1 Introduction

Most modern electronic devices are manufactured using printed circuit boards (PCBs). These boards typically consist of a glass-reinforced epoxy resin laminate substrate (FR4) onto which electrical copper circuits are stacked with mounted or attached components and conductors [1]. Traditional assembly methods for electronic components on PCBs involve press-fit processes, in which holes are drilled through the board to attach electrical components according to the amount of mechanical interference [2]. However, this method has disadvantages such as a high number of process steps, the need for pre-drilled holes on both sides of the board, as well as the use of an anvil, and connecting or mating screws [3].

In this context, the Friction Riveting method has been proposed as a promising alternative to press-fit techniques. This process offers significant advantages for PCB assembly, notably a reduction in process steps, shorter joining cycles due to the elimination of pre-drilling, and the requirement for access from only one side of the board [4]. Friction Riveting is based on the principles of mechanical fastening and friction welding [5], where frictional heat and pressure are applied to plasticize and deform a cylindrical metal rivet into one or more thermoplastic or composite components.

However, friction-assisted joining of polymers presents its own set of challenges. The low thermal conductivity and high crystallinity of polymers can weaken joint strength [6]. Low thermal conductivity results in uneven heat distribution during the process, leading to irregular melting and cooling, while high crystallinity can induce brittleness and poor bonding, resulting in weak joints and voids. To address these challenges, innovative techniques such as friction riveting [5] and friction self-riveting welding [7] have been developed, displaying their efficiency in producing strong joints.

Recent findings on Friction Riveting have demonstrated its potential application in various material combinations and have been reviewed by Lambiase et al. [8]. However, there are no studies on the potential applications of PCBs in electronic materials, which are extremely thin and have complex internal structures. So far, the development of Friction Riveting has been limited to a minimum diameter-to-thickness ratio of less than 1:1 to ensure stable anchorage of the rivet into polymeric or composite plates [9].

Current feasibility studies on Friction Riveting have demonstrated that it is possible to plasticize 5 mm diameter aluminum rivets in 1.5 mm thick FR4-PCB substrate plates [10]. Further investigation also showed successful anchoring into FR4 substrates using 4 mm diameter aluminum rivets at temperatures above 300 ºC [11]. In both studies, the friction-riveted joints exhibited significant anchorage efficiency for thicknesses and anchorage depths of less than 3.0 mm, which represented the lowest anchorage depth, achieved using this technique.

Based on these promising findings, this study investigates the Friction Riveting process applied to 1.6 mm thick PCBs, made of FR4 substrates, four copper layers, and the absence of electronic components, using reduced-diameter aluminum rivets. The objective is to understand the influence of the process parameters and reduced rivet diameters on joint formation. For this purpose, process temperature, microstructure, and mechanical properties were analyzed. The results show the potential of the application range of Friction Riveting technology, especially for attaching electronic components to PCBs, which is desired in the world of electronic assemblies.

2 Materials

2.1 PCB material

A PCB material with a thickness of 1.6 mm and without electrical components and traces, provided by Panasonic Industrial Devices Europe GmbH (Germany), was used to fabricate joints. The PCB surface is coated with a green glossy solder mask at the top and bottom to protect against corrosion, as shown in Fig. 1(a). The PCB substrate consists of distinct regions, with some containing only fiber epoxy layers stacked, Fig. 1(b), and others featuring alternating copper layers and fiber epoxy, which were selected for joint fabrication, as shown in Fig. 1(c). The copper layers, each 35-µm thick, play a crucial role in connecting the electronic components to create circuits.

Fig. 1
figure 1

a Top view of the PCB with green glossy solder mask and without electrical components and conductors. Details of its distinct regions b with fiber epoxy layers stacked and c with alternating layers of copper and fiber epoxy

Full size image

Atintoh et al. [12] experimentally determined the mechanical properties of this type of PCB. The results are summarized as follows: the modulus of elasticity in the X direction is 26.0 ± 0.4 GPa, in the Y direction 23.34 ± 0.43 GPa, and the shear modulus in-plane was determined as 5.0 ± 0.10 GPa.

