Edward Watts is a Senior Welding Engineer in the Friction & Forge Processes Group. He has worked at TWI for 16 years, mainly involved with equipment development.
Dave Nicholas is Manager of the Friction & Forge Processes Group. He has been involved with research and development of friction based processes at TWI for 28 years.
The passage of time has seen major advances in hydraulic, transmission and control systems for friction welding. The opportunity to use more complex combinations of major welding variables is now available, as described by Edward Watts and Dave Nicholas.
At the beginning of the 1960s TWI revealed its commitment to the solid phase process of friction welding. [1] Since that time, a wide variety of friction based processes and the necessary equipment have evolved, so much so that there are around 20 machines available for development and demonstration in the friction welding laboratory at Abington. Rotary friction welders cover a wide size range from 3-150mm diameter in solid bar, whilst machines capable of welding non-round components using orbital, linear [2] and angular reciprocating motions further extended applicability of the technology. In recent years friction stir welding, [3] friction hydro pillar processing [4] and friction taper stitch welding [5] have realised even greater opportunities to exploit the advantages of this solid phase joining technology.
To the present day, machine manufacturers continue to improve and increase in capacity the conventional rotary continuous drive and stored energy (inertia) machines. Quite clearly machine design is initially dictated by the diameter and cross-sectional area of the components to be joined. It is interesting to note that significant divergence of equipment size and capacity is brought about by the components' dimensions. For example, small diameters generally require much higher rotary welding speeds in order to provide a surface velocity of approximately ≅2 m/sec.
However, because the weld areas are small, the transmission power of the rotating system and the necessary axial welding forces are low. In direct contrast, as diameters and cross-sectional areas increase, rotation speed is reduced but there is an associated greater demand for transmission power and axial welding force. Consequently, machines become significantly larger to provide the necessary welding parameters whilst maintaining rigidity and dimensional tolerances of welded components. At present, the largest friction welder of the stored energy (inertia) variant is located in the USA, and is capable of delivering 20 000kN axial force.
Continuous drive rotary process
Almost all commercially available continuous drive machines run at a constant rotational speed, and adopt relatively straightforward welding force cycles, such as a first and second stage friction force followed by an upset or forge force, where their values are preset. Figure 1 illustrates a typical weld cycle. The interaction of rotary speed and friction welding forces for a given material and size will significantly affect the various torque levels (initial and equilibrium) and metal displacement rates seen during the friction heating sequence of the weld. [6] The chosen values for these major welding variables are invariably decided upon so that they result in the production of welds which exhibit acceptable metallurgical and mechanical properties. Also the values generally lie at the centre of a reasonably wide tolerance window.
It is understandable why the speed and force cycles were kept simple, since in the early days variable speed and force systems were either expensive or the control technology and hardware, certainly for hydraulics, had not been developed. However, more sophisticated control systems, such as PLCs and more power microcircuits, have been developed. Also, alongside these systems the hardware necessary to react to such control inputs has become more reliable and advanced.
A scan through the literature reveals that the Russians were investigating the idea of pulsing energy into the weld, [7] whilst in Germany researchers were investigating the use of various speeds and force levels, which were controlled by feedback from an infrared sensor looking at the progressing weld. [8] In more recent times Blakemore has produced a control system which permits many levels of speed, force and displacement, ensuring that welding is performed to all of these preset levels. [9]
It was with this background in mind that the friction welding machine, which is now available for ongoing process and applications development, was constructed. The primary aim for a machine of this type was to investigate the speed/force pulsed combinations and interactions with the overall objectives of reducing welding forces, transmission powers and residual stresses, which will reduce machine size and thus cut costs.
