Introduction

Shape memory alloy (SMA) actuators are attractive for their exceptional actuator and sensory capabilities, especially in industries like aerospace and automotive where achieving lightweight construction is crucial. SMAs, with their high-power density, play a vital role in meeting these demands. The shape memory effect (SME) is utilized to induce actuator stroke, typically activated through resistance heating or a surrounding medium. While semi-finished products such as wires, tubes, sheets, or foils are commonly used for actuator applications, their transformation into functional actuators requires expertise in shape memory alloys. Mechanical anchoring, often achieved through processes like crimping, is essential for transmitting generated forces. Additionally, electrical connections are frequently needed for activation, leading to significant custom development work and expertise in the complex field of SMAs, incurring both time and cost. A recent advancement involves a form-fit connection using laser-processed spheres at wire ends. This method, complemented by welding a tinned copper (Cu) wire on the sphere's top, ensures electrical activation, marking a significant step toward standardizing actuator components and simplifying SMA accessibility. An even more sophisticated approach is the direct connection through laser welding SMA wires onto printed circuit boards (PCBs). Leveraging widely used PCBs provides a well-researched platform for standardized actuator components, streamlining engineering processes for new actuator systems. Laser welding of nickel-titanium (NiTi) alloys onto PCBs is a reliable method due to its small heat-affected zone, enabling modification of shape memory properties. Challenges in welding NiTi and Cu, such as brittle precipitates and the inclusion of Sn in the Ni-Ti-Cu-Sn system, require careful consideration. Moreover, the thickness of tinned copper (Cu) layers on printed circuit boards is typically 70 μm or smaller, posing an additional challenge, especially when combined with the commonly used FR4 base material. Reinheimer et al. initially used a theoretical approach to underscore the fundamental feasibility of a welded joint on a printed circuit board. Subsequently, they successfully validated their approach through blind joints, demonstrating the absence of delamination defects. The key challenge lies in controlling heat input during the welding process, as excessive heat can lead to delamination processes within the FR4. To address this concern, a nanosecond laser is employed, in contrast to a previous publication that used a modulable continuous-wave fiber laser. This choice of laser technology aims to enhance precision in heat control, mitigating the risk of delamination processes in the FR4 substrate. Traditional soldering methods, often used for bonding components to PCBs, face challenges with NiTi due to poor wettability, necessitating aggressive fluxes or ultrasonic soldering. In contrast, micro laser welding, as explored in this paper, offers a stable joint with controlled heat input.

Experimental Results

The experiments were conducted with an Yb-fiber laser redEnergy from SPI Lasers, operating between 1059 nm and 1065 nm with a pulse length of 6-500 ns and a repetition rate up to 1 MHz. The experimental setup with the optical paths is shown in Fig. 1. The laser beam is guided to the workstation with an optical fiber and then collimated using a lens (L). Subsequently, the beam is transmitted through a dielectric mirror (DM) to a SCANcube 10 from Scanlab and then focused on the NiTi-wire using an f-theta-optic with a focal length of 160 mm. To visualize the process for improved positioning a camera setup with a movable objective is utilized. The experiments employ binary NiTi shape memory alloy wires containing 49.8 at.% Ni (martensitic at room temperature) with a diameter of d = 0.24 mm, manufactured by Ingpuls GmbH. The PCBs utilized are produced by JLCPCB, featuring FR4 base material and Lead-Free HASL-RoHS surface finish. Before processing, both the wires and PCBs undergo cleaning with isopropanol.

In the experimental investigation of manufactured specimens, uniaxial tensile tests were conducted until failure to assess their mechanical properties. The obtained tensile strength values consistently hovered around 705 MPa. Notably, the specimens exhibited failure predominantly in the processed area. Figure 2 shows images of a fractured sample after tensile testing. The upper images display fractures on the PCB while the corresponding bottom images highlight the wire counterpart, revealing a rupture that included part of the weld seam and attached tin flakes. Further examinations involved subjecting the manufactured actuators to lifespan testing through cyclic activation until failure with a 300 MPa load. Electrical activation parameters were set with a 3-second activation phase followed by a 15-second cooldown. Remarkably, the manufactured actuators demonstrated impressive performance, achieving up to 14000 cycles before failure.The failure analysis of the actuators revealed potential multiple fractures, categorized into primary and possible secondary fractures. The primary fracture originated in the unprocessed material, starting from surface defects. These defects were attributed to the influence of driving wheels from the integrated wire feeder, a crucial component in the manufacturing process of loop actuators. The primary failure within the unprocessed material can instigate a subsequent secondary failure at the weld spots. In instances where failure occurs at the primary locations within the untreated material, the remaining specimen undergoes acceleration, inducing additional shear forces on the welded points on the circuit board. Consequently, secondary fractures may arise at these weld seams due to the heightened stress conditions. Regardless of whether secondary fractures manifest in a given tested specimen, a distinctive undulating deformation pattern preceding the welded points is consistently observed in every sample.Differential Scanning Calorimetry (DSC) measurements were conducted to assess the extent to which the welding process alters the transformation behavior of the material. It was observed that within the welded region, the transformation behavior persists, albeit with noticeable modifications. Specifically, the transformation temperatures were found to be shifted towards higher temperature ranges in the welded areas.

In addition to mechanical testing and the DSC measurement, a comprehensive microstructural characterization was conducted utilizing Electron Backscatter Diffraction (EBSD) measurements and Transmission Electron Microscopy (TEM) investigations with Energy Dispersive X-ray Spectroscopy (EDX) for elemental analysis and phase identification. These advanced analytical techniques provided detailed insights into the internal structure and composition of the manufactured specimens and actuators. EBSD measurements were employed to analyze the crystallographic orientation of the materials, offering a thorough understanding of the grain structure and texture. TEM investigations, coupled with EDX analysis, enabled a high-resolution examination of the microscale features, including multiple sections of the weld seam. The elemental composition of these microstructural constituents was accurately determined, providing critical information on the distribution of alloying elements and potential precipitates. Furthermore, phase identification through TEM allowed for a precise characterization of the material phases present within the specimens.

Conclusion

In conclusion, the welding of NiTi wires onto printed circuit boards using nanosecond lasers with a high repetition rate proves to be a promising approach for the development of future standardized shape memory actuators. The welded specimens exhibit not only sufficient mechanical tensile strength but also impressive cyclic durability, as demonstrated by their ability to withstand a considerable number of activation cycles. The microstructural investigation, facilitated by EBSD measurements and TEM examinations with EDX has provided a profound understanding of the interconnection between the NiTi wires and the circuit boards. The EBSD measurements allowed for a detailed analysis of the crystallographic orientation and grain structure. Meanwhile, TEM investigations with EDX analysis permitted a high-resolution of multiple sections of the weld seam, offering precise information on elemental composition and phase identification. This multidisciplinary approach not only confirms the mechanical integrity and cyclic performance of the manufactured specimens but also provides a comprehensive characterization of their microstructure. The combination of sufficient mechanical properties, cyclic durability, and a well-characterized microstructure underscores the viability of the proposed manufacturing method. These findings lay the groundwork for the continued development and standardization of shape memory actuators, showcasing their potential for diverse applications in the field of advanced materials and actuator technologies.