Review of recent developments in resistance welding (Nov 02)

Paper presented at 2nd International Seminar on Advances in Resistance Welding, 6-7 November 2002, Aachen, Germany

Introduction

Resistance welding processes have a century of history and have developed from very humble beginnings into one of the most sophisticated automated welding processes of the present day. Despite the invention of a number of more exotic processes in recent decades, such as laser welding which has replaced some traditional resistance welding applications, the unique characteristics of resistance welding have kept the process in the forefront of the manufacturing industry.

Probably the two key factors in joining technology are cost and quality. In manufacturing, maintaining the quality of the joint to a level that is fit for purpose is vital. Increasing demand is being placed on reliability. The reason is clear in the case of a safety critical part. However, the cost implication of failure either at a later stage of manufacture, or in service can be significant both in terms of the remedial action needed and also the loss of customer confidence. Although no welding process can be guaranteed 100%, the pressure is put increasingly on the supplier to provide low parts-per-million risk of failure. In some cases, an isolated weld failure is unlikely to compromise the performance of the product, such as in vehicle body assembly. However, when making electrical interconnections, for example, one failure is unacceptable.

Manufacturing cost is the other key factor in today's highly competitive marketplace. The initial capital cost and the running cost must be balanced over the life of a product. As one of the major contributors to running cost is labour, particularly in Western Europe, automation plays a very important role.

This paper provides a view of developments in spot welding in a number of areas which impact on these cost and quality factors. Topics covered include equipment and power supplies, control and monitoring, electrodes and tooling, plus material challenges and innovations.

Equipment

The most significant developments in welding equipment in recent years have been related to power supplies and the force application systems. Medium frequency power supplies now dominate new automotive installations and feature widely in miniature applications ^[1,2] . They have seen gradual improvements in efficiency and control capability over a period of about 15 years. Transistorised dc supplies have also been widely applied in miniature applications and the feedback capabilities provide exceptional control of the weld pulse. The advantages to be gained of reduced transformer weight with medium frequency, in the range of 1000Hz, accounts for their widespread use in robotic equipment. However, this has been further exploited by operating at even higher frequencies, such as 20kHz systems. This enables the transformers to be even more compact, with particular benefit on long reach guns where additional weight savings can allow reduced robot size to be used and also reduce gun inertia.

Servo controlled force application systems ( Fig.1) are gaining increased interest as they provide much greater flexibility in terms of programmable force levels ^[3,4] . In addition, the ability to control electrode position and approach speeds provides potential time savings in the welding sequence. The time required from initial electrode contact to reaching the set force is also claimed to be reduced compared to pneumatic systems. Reducing the operating stroke and electrode impact speed can also have positive benefits in terms of electrode wear. Pneumatic systems are frequently set with high approach speeds to minimise cycle time but the consequent high impact forces can increase the mechanical wear of electrodes.

Fig.1. Typical servo controlled spot welding gun (Courtesy ARO)

Integrating the servo gun control with the welding robot can enable the robot to be moving while the gun is opening or closing, thus minimising the manipulation time between spots. However, when welding long flanges, there is no need to open the gun if roll spot welding is used, and this can save further time. Robot mounted seam/roll spot welders ( Fig.2) are available capable of around 120 spots per minute. 

Fig.2. Robot mountable resistance seam welding gun (Courtesy Nimak)

Automotive assembly line builders are also endeavouring to produce more positive tooling to locate components prior to welding. This has allowed greater accuracy and reproducibility of the dimensions of the welded structure. This has the added benefit of improving the fit up of components for welding, provided that this is coupled with good quality press work. The integration of laser measuring devices at key stages of the build allows further feedback of potential tooling problems if deviations are found in these dimensions. These improvements in components should allow better spot weld location, such as avoidance of welds made too close to a sheet edge, and consequently more reliable weld quality. However, this can only be achieved if the position and alignment of the electrodes relative to the component is achieved by initially programming the robot well.

Maintenance of spot welding guns is also an important aspect of reliable production. In some plants, the ideal situation exists where duplicate guns stand ready to replace any gun on the line and then the fault rectified or programmed maintenance can be carried out without interrupting production. As a means of re-testing a gun efficiently, it can be coupled to a dedicated testing station, equivalent to its robot connection. The test station can then check electrode force, water flow and electrical functions ^[5] . Additionally, the dimensional accuracy of the gun assembly can be checked with a manually operated high accuracy measuring arm. This ensures positional accuracy of the weld and correct clearances once the gun is reinstated. A means of storing welding gun specific data and welding programs within the gun itself is available ^[6] , in which off-line programmed data and maintenance records can be stored and downloaded to the welding controller ( Fig.3).

