Deflection mapping can be used to program and operate computer numerically controlled friction stir welding gantry machines (CNC FSW gantries) if they deform elastically by the high forces applied during FSW. This is an effective method, if you experience a lack of penetration (kissing bonds) in the centre of the weld, or if you expel too much flash near the start and stop of the weld. Due to varying engineering standards and design principles, it is most commonly used for low-cost machines, while high-quality machines tend to be suffiently rigid.
As a first step, you should measure the deflection using a scale or a calibrated force measurement box without rotating the tool, and check which z value should be set, to achieve the desired force. Due to the elastic deformation of the FSW machine, it is likely that the z value to be set is higher in the centre of the work table than that near the edges, to achieve the same load of for instance 25 kN (2,5 t). High quality welds can be achieved even in position controlled mode, if the path of the FSW tool is programmed keeping the expected deflection in mind.
For deflection mapping of FSW gantries the elastic deformation of the machine is measuered as a function of x-y position and varying forces and then taken into account during programming of the weld path
Friction stir welding is a solid phase pressure welding process, which requires high forces to forge and consolidate the plasticised material. The highest force is the downward pressure to keep the tool at the right height just below the surface of the workpieces, or to maintain the right ‘tool heel plunge depth’ typically 0.2 mm below the surface of the workpiece in the case of using a tilted tool with a concave shoulder. If the tool is tilted, the tool should be pressed downward so far that 75 % of the shoulder circumference is in touch with the plasticised material, i.e. some plasticised material is trapped underneath the shoulder ahead of the pin, so that the pin is not visible when looking into the gap between the shoulder and the workpiece.
The traversing force is normally only 10% of the downward force. It is required to move the tool forward at the desired welding speed. The third force, a sideward force, is a resulting force caused by the rotation of the tool. If the FSW machine or its bearings aren’t ridged enough, the tool will run a little bit off-set from the programmed path, i.e. it is to far left or to far right from the joint line, depending on the clockwise or counter-clockwise rotation of the tool.
The torque is transferred into the mounting brackets of the spindle and causes some twisting around the spindle axis.
FSW machines should be as rigid as possible, to limit deflection during welding. In some cases this is not possible, e.g. if transportable FSW machines or low-cost FSW machines have to be used, or if a conventional milling machine s being used for FSW. Deflection mapping can be used in these cases, or if the machine has been specified or designed by someone who underestimated the required forces.
Deflection of FSW gantry machines is caused by bending of the gantry’s beam, bending of the table, the slack in the bearings and tilting and twisting of the tool spindle. An elegant solution is to use force control algorithms in combination with seam tracking sensors. However, in some cases, position control is preferred, e.g. if the hardness of the workpiece varies at a crossing with a previously made weld or if the heat conductivity of the workpiece varies near ribs and spars.
The deflection may not be neglected, especially when welding in position control mode. The amount of deflection depends on the design principles and the quality of manufacture of the FSW machine. Typically, the deflection is highest in the centre of the machine due to bending moment in the beam and table. Occasionally, the beam deflects by more than 5 mm, while the tool height should be adjusted within ±0.1 mm tolerances.
If you have a deflection map, you can program the machine according to the expected deflection. In easy cases, it might be sufficient to measure the deflection only at the worst point, where the maximum deflection occurs, and then use a catenary function to interpolate the assumed deflection as a function of the x and y position and the force applied.
As an alternative, or in addition to applying deflection mapping, the following options may be considered:
Please contact firstname.lastname@example.org if you need help on deflection mapping and/or parameter optimisation.