2.2 AA 2024-T351 rivets

Extruded rivets of AA-2024-T351 alloy, with diameters of 4 mm, 3 mm, and 2.5 mm, and a length of 40 mm, were selected, as shown in Fig. 2. The decision to utilize these specific materials was based on the successful outcomes of prior process development and validation efforts in Friction Riveting process development.

Fig. 2
figure 2

AA2024-T351 rivets of different diameters a 4 mm,  3 mm,  2.5 mm, b Microstructure in the extrusion direction of rivets

Full size image

Due to the extrusion process, the microstructure of the alloy exhibited elongated grains, as shown in Fig. 2b. This characteristic significantly influences strength and deformation behavior during Friction Riveting. The yield strength of AA 2024-T351 extruded rods is typically reported to be around 324 MPa (0.2% offset) [13], which is a crucial property for understanding how the material will deform during rivet formation. Higher yield strength indicates greater resistance to plastic deformation. Additionally, the hardness of the rivets was approximately 155 HV0.2 [10]. While hardness offers some indication of the material's resistance to plastic deformation, yield strength provides a more specific measure relevant to the friction riveting process. Furthermore, Table 1 lists the relevant thermal properties of AA 2024-T351.

Table 1 Thermal properties of AA 2024-T351 [14]
Full size table

3 Experimental

The joints were manufactured using custom-made Friction Riveting laboratory equipment (Loitz-Robotik, Germany). This equipment employed a friction welding spindle RSM410 (Harms & Wende GmbH, Germany) coupled with a pneumatic clamping system (DZF-50-25-P-A-FESTO, Islandia, NY, USA) to fix the overlapped material, see in Fig. 3. All joints were produced in the PCB region, which consisted of four copper layers, as shown in Fig. 1c.

Fig. 3
figure 3

Friction Riveting laboratory equipment showing the position of the welding head, clamping, and the X-Y movable table

Full size image

To prevent the welding of the rivet to the equipment worktable due to the low thickness of the PCB, four overlapping plates were fixed in the clamping system, see Fig. 4. However, the goal was to produce joints only on a single PCB plate.

Fig. 4
figure 4

PCB overlap configuration in the clamping system for joining

Full size image

Figure 5 illustrates the schematic process steps of Friction Riveting implemented on PCBs. The technique is based on frictional heating generated by the rotating rivet into the polymeric or composite parts, Fig. 5b. As the rivet rotates, its tip reaches a suitable temperature, causing it to soften. When the metallic rivet tip plasticize, a forging force is applied, Fig. 5c. This force deforms and mechanically anchors the rivet tip into the polymeric material, resulting in a consolidated joint, Fig. 5d.

Fig. 5
figure 5

Schematic description of the Friction Riveting process steps applied to PCBs: a Positioning the components to be joined, b Friction phase where the rotating rivet is rotated through the upper part of the PCB, c Forging phase involving the plastic deformation of the rivet, and d Joint consolidation

Full size image

Different combinations of joining parameters, such as rotational speed (RS), displacement at friction (DaF), and friction force (FF), were chosen to produce the joints. The selected joining conditions, as listed in Table 2, were determined based on a prior set of parameters known to produce well-anchored joints. This prior study involved 5 mm diameter AA2024-T3 rivets with 1.5 mm thick FR4 substrate [10]. Three replicate joints were manufactured for each condition to ensure statistical repeatability 3.

Table 2 Selected joining conditions, representing the different process parameters and rivet diameters employed
Full size table

Infrared thermography was employed to monitor the temperature during the joining process of the ejected flash material. Thermograms were captured using a high infrared camera (Image IR, InfraTec GmbH, Germany) and analyzed using IRBIS 3 software.

The joint formation related to the anchoring of the rivet into the PCB and the extent of microstructural changes were analyzed by embedding cross-section samples using digital optical microscopy (VHX-6000 series, KEYENCE America, USA). The cross-section samples were prepared following standard procedures, which included cutting the samples using a diamond wire cutting machine (DWS 250, Diamond WireTec, GmbH), embedding in cold mounting epoxy resin (EpoxyCure2, Buehler Ltd., USA), and surface grinding and polishing using dedicated equipment (Tegramin-30, Struers GmbH, Germany).

