Before starting this article, let's introduce two materials science terms - shape memory alloys and gradient functional materials. Shape memory alloy refers to an alloy with shape memory effect, which has a wide range of applications in clinical and medical fields. For example, the large antenna on an artificial satellite can be made of memory alloy. Before launching the artificial satellite, the parabolic antenna is folded and installed inside the satellite body. After the rocket takes off and sends the artificial satellite to the predetermined orbit, it only needs to be heated. The folded satellite antenna naturally unfolds due to its "memory" function, restoring its parabolic shape.

(Image from the internet)

Gradient functional materials refer to a new type of material that exhibits continuous gradient changes in composition or structure. Its design requires that the function and performance vary with the internal position of the component, thereby optimizing the overall performance. For example, animal bones are a gradient structure with a tough exterior and a porous interior.

(Image from the internet)

If NiTi alloy is endowed with the dual characteristics of functional gradient structure and shape memory performance, it will greatly promote its widespread application as intelligent components. Yang Yang and others from Guangdong University of Technology used a repetitive laser processing method to obtain gradient function in NiTi shape memory alloys, in order to expand the martensitic transformation range and improve the controllability of shape memory alloys. Due to the complex coupling mechanism between microstructure and gradient function, in this study, researchers used three precision X-ray micro CT (nanoVoxel-3000 series) to characterize the characteristics of microstructure gradient in detail; At the same time, the mechanical properties and gradient function were evaluated through cyclic deformation and hardness tests, and the influence of processing parameters on its microstructure, properties, and fracture mechanism was analyzed. Finally, methods that are beneficial for designing gradient NiTi alloys with potential complex properties were discussed. This work provides a novel and effective method for manufacturing three-dimensional graded NiTi alloys, which can perform complex structural design to meet the expected functional performance in different fields.

Sanying Precision NanoVoxel-3000 Micro CT

The repeated laser processing strategy used can be divided into two steps. The first step (see Figure 1b) involves initial laser scanning, with the laser power value gradually changing from 10W to 91W. The second step (see Figure 1c) is a second repeated laser scanning step, which uses a constant laser power of 60W to remelt the same layer. Figure 1 shows a schematic diagram of the repeated laser processing strategy used to manufacture gradient NiTi alloys and a physical image of the sample.

Microscopic CT non-destructive testing (NDT) can test the external morphology and internal defects of a specimen. The nanoVoxel-3000 system was used to scan the tensile specimen. Considering that the measurement length of the tensile specimen is 20mm, this data limits the realization of higher resolution imaging. Therefore, the experiment divided the specimen into 5 parts and scanned them one by one to ensure better imaging quality. Under these conditions, the voltage and current of the X-ray source were set to 150kV and 40 μ A, respectively, with an exposure time of 0.4s. To ensure that the sample remained within the field of view during micro CT scanning, the resolution was set to 3 μ m, and image reconstruction was completed using software. Finally, Avizo software was used to further reconstruct the three-dimensional (3D) structure of the sample.

The five images in Figure 2 are continuous 3D reconstructions of each tensile specimen along the gradient direction. For convenience, the continuous five parts of the image shown in Figure 2 will be includedFor parts I, II, III, IV, and V (as shown in Figure 1d). For each image in Figure 2, the initial laser power increases from bottom to top. In Figure 2a, the coordinate system defined by BD and GD is clearly marked with red arrows. The gradient width of 0.4mm is clearly shown in Figure 2d. It is worth noting that the surface roughness of different parts varies due to different processing conditions. For the gradient regions manufactured at higher laser powers in Figure 2e, these surfaces are smoother than those manufactured at relatively lower laser powers (Figures 2a and b).