Fig.3. Welding gun data storage system MASDAT (Courtesy Matuschek)

Substantial refinements have been made in recent years in equipment for small scale welds for joining wires, foils and a range of more 'difficult' materials. Power supplies such as the higher frequency inverter type up to about 25kHz and transistorised supplies can provide fine control of short duration current pulses with feedback control in various forms: current voltage or power. Welding head characteristics can also be quite critical for such cases. Very low inertia, fast follow-up heads are available with spring, pneumatic or, more recently, electrical force application. In this latter case, variable force has been used within the welding pulse to provide active force control.

Control and monitoring

In order, to permit continuous production of automated systems, much effort has been put into the computer control systems such that faults may be immediately flagged up and rectified. This could include the sensing of failures such as tooling or gun problems, displaying the location of the fault, through the control computer, and paging the appropriate maintenance staff for immediate action. Other capabilities of the control systems allow remote interrogation of specific welding guns to check or adjust settings (with the appropriate password authority) in response to feedback from quality checks. The status of guns can be supervised as necessary, in terms of the number of welds made relative to electrode maintenance intervals or current stepping sequences.

In many cases, current measurement is built into the welding equipment and used for the process control, linking to features such as constant current or stepping functions. A range of commercial meters is available to provide the means of verification of current values on a routine basis. Force is a factor less likely to be continuously monitored but again a range of tools is available. Newer force measuring devices provide force time information enabling squeeze time to be optimised. In addition, guns have been produced which have built in force monitoring ^[7] .

Displacement monitoring has been shown to provide the most reliable indication of spot weld quality in the past, through weld expansion measurements. Weld quality monitors have been produced on this basis. However, this approach is less applicable to flexible equipment such as welding guns in production. Displacement measurement is particularly beneficial in some miniature applications where the measurement of set down in projection welding applications, such as fine wire welding, can confirm process control and weld quality.

Electrical measurements of current and voltage provide the least risk to sensor damage and have been extensively studied. Simple measurements of dynamic resistance or power have some limitations in their ability to accurately predict weld quality under the wide range of production variables encountered in the process. However, newer systems and development studies are based on more complex algorithms or the application of fuzzy logic to the recorded data ^[8,9] .

While process control forms the basis of good weld quality control, the use of adaptive control has been implemented in a number of cases. In-weld adjustment of the welding current, and if necessary weld time has been shown to compensate for variation in material, component and electrode variables, as well as the welding variables themselves. However, there is little openly published data on production case history experience with such equipment.

Although in-process weld quality control is still not universally proved, it is certainly the ideal approach. In the meantime, testing of spot welds in sheet steel is moving increasingly towards ultrasonic inspection. The PC based units now available ( Fig.4) often provide some automatic evaluation of weld quality but there is still a strong reliance on the skill of the operator to achieve a suitable signal and provide interpretation. However, modern systems do provide a comprehensive data logging capability and the weld test data can be linked directly to the part drawings ^[10] .

Fig.4. PC based ultrasonic testing equipment (Courtesy AGFA NDT)

A new approach ^[11] , to incorporate the ultrasonic sensor into the welding equipment is also being tried. While this has been attempted previously with a sensor bonded to the inside of the electrode cap, the new approach has the sensor mounted in the elbow of the electrode arm and the signal transmitted through the cooling water. A disadvantage of this approach, however, could be the difficulty of keeping the sensor square to the surface of the electrode contact. The angular alignment tolerance is very small; otherwise the return signal from the weld would not be detected fully by the sensor and would appear severely attenuated. Other techniques, such as global resonance testing ^[12] are considered to show promise and potential cost savings.

A number of systems have been devised and are being continually improved for modelling resistance weld formation ^[13,14,15] ( Fig.5). This enables potential optimisation of weld designs and reduces the number of practical trials required to set up a particular application. The predictive modelling has been taken further to incorporate it into monitoring devices and provide weld quality monitoring or feedback control of the welding pulse as a means of avoiding weld splash.