To quantify the anchoring effectiveness of the riveted joints, a combined approach of optical microscopy and a volumetric ratio (VR) calculation was employed. Optical micrographs provided precise measurements of parameters such as rivet tip width (W), penetration depth (H), and anchoring depth \(({D}_{p})\) within the PCB. Based on these measurements, VR was calculated using Eq. (1). This equation considers the interaction volume between the deformed rivet and the remaining plastic or composite material [15]. VR provides valuable insights into the joint's performance by capturing the effectiveness of the rivet's anchoring mechanism [16].

$$VR=\frac{{D}_{p} . [{W}^{2}-{D}^{2}]}{H .{ W}^{2}}$$
(1)

Furthermore, to evaluate the quality of the interface between the rivet tip and PCB, as well as any delamination and cracks in the anchoring zone (AZ), scanning electron microscopy (SEM, Quanta FEG 650, USA) was employed at 30 kV. To provide an electrically conductive thin film on the PCB for optimal imaging, the samples were sputtering with gold (Q150R ES, Quorum Technologies Ltd., England) at room temperature for 60s at a current of 65 mA.

The mechanical properties of the joints were evaluated via T-pull tensile testing using a universal testing machine (Zwick Roell 1484, Germany) at room temperature (23°C ± 2°C) and at a constant loading rate of 1 mm/min. To ensure uniform stress distribution in the PCBs during testing, a special clamping device was utilized, as described in detail by Altmeyer et al. [17]. For each joint condition, three replicate PCB samples with dimensions of approximately 7 mm x 3 mm x 1.6 mm were tested. The obtained ultimate tensile force (UTF), defined as the maximum force that a joint can withstand, was correlated with the VR. The failure modes of the joints after testing were visually inspected. The failures were identified based on the five types of failure findings by Rodrigues et al. [18].

4 Results and discussion

In this section, the influence of the Friction Riveting processing conditions and reduced-diameter rivets on the joint formation is discussed. These parameters are correlated with the evolution of process temperature, microstructural changes, and mechanical properties.

4.1 Aluminum rivets of 4 mm diameter

In the first part of this study, the process temperatures of joints produced under conditions C1 and C2, using aluminum rivets with a 4 mm diameter, were evaluated. Variations in DaF and FF values at a constant RS led to changes in heat generation and joint geometry. The results regarding the process temperature and geometrical properties of joints C1 and C2 are summarized in Table 3.

Table 3 Process temperature and geometrical features (penetration depth, H, and deformation width, W) of friction-riveted joints produced with 4 mm diameter rivets
Full size table

In the case of joints C1, the highest temperature recorded (368 ± 10 °C) resulted in a penetration depth of 1.9 ± 0.2 mm and deformation of 80% (W: 7.3 ± 0.3 mm) of the original rivet diameter. The selected process parameters caused significant penetration and deformation of the rivet tip in the PCB, as shown in Fig. 6a. On the other hand, for joints produced under reduced DaF and FF (C2), a lower processing temperature (293 ± 31 °C) was observed along with reduced penetration depth (H: 1.2 ± 0.5 mm) and deformation (W: 6.4 ± 0.7 mm) of the rivet into the PCB, as shown in Fig. 6(b). However, it is noted that for both sets of process parameters, the deformation of the rivet primarily occurred near the surface of the PCB, making it challenging to accommodate the volume of the deforming rivet within the PCB.

Fig. 6
figure 6

Representative joint geometries obtained for AA-2024-T351 rivets of 4 mm diameter and PCB material of 1.6 mm thickness, including the H and W measurements. a C1 (RS: 6000 rpm, FF: 4000 N, and DaF: 4 mm) and b C2 (RS: 6000 rpm, FF: 3000 N, and DaF: 1.5 mm). SEM images of C1 c-e and C2 d-f joints show delamination of the PCB layers and cracks between the PCB material and rivet tip

Full size image

Micrographs of the results of joint geometries are shown in Fig. 6. Delamination was observed between adjacent PCB layers, occurring between the epoxy and glass fibers, as well as between Cu layers and epoxy interface, followed by cracking around and at anchoring in both joint conditions. Such behaviors are more likely to occur in PCB when it is overheated or subjected to excessive thermal stresses [19]. This can cause the resin to detach from the glass fiber and Cu layers [20]. Figure 6c shows that this mechanism was more pronounced in the C1 joints, probably due to the local thermal stresses caused by excessive heat and severe deformation of the rivet. The lower incidence of delamination observed in Fig. 6e for C2 joints could be related to the lower process temperature, leading to less plasticization and deformation of the rivets into the PCB.

