1. Introduction
For the production of mechanical joints by means of the technology of joining by forming with auxiliary joining elements [
1], self-pierce riveting (SPR) has been the preferred mechanical joining technology over the years to produce mechanical interlockings between two or more geometries to be joined [
2]. Form-fit joints are therefore created by employing those semi-tubular rivets as unremovable mechanical fasteners (
Figure 1a).
Lightweight construction demands for multi-material design, which in turn demand the utilization of versatile joining processes that circumvent the metallurgical incompatibilities, provide a low heat input and can provide an adequate level of flexibility [
3]. At the same time, a versatile process chain required for product manufacturing with this kind of joining by forming processes [
4] encompasses the material combination (joining suitability), the design and layout of the joints (joining safety), and the adaptability and predictability of the joining process (joining possibility). Thus, it becomes necessary to reduce the amount of joining parameters and configurations, as well as the tool variants, that depend on each material combination, and present themselves as the limitations of conventional self-pierce riveting [
5].
The application of SPR to aluminium sheets of different alloys have been extensively investigated. From those studies, the effect of rivet and die shapes on the joint properties have been studied [
6,
7,
8]. The analysis of the influence of the sheet thickness on the fatigue performance of aluminium SPR joints [
9] concluded that the fatigue life of the SPR joints was extended with larger sheet thicknesses, where failure tends to occur at the bottom sheet along the joint button. In addition, the joints made from thick aluminium sheets have good mechanical stability, whereas the joints made from thin aluminium sheets have poor durability in corrosive environments [
10]. For the same aluminium alloy utilized in this research, the influence of the surface conditions on the SPR joint strength was assessed [
11], from which it was concluded that the increase of the surface roughness allows us to increase the shear strength of the SPR lap joints. Different strategies have been developed to accommodate the joining of multi-material structures and advanced materials [
12]. The corresponding joint failure mechanisms under different mechanical loading conditions [
13,
14] and the joint corrosion issues [
14] have been carefully analysed, from which it can be concluded that there is still room for improvement.
An innovative self-pierce riveting process consists of double-sided self-pierce riveting (DSSPR) which makes use of a tubular rivet with chamfered ends that are capable of producing hidden joints between sheets placed over each other (
Figure 1b). The tubular rivets are placed in-between the sheets to be joined and forced through them, while their ends are flared to create a mechanical interlocking [
15,
16]. In comparison with SPR, some important advantages are presented by DSSPR such as: the ability to join different thicknesses without limitations for larger thicknesses since is not necessary to tear up the pierced sheet as it happens for SPR; the reduced to non-existent material protrusions in the sheet surfaces; smaller levels of controlled deformation and reduced damage and stress–strain levels in the materials; the joint not being exposed to galvanic corrosion or other elements; the simpler geometry of the tubular rivet and process parameters; dedicated tools no longer being needed other than flat compression plates.
The validation of the DSSPR joining technology has been performed with tubular rivets of stainless steel AISI 304 and aluminium AA5754 sheets, to investigate the working principle and the geometric scalability of the tubular rivets. Recently, the chamfered ends of the rivets have been optimized to improve the rivet penetration and final morphology, and therefore increase the overall joint strength of the mechanical connection [
17].
Since there is no upper thickness limit in DSSPR contrary to what it is observed for SPR, sheets with a thickness of 5 mm were originally tested [
16], and more recently, sheets with a thickness of 1.5 mm were investigated [
18]. From this last study, it was concluded that the external diameter of the tubular rivet can remain constant, although its height and thickness need to be modified in order to match the thickness of the sheet. This means that the initial rivet thickness needs to be lower than the smaller sheet thickness and the initial rivet height needs to be at least equal to double the smaller sheet thickness, in order to produce a proper mechanical interlocking.
From an industrial point of view or when demanded from the material combination, some strategies may need to be developed. For instance, the authors have shown that while keeping all the other parameters constant, the chamfered angle of the tubular rivet can be modified at each end of the rivet, to account for the different resistances to penetration of the materials from the two sheets [
19]. However, when those differences are too high (for example, the combination of a PVC sheet with an Aluminium sheet), an asymmetric and reduced mechanical interlocking is formed at the harder sheet material. For those situations, another strategy is then presented in that same work [
19] to produce a symmetrical joint with a good mechanical interlocking in both sheets: the tubular rivet is pre-riveted in the harder sheet by a dedicated compression tool and then subsequently pressed against the softer sheet to produce the joint between the two sheets. This solution allows us to keep the same chamfered angle of the rivet in both sides, thus avoiding possible errors during the positioning of the rivet, and also allowing us to join sheets with very different mechanical resistances, which is a step forward towards an adequate industrial implementation.