Figure 2 Continuous 3D reconstruction of five tensile specimens along the gradient direction

The green pixels in Figure 3 represent the corresponding shapes of the three-dimensional internal defects of the five parts, each with differences between them. Figure 3 shows the morphology, distribution, and volume fraction changes of defects along the gradient direction. Specifically, the first part shown in Figure 3a corresponds to the gradient region from 10W+60W to 25W+60W, where the defect shape is irregular. On the cross-sections of the TD-GD plane (Figure 4a) and TD-BD plane (Figure 4c), these irregular defects are planar defects that also stretch along the TD and GD directions between the stacked layers. Considering that the sample shown in Figure 3a corresponds to an initial lower laser power, these defects are caused by insufficient laser energy input during the initial scanning process, despite the total input of laser energy being sufficiently high. As the initial laser power increases, the defect volume fraction in the gradient region corresponding to part II of Figure 3b, from 27W+60W to 41W+60W, is smaller compared to part I (shown as 0.74% in Figure 5). The third part shown in Figure 3c, corresponding to moderate laser power, does indeed contain the least number of defects (0.09% as shown in Figure 5). The defects in the gradient region corresponding to higher laser power in Figures 3c and d are different from other defects *. For part IV in Figure 3d and part V in Figure 3e, the geometric shape of the defect becomes smaller and spherical. This defect is referred to as a "keyhole" in metals produced by selective laser melting (SLM) and is typically associated with excessive input of laser power. Interestingly, compared to the relatively low laser power and less severe layering in the bottom area of Figure 3d, the distribution of spherical defects exhibits a layering phenomenon. The lock hole defect almost disappears at the boundary of adjacent gradient regions, and the melting region corresponds to relatively low laser power. Therefore, it can be considered that the lock hole defect is caused by the molten pool overheated by the laser power.

Figure 3 shows the three-dimensional reconstruction of the scanning results of five tensile specimens, revealing internal defects in different parts of the five sections

Based on the micro CT scan results, perform statistical analysis on the size of all defects. The dimensions of the three defects are shown in Figure 4b and Figure 4d as examples. The average defect length of Figure 4c is about 0.21mm, and the average defect length of Figure 4d is about 0.04mm. Therefore, the irregular planar defects formed by low power are nearly 10 times larger than the lock hole defects formed by high power. Simultaneously analyzed the volume fraction of defects in all five parts. The histogram in Figure 5 shows the variation of defect volume fraction in different parts. The volume fraction of Part I is 2.84%, and Part V is 7.09%. The defect volume fraction in Part III is prepared between 43W and 59W, as low as 0.09%. This indicates that it is advisable to manufacture NiTi alloys with moderate laser energy input in order to form the minimum number of internal defects.

Figure 5 Defect volume fraction of functionally graded samples

The CT scan reconstruction results are shown in Figures 3 and 4. The external morphology and internal defects of the graded sample vary with the initial laser power in the first step. At lower initial laser power, rough surfaces and irregular planar defects were formed between layers. But when the laser power input is too high, it will form smooth surfaces and spherical lock hole defects. Compared to them, the microstructure corresponding to the medium initial laser power (43W-59W) results in a relatively smooth surface with minimal defects, as shown in Figure 2c. Considering the internal defects shown in Figure 6, the low power of the first step laser scanning resulted in a large number of irregular defects, which can be partially eliminated by repeated laser scanning, as shown in Figure 6d. However, excessive input of laser power, such as initial laser scanning with a constant power of 60W and secondary laser scanning with 60W, promotes the formation of B19 'phase on one hand, and leads to the formation of more lock holes on the other hand. Therefore, we know that the first and second laser scanning steps are very important. These facts suggest that we should avoid using low laser power below 30W during the first scanning process, as this will result in too many irregular defects. But once the initial laser power is greater than 60W, the secondary laser power should not be too high to avoid the formation of lock holes.

Through high-resolution micro CT scanning, the appearance and internal defects of NiTi memory alloys with gradient function were non destructively visualized. Combined with other analysis methods, the mechanism of defect formation was revealed, providing guidance for the selection of repeated laser processing conditions. This is a testing method for studying NiTi memory alloys with gradient function.