Laser Welding in Transmission Production: How the Joint Actually Forms

• 1
  What Happens in the Melt Pool During laser welding in transmission production
• 2
  Joint Fit-Up Tolerances and Surface Conditions
• 3
  Which Alloys Weld Well and Which Ones Fight Back
• 4
  How We Verify a Production Weld Before It Ships
• 5
  Laser Welding Compared with Other Transmission Joining Methods

A transmission weld is a lap or butt joint between two rotating parts, usually a gear blank and a shaft, or two stamped shells that form a drum. The beam arrives at the interface, the surface absorbs the energy, and metal melts within microseconds. If power density is high enough, the molten surface vaporizes and opens a narrow cavity called a keyhole. That keyhole is what separates laser welding from arc processes: the beam reaches deep into the joint instead of spreading heat across the surface.

Once the keyhole is open, molten metal flows around it and solidifies behind the moving beam. The result is a narrow fusion zone with a heat-affected zone often under 0.5 mm wide on steel. That narrow HAZ is the reason a transmission gear can be welded after heat treatment without losing case hardness across the whole tooth flank. Distortion stays small because total heat input is low, not because the process is gentle.

Two modes matter in practice. Conduction mode stays below the vaporization threshold and produces a shallow, wide weld, useful for thin stamped shells around 0.8–1.5 mm. Keyhole mode penetrates and gives a depth-to-width ratio from roughly 2:1 up to 10:1 on stainless and low-alloy steel. Transmission joints that carry torque usually need keyhole mode, because the joint must fuse through the wall, not just sit on top of it.

The transition between the two modes is set by power density, not by total power. A 3 kW beam focused to Ø0.2 mm behaves differently from the same 3 kW spread over Ø0.6 mm. That is why welding parameters cannot be copied between machines without checking spot size, focal position and shielding gas flow first.

Laser welding forgives very little gap. For a 1.5 mm wall in low-alloy steel, a gap above about 0.1 mm at the interface will produce lack of fusion or a sagging root. Press-fit joints are common in transmission work for exactly this reason: the two parts locate each other and the beam only has to close the remaining clearance.

Surface condition matters as much as geometry. Cutting oil, oxide from an earlier heat-treat step, and phosphate coatings all absorb or reflect energy differently from bare metal. A wiped joint with a clean, dry surface welds consistently. A joint with a heavy oxide layer needs more power or a different shielding gas mix, and it will still show porosity.

For aluminium housings and drums, reflectivity adds a second problem. Aluminium reflects a large share of near-infrared light at room temperature, so the first milliseconds of the pulse are unstable. A slightly rough surface, a small preheat, or a ring-mode beam helps the weld start evenly. Once the keyhole opens, absorption rises sharply and the process stabilizes.

We check fit-up, wall thickness and material grade before quoting a welded assembly. A drawing that specifies a weld symbol without a gap allowance usually needs a small design change, and that is cheaper to find at the quotation stage than after the first parts are cut. The same review applies to joints that will be machined after welding, because stock left on the weld face changes the heat sink.

Low-alloy and carbon steels are the easy case. Grades such as 4130, 4140 and 4340 weld cleanly with a keyhole, and the narrow HAZ keeps hardness loss local. If the part is already through-hardened, a post-weld temper is still needed in most cases, but the zone to treat is small.

Stainless steels behave differently. Austenitic grades like 304 and 316 resist cracking but conduct heat poorly, so the weld stays tight and the HAZ stays narrow. Martensitic grades such as 420 and 440C are harder to weld without preheat and post-weld heat treatment, and 17-4PH needs a defined aging sequence after welding to recover strength.

Aluminium and magnesium are the reflectivity problem again. 6061 and 7075 weld, but 7075 is generally avoided for fusion joints because of hot cracking. ADC12 die-cast housings contain porosity and gas that expand under the beam; a welded joint on a die casting is possible but needs a porosity check on the casting first.

Titanium and nickel alloys belong in a different conversation. TC4 (Ti-6Al-4V) and Inconel weld well under full inert shielding, but oxygen pickup above a few hundred ppm embrittles the joint. That requires a glovebox or a trailing shield, which changes the cost structure. We flag those joints early rather than quoting them as routine work.

A weld that looks good under a microscope can still be unacceptable. The checks that matter are the ones tied to the failure mode of the joint. For a transmission drum, that usually means penetration depth, lack of fusion, and porosity at the interface.

Penetration is checked by macrosection on the first article and on periodic samples. The section is mounted, polished and etched, then measured at the microscope. A target of 70–80 percent of wall thickness is typical for a load-bearing lap joint. Anything below 60 percent is treated as a process drift, not a one-off.

Porosity is measured against an agreed limit, not a vague standard. A common production limit is total porosity under 2 percent of the weld cross-section with no single pore above 0.3 mm in the load path. Dye penetrant and micro-focus X-ray cover the surface and internal cases respectively.

Hardness mapping across the HAZ confirms that the heat cycle did not soften a case-hardened flank beyond the drawing allowance. We run these checks on first article, then at a defined interval during the run. Inspection reports go out with the shipment on request, which is what most transmission programs actually need.