Laser Assisted Friction Stimulation Welding: How Preheating Changes the Weld

• 1
  What the Laser Actually Does Before the Pin Arrives
• 2
  Parameters That Decide Whether the Process Works
• 3
  Which Alloys Gain the Most From Preheating
• 4
  Limits, Tool Wear, and Where the Process Stops Paying
• 5
  How to Prove the Joint Before You Commit to a Run
• 6
  Laser Assisted Friction Stimulation Welding vs Plain Friction Stir Welding

In standard friction stir welding, the rotating tool does all the work. Shoulder friction and pin stirring generate the heat that softens the metal, and that heat has to be generated in the same place the tool is pushing. The problem is that the tool has to get there first. Axial force stays high, and the machine frame and backing bar carry the load.

Laser assisted friction stimulation welding splits the job in two. A fiber laser tracks a few millimetres ahead of the pin and deposits energy into the surface. The metal arrives at the tool already warm, so the tool needs less torque and less downward force to reach the same stirring temperature.

The practical result is a narrower heat-affected zone around the stir zone. Because the laser heats only the strip that the tool is about to work, the rest of the plate stays cool. Distortion drops, and the clamp fixtures can be lighter.

Preheating also changes the flow stress of the alloy at the tool interface. Softer metal flows around the pin more easily, which reduces the risk of void formation behind the pin at higher travel speeds. That is where the throughput gain comes from.

• 1
  Laser leads the pinBeam spot sits 5–20 mm ahead of the shoulder edge.
• 2
  Heat is localOnly the weld strip is heated, so the plate stays flat.
• 3
  Force dropsLower torque and axial load on the same joint.

Laser power per millimetre of thickness is the first number to set. For 6xxx aluminium, a common starting band is 0.8–1.5 kW for 6 mm plate, which raises the surface to roughly 200–300 °C ahead of the pin. Too little power and the laser does nothing; too much and you form a surface melt pool that the pin then drags into the stir zone.

Lead distance matters as much as power. If the beam sits too close to the shoulder, the heated zone is undercut by the tool before it can conduct downward. If it sits too far ahead, the plate cools again before the pin arrives. A 5–20 mm lead over a 1–3 s conduction window is the usual working range.

Tool rotation and travel speed trade against laser power. Raising travel speed without raising laser power starves the joint and produces wormhole defects. Raising laser power without raising travel speed overheats the surface and burns the oxide layer. Change one variable at a time and log the result.

Backing bar material sets the ceiling on all of this. A steel backing bar pulls heat out of the root, so the root lags the top surface. On thick sections that gap is what limits penetration, not the laser.

• 1
  Power band0.8–1.5 kW per 6 mm of 6xxx aluminium.
• 2
  Lead distance5–20 mm ahead of the shoulder edge.
• 3
  Preheat targetAbout 200–300 °C at the surface.

High-strength 7xxx and 2xxx alloys gain the most. They are the ones that resist plastic flow at room temperature, so the tool has to push hard. Preheating drops the flow stress enough that a smaller machine can weld the same section, and the stir zone stays closer to the parent metal in hardness.

Magnesium alloys gain for a different reason. AZ31B and AZ91D have low melting points and a narrow hot-working window. A laser lets you reach the bottom of that window from the surface instead of relying on friction heat that can overshoot into the shoulder.

Dissimilar joints, such as aluminium to copper or aluminium to steel, also benefit. The laser can bias heat toward the harder, higher-melting side, which balances the two flow stresses at the interface. That is usually the difference between a continuous joint and a brittle intermetallic layer.

Casting alloys with high silicon content, like ADC12, are a weaker candidate. Silicon particles do not soften much with preheat, and the tool wear stays high regardless of laser power.

• 1
  Best fit7075, 2024, AZ31B, AZ91D, aluminium-to-copper.
• 2
  Weaker fitHigh-silicon die casting alloys such as ADC12.
• 3
  ReasonPreheat lowers flow stress, not particle hardness.

The laser adds a second heat source that has to be aligned every time the fixture changes. On a 2 m seam, a 1 mm drift in beam position shifts the preheat profile and shows up as a hardness band in the stir zone. Beam delivery optics need cleaning on a schedule, not on a hunch.

Tool wear does not disappear. The pin still contacts hot, abrasive metal, and in 7xxx alloys the tool life is measured in metres of weld rather than shifts. Preheating lowers the mechanical load, but the thermal load on the pin coating can rise.

Energy cost is real but small next to the machine load. A 1.5 kW laser adds little against a spindle that draws far more. The savings come from lower fixture stiffness, fewer re-clamps, and less post-weld straightening.

For thin sheet below about 3 mm, the laser is usually unnecessary. The tool reaches welding temperature on its own within a few millimetres of travel, and the extra heat source only adds setup complexity and a distortion risk you did not have before.

• 1
  AlignmentBeam drift of 1 mm changes the hardness profile.
• 2
  Tool lifeStill limited in 7xxx; measure in metres of weld.
• 3
  Skip itBelow about 3 mm sheet, plain FSW is enough.

Cut a transverse section from the first article and look at the stir zone shape. A properly preheated weld shows a symmetric, fully consolidated nugget with no tunnel defect at the retreating side. If the nugget is asymmetric, the laser is biased to one side of the seam.

Run a hardness traverse across the joint at 0.5 mm steps. In 6xxx alloys, the minimum hardness in the heat-affected zone should sit within about 10–15 HV of the parent metal when laser preheat is used, because the total heat input per unit length is lower.

Pull three tensile coupons per parameter set. Record ultimate strength, yield strength, and elongation, then compare against the parent metal. A joint that reaches 70–80% of parent yield with good elongation is behaving normally for a stirred aluminium weld.

Bend tests catch what tensile tests miss. A root bend over a 4 mm mandrel will open a lack-of-penetration defect that a tensile bar can pass straight through, especially on thick sections with a cold root.

• 1
  Macro sectionCheck nugget symmetry and tunnelling.
• 2
  Hardness traverse0.5 mm steps across the joint.
• 3
  Bend testRoot bend finds cold-root defects.