Influence of local heat treatment of rivets on the joint formation of a versatile joining process

An efficient lightweight construction method is the combination of different materials in order to adapt the structure to the applied load. To join these multi-material structures mechanical joining technologies are applied. However, the rigid tooling systems cannot be adjusted to changing boundary conditions which is why new, versatile joining technologies are required. In the versatile self-piercing riveting (V-SPR) process presented in [1] different material combination are joined by using a multi-range capable rivet. The rivet head is formed onto the respective thickness of the joint by an outer punch. In order to punch thru the upper sheet a great rivet hardness is required whereas a lower hardness is required for the subsequent forming of the rivet head. To achieve a combination of these requirements, this study investigates a local heat treatment of the rivet. The aim is to determine the feasibility of such a heat treatment as well as to investigate the influence on the joint formation.


Introduction
Due to the ecological and social developments of the recent years as well as the climate change, greater focus is being placed on resource-saving technologies to reduce CO2 emissions significantly.Also, increasing efficiency is an objective in many areas [2].A far-reaching change is also occurring in the mobility sector.As this sector is a major contributor to carbon dioxide, various strategies, such as the use of lightweight constructions are being pursued to reduce these emissions.Multi-material design, in which materials with different properties are combined, also belongs to this type of construction technology.However, due to their properties, the different materials can no longer be joined using conventional thermal joining methods, For this reason, mechanical joining processes such as selfpiercing riveting are increasingly being used.

Self-piercing riveting
For joining two or more sheet with auxiliary joining element, self-piercing riveting (SPR) is of particular interest.The process sequence can by divided into four stages (see Figure 1) and begins by fixing the sheets on the die.Subsequently, the rivet penetrates the punch-sided sheet.The resulting slug is stored in the rivet shank.The further feed leads to a plastic deformation of the rivet, which expands radially into the die-sided sheet.Hereby, form-fit and force-fit joint is created [3].Several works have focused on the extension of the process limits of the SPR process [4].These focus either on the rivet geometry [5] or the process kinematics [6].Also the combination of rivet geometry and kinematics were investigated as shown in [7].However, the strategies for increasing the process limits described above IOP Publishing doi:10.1088/1757-899X/1307/1/012009 2 often lead to increased process complexity and a great degree of specialisation of the process due to the rigid tools used.

Versatile self-piercing riveting
With regard to versatile process chains joining processes have to be able to react to changing boundary conditions.The V-SPR process presented in [1] allows these changes by combining an extended tool actuator technology with a multi-range capable rivet.To form the joint, a two-piece punch is used.The process itself relies on the process principle of the conventional SPR process (see Figure 2).Using the inner punch the rivet is pushed into the parts to be joined whereby an form and force-fit joint is created.The rivet head is formed after setting using the outer punch.Subsequent forming allows flexible reaction to changing sheet thicknesses.The multi-range capable rivets are currently produced by machining.As with the production of conventional SPR rivets, they are subsequently heat treated and tempered to a hardness of 480 HV.The friction and corrosion resistance between the rivet and the joining parts is improved by using an ALMAC coating.
Figure 2. Process sequence of the versatile self-piercing riveting process [1] Previous studies on the V-SPR process have mainly focused on the process development and joint formation analysis.Especially for aluminium joints, large material thickness ranges could be joined with a single rivet-matrix combination [1].In addition, the preferred joining direction thin to thick was identified with the help of a numerical analysis [8].The joint formation when using multi-material joints could also be proven [9].However, it was shown that the use of steel as the die-sided joining part requires a large deformation in the head area, as the rivet cannot penetrate deep into the joining parts.This increases the risk of cracking in the setting head during the formation of the rivet head.For this reason, the use of a local heat treatment instead of the current heating and tempering process is to be investigated in this study.A ductile head could also improve the forming behaviour of the rivet head to the punchsided joining part.