Equipment
Some years ago TWI took delivery of a special Thompson friction welding machine that was commissioned by the then CEGB's Marchwood Research Laboratory. The machine was initially purchased for research and development, so consequently was fitted with a hydraulic transmission system to provide infinite speed variation up to 1500 rev/min and at that time a state-of-the-art hydraulic force system capable of delivering a maximum welding force of 400kN. After installation a Lynx infinitely variable speed electrical transmission system capable of delivering 50kW, a special hydraulic system, and control system which uses a Mitsubishi Melsec F1 series PLC, were fitted to permit the machine ( Fig.2) to operate at the following nominal specification (see Table).
| Table: |
|---|
| Rotation speed | High range 0-1500 rev/min, low range
0-750 rev/min, three levels preset for friction cycle |
| Welding force | 10-500kN infinitely variable |
| Welding stroke | 200mm |
| Welding force oscillation | ±25% about three preset friction force levels at frequencies from 0-100Hz |
| Head advancement rate | 0-25mm/sec force maximum with both servos fitted |
| Rotary component size | 10-200mm diameter using manual chucking |
| Stationary component size | 10-160mm diameter using hydraulic chucking |
The preset levels of rotational speed and the preset force levels (three friction and one upset), can be controlled by either displacement or time. An example of possible combinations of the main welding variables is illustrated in Figure 3.
A three stage force level with approach force prior to contact, and variable flow control during each of the stages is also possible. For each force level, oscillation of the force (even separation of the weld components if so required) is possible, where the frequency, offset gain and mark space ratio are all independently adjustable. During each stage of force, the rotation speed can be infinitely varied from zero to the maximum speed of the range set.
The machine is fully instrumented to obtain dynamic outputs of axial displacement, rotation speed, welding force and torque. These outputs are fed to a PC-based data acquisition system for storage and further processing.
The future
Already, project work which involved development trials with one of TWI's emerging processes, that of friction hydro pillar welding, has benefited from the greater flexibility afforded to the permutation of welding variables, improving significantly the joint properties.
The machine can be used for a range of friction welding processes, which includes conventional rotary friction welding, friction surfacing, friction stir welding, friction plunge welding, third body friction welding and friction hydro pillar processing of metals.
Future trials are planned to assess the various combination patterns of the variables on weld operating characteristics and the quality of welds produced. Also under consideration are further modifications to the machine's control system to permit even more complex variable combinations.
References
| N° | Author | Title |
|
| 1 | Anon: | 'Exploiting friction welding in production.' Information package series, published by The Welding Institute, 1979. | Return to text |
| 2 | Nicholas E D: | 'An introduction to linear friction welding.' 1st European Conference on joining Technology, November 1991, Strasbourg, 424-432. | Return to text |
| 3 | Dawes C J and Thomas W M: | 'Friction stir process welds in aluminium alloys.' Welding Journal 1996 March, 41-45. | Return to text |
| 4 | Nicholas E D: | 'Friction hydro pillar processing.' 11th Annual North American Research Conference, Advances in Welding Technology, 7-9 November 1995. Columbus, Ohio, USA, 313-322. | Return to text |
| 5 | Andrews R E and Mitchell J S: | 'Underwater repair by friction stitch welding.' Metals and Materials, Marine Technology 1990 6 (12). | Return to text |
| 6 | Ellis C R E: | 'Continuous drive friction welding of mild steel.' Welding Journal 1972 51 (4) 183-197. | Return to text |
| 7 | Voinov V P et al: | 'Pulsed friction welding of alloy ZLS6-K and steel 40G.' Welding Production 1976 23 39-41. | Return to text |
| 8 | Drews P and Schmidt J: | 'Feedback control in friction welding - operation of computer and microprocessor units to enlarge the field of application.' IIW Doc IIIJ-33-80, May 1980. | Return to text |
| 9 | Blakemore C R: | 'Design and implementation of a total control system for a portable friction welding machine.' 4th International Conference. Computers in Welding, TWI, Cambridge, UK, June 1992, Paper 6. | Return to text |
TWI Industrial Members are reminded that the machine is available for development trials that may be specifically orientated to their application needs for the continuous drive process, and be used to study variants on other friction processes - as listed previously.
