Laser and electron-beam welding concentrate enormous energy into a keyhole of molten metal — precise, deep and fast, but still fusion. In crack-sensitive materials (nickel superalloys, high-strength steels, thick copper) the melt-resolidify cycle is exactly what drives hot cracking, porosity and the heavy NDT regimes that go with them. Friction stir welding never melts the metal: solidification defects are eliminated by mechanism, the forged weld zone is often stronger than parent, and no vacuum chamber is needed. Beam processes keep the advantage for very deep single-pass penetration, micro-joints and hermetic sealing of tiny packages.
Why beam welds crack in the materials that matter most
High-value alloys are optimised metallurgies. Melting them locally and letting them re-solidify undoes that optimisation at the worst possible place — the joint:
- Nickel superalloys (the Inconel and Haynes families) suffer liquation and strain-age cracking during welding and post-weld heat treatment; the standard response is tightly controlled beam parameters plus extensive NDT on every weld.
- Creep-resistant and reduced-activation steels (9Cr class) derive their high-temperature strength from carefully engineered microstructures. A fused and re-solidified zone is a different, weaker material — the well-known soft-zone problem of fusion-welded 9Cr steels.
- Copper conducts heat so well that fusion processes struggle to hold a stable pool; electron beam does it, but only inside a vacuum chamber sized to the part.
Each of these is a melting problem. Removing the melting removes the problem class.
What the solid-state route changes
Published results in exactly these material families show what forging below the melting point means in practice:
- Superalloys. Friction stir welds in Inconel 718 have achieved around 100% joint efficiency, with stir-zone grain refinement raising tensile strength from 886 MPa (parent) to 1,135 MPa and hardness from about 273 HV to 352 HV — the joint outperforming the base metal (Song and Nakata). Work on Haynes 282 reports full joint efficiency, with post-weld heat treatment increasing strength further and no grain growth during PWHT.
- Steels. The process produces welds in creep-strength-enhanced and reduced-activation ferritic-martensitic steels without the melted-and-recast zone that degrades fusion welds — which is why fusion-averse sectors such as fusion-energy research evaluate it for structural steels.
- Copper. The canonical case: Sweden’s SKB selected friction stir welding over electron beam welding to seal the 50 mm-thick copper canisters for spent nuclear fuel, after years of side-by-side development (SKB report TR-04-16). Thick-section copper, including single-pass and bobbin-tool approaches, is established territory.
- Dissimilar joints. Aluminium to copper — a combination fusion processes handle poorly because of brittle intermetallics from mixing in the melt — is a routine friction stir welding application for thermal and electrical hardware, treated at StirLight as a development activity rather than a catalogue procedure.
| Criterion | Laser welding | Electron beam | Friction stir welding |
|---|---|---|---|
| Physical state | Fusion (keyhole melt) | Fusion (keyhole melt) | Solid state |
| Hot / liquation cracking in Ni alloys and steels | Managed, not eliminated | Managed, not eliminated | Eliminated by mechanism |
| Porosity | Process-dependent | Low, but present | None from solidification |
| Weld zone versus parent | Cast microstructure, typically weaker | Cast microstructure, typically weaker | Forged, fine-grained — can exceed parent |
| NDT burden | High in critical alloys | High in critical alloys | Fewer defect classes to inspect |
| Vacuum requirement | No (shield gas) | Yes — chamber sizes the part | No |
| Very deep single-pass steel penetration | Moderate | Excellent (100 mm+) | Thickness limited by tool technology |
| Reflective / conductive metals (Cu, Al) | Difficult (reflectivity) | Good, in vacuum | Good, in air |
| Micro-joints, hermetic packages | Excellent | Excellent | Not the right tool |
| Capital and running profile | Beam source, chiller, fume control | Chamber plus pump-down time | Machine-tool class; no consumables |
When the beam is still the right answer
- Very deep, narrow, single-pass penetration in steel — electron beam’s parallel-sided welds through 100 mm and more remain unmatched.
- Micro-scale and hermetic package welding — features far below tool scale.
- Fully enclosed geometries where no backing support or tool access exists.
- Established beam lines with qualified procedures and acceptable yields — the argument for friction stir welding is strongest where cracking, NDT cost or vacuum-chamber logistics are actively hurting.
It is also fair to state this process’s own constraint in these materials: welding steels, superalloys and thick copper demands advanced tool materials (PCBN, W-Re and similar) and serious machine stiffness. Tool cost and wear are real engineering line items. They are managed with parameter development, pre-heating and in-process monitoring, but a supplier who does not mention them is not being straight.
Frequently asked questions
Can friction stir welds in superalloys match electron beam quality?
In published studies the joint does not just match the parent. Grain refinement can make the stir zone stronger and harder than the base material. Beam welds, being cast structures, rarely achieve that.
Does friction stir welding of steel or Inconel need shielding gas?
An inert shroud, typically argon, is used to protect the tool from oxidation at high temperature, a modest provision compared with a vacuum chamber.
When is electron beam still the right answer?
Very deep single-pass steel penetration, micro-scale hermetic packages, and fully enclosed geometries with no tool access or backing.
Next step A structured feasibility trial answers the question on your geometry, with measurements rather than argument.