Experimental details
For the combination of a high shaft hardness with a ductile and formable rivet head, the setting of graded rivet properties is required.A possible method to achieve this is the local heat treatment of the rivet by induction.In this study, the feasibility of such a local heat treatment is investigated.The following section discusses the applied material and the general experimental setup in more detail.

Test materials
The multi-range-capable rivet was manufactured using C35 B+Cr steel.Due to its good ductility and strength properties even in the hardened state this material is well-suited for use as rivet material.The base material is a cold-drawn wire with a thickness of 10 mm.The rivet geometry is produced from this material by means of machining processes.Joining tests are carried out to prove the feasibility of local heat treatment of the rivet.For this purpose, joining parts of the aluminium material EN AW-6014 in temper T4 are used.This aluminium alloy is often used in crash-relevant components in the automotive industry due to their good formability and high energy absorption.The chemical composition of the rivet as well as the sheet material is shown in Table 1.During the entire hardening process, the temperature of the rivet element is measured using the pyrometer positioned above the induction coil.A heating time of 0.2 seconds was specified for all tests carried out.The heating time is followed by a holding time at the specified temperature.During the holding phase, the temperature is regulated using the pyrometer.The temperature is varied in the different test series.After the holding time, the riveting element is rapidly quenched to room temperature within a few seconds using a defined air volume flow.
The feasibility of the local heat treated rivets is analysed using a joining system with an extended tool-actuator technology.It was presented in detail in [1].In order to create the joint, the inner punch sets the rivet.Subsequently the outer punch forms the rivet head onto the respective thickness of the punch-sided joining part.To produce these specimens the joining speed was set to 5 mm/s.The defined displacements for the inner and outer punch were kept constant for all joining tests.In order to define the process parameters specimens with the dimensions 45 mm x 45 mm are used.The joint is set in the geometric centre of the joint.
The specimens are examined with regard to their joint formation, the resulting microstructure and the ensuing hardness after joint formation.Vickers hardness testing was conducted using a hardness testing machine KS 30 S and according to DIN EN ISO 6507-1.

