).
According to the formula, log value of wall reduction rate is radial true strain and log value of diameter reduction rate is circumferential true strain. ‘Q’ ratio is a parameter reflecting the relative amount of radial true strain and circumferential true strain. Thus, ‘Q’ ratio mainly controls the radial and circumferential strain of tube deformation. The true radial and circumferential strains can be described by a plane strain ellipse perpendicular to the axial direction of the tube [
7,
9], as shown in
Figure 7. In the rolling process of tube, both radial and circumferential strains are compressive strains. For the deformation with high ‘Q’ ratio (Q > 1), the radial true strain of tube is greater than the circumferential true strain, and the tube will acquire the radial {0001} texture whose normal line of the basal plane is nearly parallel to the radial. For Q = 1, the radial true strain of the deformed tube is equivalent to the circumferential true strain, and the {0001} basal is evenly distributed in the radial-circumferential plane. For Q < 1, the radial true strain of the deformed tube is smaller than the circumferential true strain, resulting in the basal parallel to the circumferential direction and the formation of {0001} texture with circumferential alignment.
Due to the special requirements of aircraft hydraulic systems, the optimal orientation of Ti-3Al-2.5V tube is the basal texture distribution along the radial direction of tube. Through the discussion of the experimental results, the high ‘Q’ ratio rolling can effectively increase the radial strain and reduce the circumferential strain. This strain mode can help the grain basal pole orient along the radial direction, so as to obtain the strong radial texture. Moreover, each ‘Q’ ratio corresponds to a particular dies in Pilger rolling, which makes it difficult to do much research. The understanding of strain vector can not only explain the relationship between ‘Q’ ratio and texture, but also predict other ‘Q’ ratio. Therefore, it is of great practical significance. It can be seen from the research that when the initial texture of the tube is radial-circumferential equally distributed, in order to obtain the radial texture of the tube, the ‘Q’ ratio has to be greater than 2.0, which can provide reference for the production process.