Several studies on PCB component soldering have reported delamination at interfaces, particularly in areas exposed to high temperatures. Furthermore, Verma et al. [21] showed that PCB delamination increases significantly with temperature. Xie & Sitaraman [22] observed that delamination predominantly occurred between epoxy glass-reinforced and copper layers. Figure 6(d) and (f) show the weak bonding mechanisms and cracks along the interface between the PCB substrate and rivet tip, indicating that the edges of the joints are susceptible to such issues. Because PCBs are made of glass fibers, copper, and epoxy resin, each with distinct elastic properties and coefficients of thermal expansion (CTE), temperature gradients can induce significant stresses at the interfaces, resulting in weak material bonding [23]. Another factor contributing to this mechanism is the deformed rivet area with discontinuous edges, which can also negatively influence the bonding behavior and consequently generate non-hermeticity and reduced bonding strength [24].

4.2 Aluminum rivet of 3 mm diameter

To mitigate delamination and cracking, reduced rivet diameters of 3 mm were used in combination with low values of FF (1600 N) and RS (4000 rpm and 3000 rpm), i.e., C3 and C4 joints. These conditions and the reduced rivet diameter were chosen to prevent high processing temperatures and severe deformation of the rivet tips in the PCB.

Table 4 shows that joints C3 and C4 reached process temperatures of 331 ºC and 341 ºC, respectively, which are lower than those of C1 owing to the lower RS and FF values used. However, the observed temperatures are higher than those in C2, which suggests that a higher DaF (4 mm) increased the contact area, thus increasing the heat generation, as also observed by De Proença et al. [16]. C3 and C4 exhibited similar joint rivet penetration and deformation, primarily at the PCB surface. Figure 7 shows the representative geometries of joints C3 and C4. The joints produced under C4 exhibited buckled rivet formation owing to rivet deflection from the center axis of the spindle. This was observed in the initial stage of the process when the rivet touched the upper surface of the PCB. The buckling of the rivet can be explained by the excessive rivet load caused by the weight of the spindle, which occurs mainly at low rotational speeds (3000 rpm). Ma et al. [25] also addressed the issue of rivet buckling in friction-based hybrid joining processes, including Friction Riveting, and linked it to either insufficient heat input or excessively softened lower composite plates.

Table 4 Process temperature and geometrical features (penetration depth, H, and deformation width, W) of friction-riveted joints produced with 3 mm diameter rivets
Full size table
Fig. 7
figure 7

Representative joint geometries obtained for AA-2024-T351 rivets of 3 mm diameter and PCB material of 1.6 mm thickness, including H and W measurements. a) C3 (RS: 4000 rpm, FF: 1600 N, DaF: 4 mm) and b) C4 (RS: 3000 rpm, FF: 1600 N, DaF: 1.5 mm). SEM images of the joint formation obtained from C3 (c-e) and C4 (d-f) showing delamination of PCB layers and cracks between the PCB and rivet tip

Full size image

Figure 7(c) and (e) reveal that both joints exhibited delamination at the interface between the solder mask and the first layers of the PCB substrate. This mechanism was more pronounced in the C3 joints, Fig. 7(a), where the rivet leg deformation was directed upward. This directional deformation likely resulted in more delamination compared with C4, where the rivet leg deformation was towards the bottom part of the PCB, as shown in (Fig. 7b).

Furthermore, in Fig. 7(d) and (f), there are indications of the onset of cracking in the deformed parts of the rivet at the bottom of the anchoring zone. It can be tentatively concluded that the reduction in the rivet diameter (from 4 mm to 3 mm) and decrease in the rotational speed (from 6000 rpm to 4000 rpm) and FF (from 4000 N to 1600 N) did not lead to a proportional reduction in the heat generation. Therefore, delamination and cracking were still observed on the PCB.

Galińska & Galiński et al. [26] highlighted the similarities between delamination, cracking, and fiber/matrix damage that occur in composite riveting and drilling processes. They also suggested that using carbon nanotubes, which are already employed in electronic applications, could potentially help prevent delamination and other defects during riveting and drilling processes.