Results and Discussion
In order to prove the feasibility of the local heat treatment of the versatile rivet geometry and to be able to determine the influencing parameters on the hardening process, first experimental tests were carried out.The results of the hardening tests are shown in Figure 4 compared to a conventionally hardened rivet which consists of a hardness of 480 HV (Reference).Generally, it is shown that the hardening of the rivet is possible by means of induction heating.The local heat treatment including the following quenching process leads to an increased hardness in the shank area of the rivet.In the rivet foot area a hardness of 600 HV to 700 HV is reached.It can be assumed that in this area, due to the high hardness and the rapid cooling rate as a result of the subsequent quenching process, a completely martensitic structure has been formed.The hardness decreases with increasing distance from the rivet foot towards the rivet head.The transition range between the hardness levels is relatively short.The initial hardness of the base material is present in the rivet head.
The position of the induction coil (middle and top, Figure 3, e) and the heating temperature used (860 °C, 900 °C, 950 °C; measured using the pyrometer) have a decisive influence on the hardened area of the rivet shank.The examined temperatures do not correspond to the austenitization temperature (850 °C) of the rivet material due to the short heating rate (0.2 seconds).If the coil is positioned at the top position, less heat is applied to the rivet and a smaller area is hardened.Increasing the temperature while remaining at the top coil position can increase the hardened area.If the coil is positioned in the middle, the hardening temperature can be lowered to obtain the same hardening result (Figure 4 II and III).The increase in temperature also leads to an enlarged hardened area for a middle coil position.However, the differences in hardened areas are no longer significant when comparing the hardening temperature of 900 °C and 950 °C.This effect can be attributed to the use of copper as a rivet holder material during this test series.The good heat conduction properties of copper ensure rapid heat dissipation from the rivet head, which is placed onto the rivet holder, into the copper material.The heat introduced into the rivet shank, where the coil is positioned is therefore not sufficient to heat the transition area between the rivet shank and the rivet head as well as the rivet head sufficiently.For this reason, the following quenching process after the heat treatment does not result in a microstructural transformation.All the hardness parameters examined show that the rivet head does not undergo any change in hardness.The basic ductility of the material is maintained.At the same time, this can lead to deformation of the rivet head as a result of the high joining forces when setting the joint.In order to be able to exclude this, the joint formation was investigated experimentally.The result is shown as an example for a rivet with hardness parameters IV in Figure 5.It can be seen, that the hardness of the rivet shank is sufficient to punch through the punch-sided joining part.However, the shank collapses at the point where the hardened area transitions to the non-hardened area.The strength of the material is not great enough to absorb the joining forces, which increase exponentially during the joint formation.As a result, the inner punch, which introduces the load into the rivet, embosses into the rivet head.The good ductility of the material improves the forming of the rivet head onto the punch-sided joining part using the outer punch.The aim of the local heat treatment is thus to harden the shank along its entire length to prevent collapsing during the formation of the joint.Since the rivet holder material copper prevents an increased hardening range in the shaft area due to its good thermal conductivity, chamotte was used as the rivet holder material in further tests.The lower thermal conductivity aims to prevent the heat from flowing out of the rivet head in order to achieve continuous hardening of the shank.
In order to identify the effects of the local heat treatment, the joints made with chamotte as rivet holder material were examined metallographically.An exemplary result is shown in Figure 6 for a hardening temperature of 860°C.The ferritic basic structure of the material is visible in the central area of the rivet head (Figure 6, a).The ferrite appears as white round grains in the micrograph.The individual grains are delimited from each other by the black grain boundaries.In addition, the grooves resulting from the manufacturing process by means of drawing are visible.In this area, the microstructure was not affected by the heat treatment.
In the rivet head protrusion (Figure 6, b), which is formed onto the joining partner on the punch side, a mixed microstructure of ferrite (white) and pearlite (black) is found.The microstructural transformation here shows that the induced heat has penetrated into the rivet head.Since only pearlite is present here in addition to the ferrite, the hardness has not increased too much.Sufficient ductility, which is required for the forming process, is present.
In the transition area between the rivet head and the rivet shank, seen in Figure 6, c), a mixed structure of ferrite, pearlite and bainite is present.In parts, individual martensite needles may also be present.Due to the mixed structure, a high hardness is already present in this shank area, which prevents the shank from collapsing when the rivet is set.
In the rivet foot (Figure 6, d), a martensitic structure is found exclusively.The individual martensite needles, which spread from one side of the grain to the other, are clearly visible.This is also the area with the greatest heat input since the induction coil is placed here.As a result of the martensite, the highest hardness values are found in the area of the rivet foot and ascending to the rivet head.
In addition to the analysis of the microstructure, joining tests were carried out for the locally heat treated rivets with chamotte as rivet holder material for different temperatures.Again, after the heating (0.2 seconds) and holding time (3.0 seconds) a quenching process is carried out using a defined air volume flow.The results are shown in Figure 7.In addition, the hardness distribution is shown next to the micrograph.Despite the use of chamotte, a heating temperature of 800 °C is not sufficient to form a proper joint.Again, collapsing of the shank occurs.However, if the temperature is increased at the same induction coil position (middle), the shaft collapse is prevented.The prevented heat dissipation also provides a softer transition between the hardness regions in the shank.The resulting maximum hardness does not change.
The evaluation of the hardness reveals that the use of chamotte leads to an increased hardness in the rivet head area.This is caused by the accumulating heat as a result of the rivet holder material used.At a hardening temperature of 900 °C, a hardness of 520 HV to 600 HV is obtained in the head.The high hardness of the entire rivet suggests that a pure martensitic structure has been produced in large areas.The great hardness has in turn a negative effect on the formation of the rivet head.Equal punch displacements result in a small gap between the rivet head and the punch-sided joining part.If a hardening temperature of 860°C is used, the result is a head hardness of 320 HV to 420 HV, which ensures a good forming behaviour of the rivet head.At this hardening temperature, the hardness of the base material is still present in the centre of the rivet head, which suggests that no microstructure transformation has occurred here.
In general, the hardness of the shank prevents the spreading of the rivet in the pure aluminium joint.At both 860°C and 900°C, only a small interlock is achieved.In addition, a low minimum die-side material thickness is formed in both joints.In principle, the results showed that local heat treatment of the rivet element can be used to achieve a joint formation.The advantage of this procedure is that little energy is required to harden the rivet element.Complex hardening and tempering processes in large furnaces can therefore be partially avoided, which can have a positive effect on ecological aspects and the overall energy consumption.In addition, the local heat treatment of the rivets can increase the application range of the multi-range capable rivets.The increased design freedom therefore allows the increase in the degree of lightweight construction, as different rivet elements can be avoided and shorter rivets can be used due to the multirange capability and the subsequent forming of the rivet head.