Another possible solution is the introduction of flat bottom holes in the strongest sheet by means of machining or forming, in order to correctly position the tubular rivets before piercing. This promotes the penetration in harder materials and their subsequent assembly to softer materials by means of the opposite end of the tubular rivet [
20]. At the same time, the excess volume of sheet material during the rivet penetration will flow through the empty spaces of the flat-bottom hole, thus eliminating any protrusion above the sheet surfaces. In terms of industrial implementation, this solution allows us to easily replicate and inspect the joining process, since the position of the rivet is well-defined and no additional joining stages or changes to the process parameters are needed. This way, it is possible to join different materials in one single joining operation while the rivet is secured in the correct position, which is essential when the sheets to be joined are not horizontal.
Therefore, the objective of this work is to combine the different previous strategies and compare their performance when subjected to static loads. To support the investigation, numerical predictions are employed to analyse the mechanics of the deformation and the stress–strain levels for the different strategies and modifications introduced. Along with the numerical analysis, different specimens are produced and subjected to lap shear strength tests to determine the performance produced by each modification. This will allow us to define the strategy to be followed in accordance with the load and energy requirements for both the joining process and the intended application, towards a closer industrial implementation of DSSPR. Different compromises are made with each joining strategy, although the introduction of flat-bottom holes in both sheets is able to provide a similar performance to conventional DSSPR joints without holes, that offers the best performance among all strategies. The utilization of pre-riveting follows the introduction of flat-bottom holes in terms of performance and its applicability regards the utilization of sheets of very different strengths or thicknesses.
3. Results
Comparisons between the Different Strategies
Through the combination of finite element modelling and experimentation, it was possible to identify the mechanics of deformation for the different strategies analysed. Generally, the thickness of the tubular rivet increases along the deformation due to compression and strain hardening of the sheet and rivet, the latter being responsible for promoting a combined piercing and flaring of the tubular rivet, which results in the formation of a mechanical interlocking [
16]. As the rivet penetrates through the sheets, the sheet material flows over the rivet to accommodate the volume of the rivet being pushed into the sheets. For plain sheets (
Figure 5), the material flow is very constrained and as a result the stress levels increase as well as the joining force.
After the tools were removed, a very small protrusion is visible in the top of the sheet surfaces which results from the elastic recovery of the materials from those sheets (refer to the arrows in the photograph in
Figure 5) in the opposite direction through which they were forced while the rivet was pushed through them. This causes the rivet ends to curl for compensating the constraint caused by the strong contact between the two overlapped sheets at the centre of the joining region.
To overcome the constraints in material flow, flat-bottom holes were introduced in the sheets to allow for positioning and aligning both the rivet and the two sheets to be joined. As seen in
Figure 6, which discloses the experimental and finite element predicted cross-sections of two joints with flat-bottom holes of different depths
, the amount of unfilled volume between the outer rivet wall and adjacent sheet material (that is normally observed for conventional DSSPR) is significantly reduced.
The gap created by the flat-bottom holes allows us to reduce the upward elastic recovery movement, which in turn results in even smaller protrusions than those created by conventional DSSPR. This happens because the pressures at the sheet surfaces are highly reduced with the introduction of the holes and the sheet material can flow towards the inside of the rivet diameter without making contact with the opposite sheet. The resulting form-closed joining mechanism is now responsible for a smaller mechanical interlocking than in conventional DSSPR, although a stronger force-closed mechanism develops at the contact interface between the rivet and the flat-bottom hole due to the residual normal pressures after the unloading of the tools. Depending on the depth
of the flat-bottom hole, the radial stresses developed will prevent tangential movement due to friction and may produce higher mechanical performances than other strategies, as it will be seen in
Section 4.