Nevertheless, as demonstrated in previous studies on joint formation in Friction Riveting, the anchoring efficiency related to the deformation achieved (PCB thickness 1.6 mm, 3 mm diameter) exceeds the required needs in terms of mechanical anchoring [27]. Therefore, it is reasonable to explore the potential of this technology to achieve similar or superior rivet anchoring using an even smaller rivet diameter.

4.3 Aluminum rivet with a diameter of 2.5 mm

To enhance the joint quality by addressing delamination, cracking, and rivet anchoring concerns, reduced rivets with 2.5 mm diameter were used. The parameters RS (4000 rpm) and DaF (4 mm) used in C3 for a 3 mm diameter rivet, were applied to 2.5 mm diameter rivets (C5), along with a decrease in FF from 1600 N to 900 N.

Table 5 shows that the C5 joints reached a process temperature of approximately 285 ± 25 ºC. This condition facilitated rivet penetration of 1.2 ± 0.2 mm into the PCB, while reducing deformation (W: 2.8 ± 0.2 mm), as seen in Fig. 7. Despite the lower diameter-to-thickness ratio (D/t: 1.56), the selected combination of process parameters resulted in a relatively low temperature, which was insufficient to ensure the anchoring of the rivet to the PCB.

Table 5 Average process temperature and geometrical features (penetration depth, H, and deformation width, W) of friction-riveted joints produced with 2.5 mm diameter rivets
Full size table

Figure 8(b) illustrates the presence of cracks along the anchoring zone between the PCB and rivet tip, and Fig. 8c shows fewer instances of delamination within the PCB layers. This can be directly attributed to the lower process temperature and minimal deformation of the rivet.

Fig. 8
figure 8

Representative joint geometry obtained for AA-2024-T351 rivets of 2.5 mm diameter and PCB material of 1.6 mm thickness, including H and W measurements (joint obtained under C5: RS: 4000 rpm, FF: 900 N, DaF: 4 mm). SEM images of joint formation obtained under C5 showing a) cracks between the PCB material and rivet tip and b) delamination in adjacent PCB layers

Full size image

Overall, the results demonstrate the impact of reducing the rivet diameter, a process known as downscaling, on the formation of friction-riveted joints. The series of adjustments in the process parameters resulted in notable changes in the process temperature and the deformation characteristics of the joints. The lower process temperature, which was initially perceived as an advantage, raised concerns regarding the anchoring efficiency of the rivet within the PCB substrate for smaller rivet diameters.

This investigation highlights the important balance between rivet diameter, process parameters, and joint performance. The proper selection of friction riveting parameters is crucial to achieve a sufficient temperature at which the rivet becomes plasticized and firmly anchored within the PCB. However, it is important to note that high temperatures can induce PCB overheating, leading to changes in material properties and consequent delamination and crack formation. These issues may be mitigated by considering PCB substrates with greater resistance to elevated temperatures (> 300 ºC) or by selecting rivets with plasticizing effects at lower temperatures (< 300 ºC).

Furthermore, the parameters chosen for the reduced rivet diameter showed a slight reduction in cracking and delamination but failed to achieve satisfactory anchoring of the rivet within the PCB. This can be attributed to the high diameter-to-thickness ratio (D/t>1) and insufficient level of heat generation, particularly for 2.5 mm diameter rivets.

Further research in this domain should aim to balance, and optimize rivet anchoring while reducing rivet diameter, thus obtaining superior joint quality and reliability.

4.4 Mechanical properties

The Friction Riveting process applied to 1.6 mm thick PCBs using reduced-diameter aluminum rivets demonstrated varied mechanical properties based on the process parameters. To understand the impact of the rivet diameter on joint strength, T-pull tensile tests were carried out on joints obtained with different rivet diameters (C2, C3, and C5). Table 6 lists the average UTF values for these joints, along with their VR.

Table 6 Selected joining conditions and rivet diameters, their UTF, and VR values
Full size table

Notably, C3 joints exhibited the highest UTF values, 276 ± 18 N, followed by C2 with 183 ± 12 N, and C5 with a relatively modest 33 ± 12 N. C3 joints were assembled with 3 mm rivet diameters, whereas C2 joints had a diameter of 4 mm, and C5 a diameter of 2.5 mm. This outcome is expected, as C3 joints demonstrated superior anchorage efficiency, as quantified by a high VR (0.41), and showed fewer instances of delamination and cracking on the PCB substrate.