Conclusion
In the study presented the local heat treatment of multi-range capable self-piercing rivets was investigated.In the process sequence, the rivet head is formed to the respective thickness of the joining partners after the joining process.This subsequent forming process requires a certain ductility.However, a corresponding hardness is necessary for the punching of the punch-sided joining partner.The local heat treatment aims to influence the properties of the rivet in such a way that a hard rivet shank is combined with a ductile rivet head.The heat treatment was carried out inductively with the induction coil placed around the rivet shank.After reaching the hardening temperature and the expiry of the holding time, a quenching process using a defined air volume flow was carried out.Basically, it was shown that both the position of the induction coil as well as the hardening temperature used influence the hardness distribution in the rivet.
During heat input, the rivet head was placed on a rivet holder.Copper and chamotte were tested as holder materials.It was found that copper as material prevents the complete hardening of the shank.When using chamotte, the heat can rise higher in the rivet, which prevented the shank from collapsing during joint formation.However, the high hardness due to induction heating inhibits the spreading of the rivet and thus the joint formation in pure aluminium joints.The reduced hardness in the rivet head area could improve the forming of the rivet head.Future work aims to include a tempering process in the process sequence in addition to the inductive hardening process.Furthermore, the use of inductionhardened rivets will be investigated for joining ultra-high strength materials.

Figure 1 .
Figure1.Tooling for the self-piercing riveting process (left) and process sequence for self-piercing riveting and exemplary joining force-displacement curve (right)[3]

2. 2 .
Experimental set-up Multi-range capable rivets are used for the investigations of local heat treatment.The geometry of these rivets is shown in Figure3, a).In addition, Figure3, b) shows the experimental set-up for the heat treatment of the rivets.For the experiments, the rivets are positioned below the induction coil using a rivet holder.Both copper (Figure3, c) and chamotte (Figure3, d) were used as rivet holder material to influence the heat flow into the rivet head.The distance between the rivet and the coil is adjusted by means of a spacer.In order to be able to analyse the influence of the coil position, different series of tests were carried out with varying coil positions.Both a positioning in the area of the rivet foot (top) and in the area of the rivet shank (middle) were selected (see Figure3 e).

Figure 3 .
Figure 3. Illustration of the rivet geometry used (a), experimental set-up for local heat treatment by means of induction (b), and rivet holder made of copper (c) and chamotte (d) and position of the coil investigated (e)

Figure 4 .
Figure 4. Influence of the induction parameters induction coil position and temperature on the resulting rivet hardness when using a copper rivet holder compared to a conventionally hardened rivet

Figure 5 .
Figure 5. Joint formation using a local heat treated rivet; middle coil position, copper as rivet holder, 900°C, 1.5 s

Figure 6 .
Figure 6.Resulting microstructure using inductive local heat treatment at 860°C for 3.0 seconds and a middle coil position with chamotte as rivet holder