Fig.5. Example of a spot weld model showing temperature distribution (Courtesy Swantec)

Electrodes and tooling

There has been long discussion on the merits of the electrode materials and tip shapes available, and there are different preferences between users. The choice depends on cost and overall electrode life under the conditions used. While the copper zirconium and alumina dispersion strengthened copper types are frequently used for coated steels and to cope with hot conditions, electrode life can be controlled by the dressing procedures used.

Dressing can be conducted frequently with little material removal, as a means of maintaining the tips in perfect condition, or by allowing substantial wear within the life of the tip then dressing substantially back to the original shape. In the latter case, the procedure may be accompanied by a current stepping operation. As a means of minimising the time required for electrode changes, automatic tip changers are now available. In addition to the dressing stations providing the cutting tool or tools to redress the electrode, additional features may be built in such as tip size and alignment checking and electrode force and current measuring devices. Such accessories offer the possibility of providing routine quality checks.

Many different applications in resistance spot welding incorporate designs, which are not ideal in terms of spot welding assembly, because they limit access and require the use of non-preferred electrode configurations. This might include forward angled electrodes with small diameter adaptors that increase the flexibility of the electrode assembly under load leading to electrode skidding or misalignment.

Materials challenges and innovations

In the competition for lightweight materials, particularly for automotive applications, ultra-high strength (UHS) steels and aluminium alloys are receiving much attention. In general, steels with a tensile strength up to 600N/mm² are well established and do not cause problems on resistance spot welding. However, there has been limited spot welded application of steels over 600N/mm² and there is a still a need for weldability and performance data for these steels.

Plug failure is a standard weld quality criterion and allows the weld size to be verified after routine testing. This avoids the difficulty of interpreting the size and strength of the weld from an interface or part interface fracture. The occurrence of interface fractures can indicate excessive hardening and potential embrittlement of such welds. However, the fracture behaviour is dependent on sheet thickness, material type and weld size. Work at TWI on steels up to 1200MPa tensile has shown that good weldability with plug failure can be achieved in some grades in thin sheet, despite relatively high hardness 400 to 450HV. Other types, such as TRIP steels with a higher carbon content, and that harden excessively (around 600HV) remain a problem. Special welding in-process heat treatment procedures are then needed which may not be compatible with high rate production. Long weld times and pulsed conditions, controlled cooling, in-process tempering, or weld dilution methods are the means of modifying weld hardness and fracture behaviour in UHS steels.

In the highest strength steels, softening of the heat affected zone can occur and this can limit weld strength. In other cases, an interface fracture is not always associated with weld brittleness and high cross tension strength can be achieved. The potential range of material and thickness combinations likely to be found in automotive applications complicates the picture.

Aluminium alloys have more recently been self-piercing riveted instead of being spot welded in a number of major applications. However, spot welding provides a lower cost alternative, with matching properties if coupled with structural adhesives. The key factors in making reliable spot welds in aluminium are good control of the material surface and the electrode tip. Successful production applications are achieved where sufficient care is taken over these factors. A novel approach to tip dressing by Alcan, reinforces the benefit of such control. Frequent buffing of the tips was shown to completely prevent progressive tip contamination and reproducible weld quality was achieved over tip lives of several thousand welds.

The other material types gaining interest are the organic coated, pre-primed steels designed to eliminate the initial stages of the painting process, with consequent cost and environmental savings ^[16] . The very thin organic coatings around 1 to 2µm are readily weldable but some of the newer much thicker coatings make spot welding more difficult. In addition, the underlying chromium free pre-treatments add to the electrical resistance of the coating. The coating suppliers are working to refine the coatings and conductive particle filling to make spot welding more reliable. This is again a case where good electrode dressing practice, programmed pulse shapes, including upslope, and advanced controller types may contribute to the successful welding of such materials.

There is still scope for novel approaches to resistance welding and recent developments such as conductive seam welding of aluminium alloys ^[17] ( Fig.6) and innovative resistance brazing methods ^[18] demonstrate this. In the former case, a butt seam weld in aluminium can be achieved by using consumable steel foils each side of the joint, which help generate the heat and contain the fused seam. The resistance brazing approach allows brazed spots to be made with similar static and higher fatigue strength than an equivalent spot weld.