Figure 9 shows a common failure mode among all joints (C2, C3, and C5) by complete rivet pull-out, resulting in a cavity with a similar diameter to the deformed end of the rivet tip, as previously described by Rodrigues et al. [18] and Vilas Boas et al [10]. This failure mode can be attributed to the lower deformation of the rivet, particularly near the surface, and the formation of cracks in the anchoring zone. These factors potentially compromise the bond quality at the interface between the PCB and rivet anchoring zone, facilitating the detachment of the bottom of the rivet tip from the PCB substrate under load. This failure mechanism aligns with the observations made by Blaga et al. [5] and typically leads to a moderate to poor UTF. The fracture was often initiated around the anchoring zone and propagated through the PCB material, weakening the mechanical bond and leading to a complete pull-out of the rivet. This behavior underlines the importance of optimizing the deformation of the rivet tip and minimizing cracks in the anchoring zone to enhance joint performance.

Fig. 9
figure 9

Appearance of the joints after T pull tensile test C2 (4 mm rivet diameter), C3 (3 mm rivet diameter), and C5 (2.5 mm rivet diameter) with corresponding extracted rivets placed over the joined plates. Friction riveted joining area marked with a dotted circle

Full size image

The results suggest that the relationship between rivet diameter and joint performance is more complex than initially expected. While larger rivet diameter and joint performance may improve joint strength by increasing the bonding area, they can also introduce s thermal effects, which affect the formation of cracks and delamination, as mentioned previously. The findings demonstrated the potential for obtaining robust joints by adjusting the diameter of the rivets and reducing process temperatures, as exemplified by UTF values for C3 joints.

As friction assists in joining methods of polymer material, factors such as low thermal conductivity and crystallinity tend to affect mechanical properties. The interaction of thermal-mechanical and material flow behaviors observed in this study was complex due to the high process temperature and significant shear forces involved. These factors significantly affected the formation of cracks and delamination, which, in turn, influenced the mechanical properties of the joints. Thus, by refining Friction Riveting parameters, exploring advanced materials, and optimizing heat generation, the boundaries of this joining technology can be pushed to obtain joints with good strength and reliability, suitable for application in the electronic industry.

5 Conclusion

This study explored the feasibility of the Friction Riveting process for 1.6 mm thick printed circuit boards (PCBs) using reduced-diameter aluminum rivets. In this context, this study investigated the effects of process parameters and rivet diameter on the process temperature, joint formation, and mechanical properties of friction-riveted joints. Employing a 4 mm rivet diameter, extensive delamination, and cracks were observed along the interface between the PCB substrate and rivet tip. These issues were largely attributed to high-temperature processes, reaching up to 300 °C, and significant deformation of the rivets, particularly near the PCB surface. Reducing the rivet diameter to 3 mm did not proportionally reduce heat generation, therefore, delamination and cracking were still observed in the PCB joints, whereas the mechanical properties increased due to the improve anchoring efficiency. Further downsizing to 2.5 mm rivet diameter, despite reducing delamination, revealed challenges with anchoring efficiency and introduced new cracks between the PCB and rivet tip, negatively affecting the mechanical properties of the joint. Ultimately, joints obtained with 3 mm rivet diameter achieved the highest ultimate tensile force, emphasizing the significance of optimizing the diameter-to-thickness ratio for effective joint performance in thin PCBs. The mechanical properties of Friction Riveted joints in PCBs depend on the anchoring efficiency of the deformed rivet and variables, such as delamination and cracking within the composite plate. This technique demonstrates substantial anchoring potential, resulting in rivet deformations that exceed the requirements for such thin materials. Therefore, diameter reduction is necessary, along with gaining a deeper understanding of joint formation, delamination, and cracking for these reduced geometries. The quality of joints produced by the Friction Riveting process using reduced rivets depends on several factors, including the selection of appropriate process parameters, rivets with lower plasticizing temperatures, and PCB with better heat resistance. Careful consideration and optimization of these factors can lead to adequate joint anchorage while minimizing the risk of delamination and cracking.