Wobble Welding: What Oscillating The Beam Actually Buys You
2026/11/02
Wobble Welding: What Oscillating the Beam Actually Buys You
A laser seam with a fine ripple across it is often assumed to be a hand-fed bead, or a pulsed schedule. In a robot cell it is usually neither. The head is being oscillated: the beam sweeps across the joint while the cell advances along it, and the ripple is the trace of that sweep.
Wobble, or beam oscillation, is one of the few laser settings that changes the process rather than just the number. This article sets out what it does, what it costs, and the jobs where it is the wrong tool.

A robot laser cell at a customer plant: the welding head on the arm, the aluminium tray clamped to the table, and the laser source cabinet behind.
What the Head Is Actually Doing
The oscillation can come from a scanner mirror in the head or from the robot itself, and the two are not equivalent. A scanner can oscillate at hundreds of hertz with a stable amplitude and a choice of pattern: line, circle, infinity, figure of eight. A robot moving the whole wrist is limited by mass and by the accelerations the arm can hold, so the amplitude is larger and the frequency much lower.
The important consequence is that the keyhole is no longer stationary relative to the metal. It is dragged across the joint at a controlled amplitude, so the melt pool is wider than the beam and stays open longer behind the keyhole instead of collapsing immediately. Almost everything wobble is good for follows from that one change.
What It Buys
The first gain is fit-up tolerance. A static beam needs the two edges within a fraction of the spot size, which on a 0.2 mm focus is a demanding thing to ask of a folded and stamped part. A wider pool bridges gaps that a static beam would drop through, and on real parts with real tolerance that is often the difference between a production process and a prototype.
The second is porosity. Stirring the pool gives dissolved gas and zinc vapour a path out before the metal freezes. On aluminium and on galvanised steel this shows up immediately in radiographs, and it is the main reason wobble is common on battery housings and coated enclosures.
The third is bead shape. A wider, flatter bead with less undercut means less grinding afterwards, and on a visible joint the reduction in finishing time frequently pays for the oscillation optics on its own.
The fourth is cracking. A wider pool cools more slowly, and in some steels that is enough to move a centreline crack out of the process window. It is not a universal fix, but it is a lever worth trying before a preheat is added.
What It Costs
Penetration is the first bill. The same power spread over a wider area gives a shallower keyhole, so a three millimetre job that welded comfortably with a static beam may not weld at all when the amplitude is opened up. Wobble is not a way to get more penetration; it is usually a way to trade some of it for tolerance.
Speed is the second. The cell has to advance slowly enough for the oscillation to do its work at each point along the seam, so the gain in tolerance is paid for in millimetres per minute. On a long seam that arithmetic has to be done before the cycle time is promised.
The third cost is cosmetic. The ripple is visible, and a customer who expected a smooth bead will notice it. That conversation is much easier to have before the first sample than after the first shipment.
A last limitation is geometric. The sweep happens inside the scanner field, so working distance and head angle both have to stay within that field. On a deep channel or an inside corner the head cannot get close enough to keep the whole sweep on the joint, and the amplitude that worked on a flat panel has to be reduced for the joint to be reachable at all.

The finished seam along the folded joint of an aluminium battery-box tray. The ripple across the bead is the trace the process leaves as the cell advances.
The Numbers We Actually Set
Five settings decide what wobble does on a given joint. They interact, so we change them one at a time and record the result rather than dialling in a recipe from a table.
Setting |
What it changes |
The mistake we see |
Amplitude |
Width of the stirred pool and gap tolerance |
Opened up to close a gap that is really a fixture problem |
Frequency |
How many sweeps per millimetre of seam |
Too low leaves a zig-zag rather than a blend |
Pattern |
Shape of the sweep: line, circle, infinity |
Circle used on a narrow flange, welding past the edge |
Focus position |
Spot size on the work, so energy density |
Left at the static-beam value after the amplitude changed |
Travel speed |
Heat input per millimetre |
Raised to recover cycle time, which undoes the tolerance gain |
The focus row is the one that catches people. Amplitude and focus are not independent: the same nominal setting puts a different energy density on the work once the beam is sweeping, and a schedule copied from a static job will run hot at the edges of the sweep and cold in the middle.
There is one more practical point that does not appear on any datasheet. Wobble makes the process more forgiving of fit-up and less forgiving of programming: the sweep has a defined amplitude, and on a narrow flange that amplitude can carry the beam past the edge of the part. We check the swept envelope against the flange width on the drawing before the first program is written, not after the first burnt edge.
What Changes on the Machine Side
Oscillation is not free at the head either. A scanner head carries a mirror pair, a focusing lens and a protective window, and the window is the part that fails first. Spatter that would have sailed past a static nozzle lands on it as the beam sweeps, and a contaminated window absorbs energy: a schedule set on a clean window runs cold after a shift, the operator compensates by turning the power up, and a bad loop has started.
Our rule on wobble jobs is a window check at the start of every shift and a logged replacement interval, rather than a look at it when something seems wrong. Cross-jet air aimed across the window is not optional on these heads; in our experience it is the difference between a week of window life and a month.
Where We Do Not Use It
Deep single-pass penetration is the clearest case. If the joint needs the full keyhole depth of the machine, spreading the beam is working against the requirement. Narrow heat-affected zones behave the same way, and so does any joint where the specification asks for a convex, narrow bead.
Thin foil is a subtler case. The pool is small enough that the sweep can move the metal rather than stir it, and the result is a wavy seam with variable penetration. Below a certain thickness, a static beam with a good fixture is the more reliable answer, and we say so rather than quoting the more impressive option.
Requirements vary by destination market. Always confirm against the current edition of the applicable standard and with your certification body.
PDKJ continues to provide customized laser welding and resistance welding solutions based on different materials, workpiece structures, welding requirements, and production applications.
Related Products

Part of our assembly floor in Dongguan. Machines are built, wired and run off here before they are packed, and the same bench is used to try a new schedule on a customer sample.
Contact PDKJ
Send us your part drawing or a sample, and our engineers will run a welding feasibility trial on your own material and recommend the equipment configured for your application.
WhatsApp: +86 136 1279 3248
Email: jay@gd-pw.com
Website: https://www.spotlaserwelding.com/contactus.html
Address: 1-2F, Building 3, Qichen Industrial Park, No. 26 Luxi 1st Road, Liaobu Town, Dongguan City, Guangdong Province, China