The finite-element-predicted distributions and experimental results in
Figure 6 allow us to discuss the plastic deformation for different variations of the depth
. For a depth
= 1 mm (
Figure 6a), the rivet is more restrained by the material from the sheet which will provide a sounder joint that is able to support larger tangential movements due to the development of larger levels of radial stress. Nevertheless, it was previously seen by the authors [
20] that, for sheets made from materials of very different mechanical strengths, these smaller depths may result in the rivet being mainly pierced through the softer sheet without producing a proper mechanical interlocking if the chamfered angle of the rivet is not properly controlled. A larger depth
of 2 mm (
Figure 6b) is not able to offer enough constriction between the rivet and the sheets, which will result in the rivet detaching more easily from the sheets as it is forced to unbend during its lap shear destruction. The radial pressures are very low in comparison with a depth
of 1 mm which demonstrate the lack of rivet penetration into the sheets. Therefore, a larger increase of the depth
will not bring any advantages and may compromise the integrity of the sheet while limiting the industrial application of this strategy for smaller thicknesses.
Regarding the strategy of pre-riveting the rivet in one of the sheets, a dedicated compression tool forces the tubular rivet into the sheet and creates a mechanical interlocking between those two elements, as seen in
Figure 7.
Then, the other sheet is placed over the opposite free end of the tubular rivet and compressed until a point where the sheets contact with each other, and the joint is produced. Generally, the pre-riveting should be applied to the harder sheet and/or thicker sheet in order to guarantee an adequate penetration of that sheet, since after the pre-riveting operation, the already joined and strain-hardened region will provide a better resistance to deformation and force it to occur mostly at the free region of the opposite rivet end.
Two modifications were analysed for the pre-riveting strategy: one of them consists of joining the pre-riveted sheet with a plain sheet (
Figure 8a), while the other consists of joining the pre-riveted combination with a sheet having a pre-drilled hole with a depth
= 1 mm (
Figure 8b), that, as seen before, is able to improve the sheet material constraint around the tubular rivet. In both cases, the radial pressures to which the pre-riveted sheet was subjected increased mainly at the inner rivet diameter, where the sheets were compressed against each other (refer to the comparison between
Figure 7 and
Figure 8).
In comparison with the results previously obtained without the pre-riveting stage, the mechanical interlockings are now larger (0.699 mm for the case of the plain sheet and 0.521 mm for the sheet with a depth
of 1 mm), mainly due to the fact that the other two elements (rivet and pre-riveted sheet) will act as a single element and their strain-hardened material regions will concentrate the deformation in the opposite ends of the rivet where the joining with the sheet occurs (refer to the upper sheet in
Figure 8a,b). Despite the differences in the values of mechanical interlocking and in the joint morphology of these two variations with a pre-riveting stage, the distribution of radial stress is similar, which may justify the similarity of their destructive performances, as it will be later seen.
Overall, the agreement between the finite element predictions and their respective experimental specimens is suitable for all the different strategies. Minor variations are attributed to small differences in dimensions resulting from the manufacturing process of the different geometries and/or to the elastic recovery of the materials after being cut lengthwise to reveal the cross-section of the joint, and allow a comparison between the numerical and experimental results.
5. Conclusions
Different strategies were analysed to produce joints in overlapped sheets with DSSPR, making use of the strong advantages of this joining technology while ensuring the conditions for its industrial implementation in a wide range of scenarios.
The introduction of flat-bottom holes in the sheets ensures both positioning and alignment of the rivets, while eliminating any protrusions above the sheet surfaces. As a result of the gap created by the holes, the joining forces and energies are lower than for conventional DSSPR, while their performances are very similar. However, the depth of the flat-bottom holes cannot be so high because it can compromise the performance of the mechanical connection or may even not be feasible for thinner sheet thicknesses. In the latter case, pre-riveting of one of the sheets (generally the harder and/or thicker sheet) should be employed instead to ensure a proper riveted joint.
For the pre-riveting strategy, two iterations can be utilized without any relevant differences in terms of performance, other than the advantages that arise from the simplicity of placement of the opposite rivet end when a localized flat-bottom hole is already present in the opposite sheet. Nevertheless, if the opposite sheet material is much softer than the pre-riveted sheet material, a flat-bottom hole may create a weaker region in the softer sheet that can compromise the joining process. The selection of the pre-riveting strategy comes at a cost of a slightly reduced destructive performance and the need to have an additional stage other than the single stroke in which the other strategies are produced. Therefore, this strategy is more suited for different material and thickness combinations where a greater or lesser penetration of the rivets into the sheets may be desired.
In conclusion, the guidelines established during this work help to create the conditions for selecting a suitable joining strategy with DSSPR depending on the material and geometry specifications.