 Fig.6. Conductive heat resistance seam welding of 2mm thick 7075-T6 aluminium alloy sheet

Summary

The computing power available in integrated systems allows an increasing amount of data to be handled regarding the welds made and enables monitoring and test data to be linked. Statistical handling of the information allows trends to be identified and action taken as appropriate. The refinement of modelling and monitoring algorithms offers the potential for increasingly reliable weld quality monitoring and adaptive control. In support of this, verification of quality by non-destructive testing is increasingly being implemented but more reliable automatic interpretation is needed to reduce the demand on skilled operators.

The resistance welding processes provide relatively low cost, high productivity solutions to sheet metal joining and a range of other application areas. The tools are available to set up procedures and control the process to meet stringent customer demands and also to meet the European standards on weld quality and procedures, currently being finalised. It will be increasingly important to ensure that job knowledge and training are provided to enable the processes to be operated effectively.

The prospects for resistance welding processes seem secure in the foreseeable future but will be influenced to some extent by materials choices in the automotive industry. However, joining technology and materials development must go hand in hand, as cost effective processes are as important in manufacturing industry as the design and materials used.

References

1. Abbott G: 'The evolution of welding to support new body structures'. Automotive Manufacturing Solutions June 2001, pp.60-67.
2. Mewborne M: 'Controlling weld energy'. Practical Welding Today May/June 1999, pp.21-25.
3. Slavik S: 'Using servoguns for automated resistance welding'. Welding Journal July 1999, pp.29-33.
4. Schmidt-Dörnte J: 'Der Servoschweisskopf - Dynamik in Schweissprozess, Regelbarkeit des Servoantriebs'. DVS Berichte 189 Duisburg, May 1998.
5. Drechsel T: 'Instandhaltung von Widerstandsschweisszangen'. Der Praktiker November 2001, p.438.
6. Pöll K and Matuschek U: 'Vorteile der Datenspeicherung auf Schweisszangen - das intelligente Schweisszangen-Identificationssystem MASDAT'. DVS Berichte 213, Duisburg May 2001, pp.143-146.
7. 'Electrode force as an important process variable in resistance welding'. Schweissen u. Schneiden March 2001, pp.130-131.
8. Pöll K and Matuschek U: 'Adaptive control for resistance welding in the automotive industry'. Sheet Metal Welding Conference IX, Sterling Heights, Mi. October 2000.
9. Straube A and Winzen B: 'Fuzzy basierende Überwachungssysteme für Punkt- und Buckelschweissanwendungen, Möglichkeiten und praktische Anwendungen' DVS Berichte 213, Duisburg May 2001, pp.98-103.
10. Roye W: 'Ultrasonic testing of spot welds in the automotive industry'. Eurojoin 4, 4 ^th International Conference on Welding Joining and Cutting, Zagreb May 2001, pp.159-166.
11. 'Checking weld spots during the welding process'. Schweissen u. Schneiden March 2002, p.118.
12. Hatmaker T and Das S: 'Economic viability of non-destructive testing for testing joints'. Materials Evaluation October 2001, pp.1208-1213.
13. Zhang W and Kristensen L: 'Finite element modelling of resistance spot and projection welding processes'. NIST Special Publication 949. Proceedings, Ninth International Conference on Computer Technologyin Welding, Detroit 28-30 Sept.1999 pp.15-23.
14. Dilthey U et al: 'MAGSIM and SPOTSIM - Simulation of GMA and spot welding for training and application'. Proceedings of the 7 ^th International Conference of the Japan Welding Society, Kobe November 2001, pp.143-149.
15. Matsuyama K: 'Real time visualisation of welding state in spot welding'. NIST Special Publication 949. Proceedings, Ninth International Conference on Computer Technology in Welding, Detroit 28-30 September 1999, pp.3-14.
16. Müller M: 'Organisch vorbeschichtete Stahlbleche - Übersicht, Betriebs-erfahrungen, Entwicklungstendenzen'. DVS Berichte 213, Duisburg May 2001, pp.39-45.
17. Gould J et al: 'Application of conductive heat resistance seam welding'. Sheet Metal Welding Conference IX, Sterling Heights, Mi. October 2000.
18. Sitte G and Koppe K: 'Punktförmiges Widerstandsloten - oberfläschenschönende Prozesse für Überlappverbindungen'. DVS Berichte 213, Duisburg May 2001, pp.81-86.