Application of TRUMPF laser welding technology to inflatable cabinets
This page explains where TRUMPF laser welding fits in C-GIS inflatable cabinet production and where it does not. It is written for design engineers and process planners who must choose a joining method for thin stainless enclosures. After reading, you can judge seam type, filler choice, and the parts that are better CNC machined.

What this article covers
Seam design, material behavior, gas-tight limits, and the machining steps around a welded cabinet shell.
Why C-GIS cabinet shells are welded, not bolted
A cabinet-type gas insulated switchgear enclosure holds pressurized SF6 or a dry-air mix. The shell is the pressure boundary, so every joint has to hold a leak rate spec, not just a mechanical load. Bolted flanges with gaskets work, but each gasket is a potential leak path and a service item that ages. A continuous welded seam removes that path.
The walls are thin. Most inflatable cabinet shells run 1.5 mm to 3 mm in 304 or 316L stainless, sometimes with a 4 mm to 6 mm reinforcement ring at the flange. Heat input has to stay low or the panel distorts and the flange face will not seat. This is the constraint that shapes every process choice downstream.
The enclosure also carries internal busbars, insulators, and a bursting disc. Weld spatter inside the tank is a dielectric risk, and grinding dust is worse. A process that welds without spatter and needs little post-weld cleanup has a real advantage here.
How the TRUMPF process is applied to a cabinet shell
TRUMPF solid-state laser sources operate at 1030 nm to 1080 nm, which couples well into stainless steel. For a 2 mm 304 wall, a typical bead-on-plate weld runs 1.5 kW to 3 kW at 3 m/min to 6 m/min, with a spot of 0.3 mm to 0.6 mm. Those are starting points, not fixed recipes. Welding speed is set by penetration depth and by how much distortion the fixture can hold.
Keyhole mode is used where full penetration is needed, such as the corner joints of the tank body. Conduction mode is used on visible outer seams and on thin return flanges, where a smooth cap bead matters more than depth. Filler wire is usually not needed on 1.5 mm to 2 mm sheet, but 1.4301 and 1.4404 grades can crack in the crater without a small amount of 316L filler or a ramp-down profile.
The joint design is the part most often wrong on first drawings. A square butt with zero gap works up to about 2 mm. Above that, a small V or a stepped lap gives the beam somewhere to go. A lap joint on 1.5 mm sheet is forgiving on gap, which is why many cabinet corner details are laps rather than butts.
Process comparison for cabinet shell joints
Typical values for 1.5–3 mm 304 / 316L stainless in an enclosure context.
| Method | Wall thickness | Seam speed | Post-weld work |
|---|---|---|---|
| TRUMPF fiber laser | 1.5–3 mm | 3–6 m/min | Light brush, no spatter |
| TIG | 1.5–3 mm | 0.15–0.4 m/min | Pickle and passivate |
| MIG | 3 mm and up | 0.4–0.8 m/min | Grind spatter, passivate |
| Resistance spot | 0.8–2 mm | Fast per spot | Sealant required |
| Laser-hybrid | 3–6 mm | 1–2.5 m/min | Wire cleanup |
When laser welding is the wrong answer
Thick flanges and cast bosses are the first case. A 10 mm 316L flange welded to a 2 mm wall needs either a hybrid process or a machined step so the beam sees a consistent thickness. The laser does not care that the wall is thin; it will punch through if the parameter set is tuned for the flange.
Deep, closed geometries are the second case. A beam needs line of sight. If a stiffener sits inside a 200 mm deep box behind a lip, no fiber delivery head reaches it. Those joints go to TIG, or the design is changed to move the stiffener outside.
The third case is cost. A laser cell earns its keep on repeat production with a stable fixture. For a one-off prototype shell, TIG plus a machined flange is usually cheaper and faster, even with the extra cleanup.
Where CNC machining sits in the same build
Welding does not produce a finished cabinet. The flange faces, the busbar penetrations, the insulator mounting holes, and the seam where the cover seats all need metal removed to a tolerance. A welded shell that has been distorted by heat is a poor starting point for a tight bore, so the sequence matters.
The usual order is: form the sheet, weld the shell, stress-relieve if the grade allows, then machine the flange face and the penetration bores. Flange flatness on a C-GIS tank is commonly held to 0.05 mm to 0.1 mm across the sealing face. Bore position for busbar penetrations is often ±0.05 mm, and the bore itself may be held to ±0.005 mm on diameter when it takes an O-ring or a ceramic insulator.
GreatLight machines these welded shells on 3-axis and 4-axis mills and on mill-turn centers, with a Ø400 mm rotary table for the flange work. Wall thickness on a welded tank is not uniform, so we touch off on the actual surface rather than trusting the drawing. A 4,000 mm maximum processing size covers most cabinet lengths. If the shell arrives already welded, we check flatness before the first cut.
Material and finish notes for welded enclosures
304 and 316L are the default grades. 316L resists pitting better in coastal or humid switchgear rooms, and it welds with less sensitization because of the low carbon. 304 is cheaper and fine for indoor, dry conditions.
Aluminium shells appear on some dry-air designs. ADC12 die castings are used for end plates and small covers rather than the main tank, because cast aluminium porosity creates leak paths that a laser seam cannot close. If a casting is welded, the porosity has to be mapped first.
After welding and machining, the inside of the tank is usually cleaned and passivated. Outside surfaces get bead blasting, brushing, or powder coating. Laser marking for nameplates and terminal IDs is done after finishing, with a minimum character height of 1.5 mm so it stays legible after coating.
Questions engineers ask before specifying this process
Can a laser-welded cabinet shell pass a helium leak test?
Yes, if the seam is full penetration and the parameter set is stable. Helium leak testing is done after cleaning and before final assembly, and it catches porosity and crater cracks that a visual check misses.
The common failure is not the seam itself but the start and stop point. Overlap the start by 3 mm to 5 mm or program a ramp-down so the crater does not stay open.
Do we need filler wire on 1.5 mm stainless?
Usually not on a square butt with tight fit-up. Autogenous welds on 304 and 316L are common at that thickness.
Add a small amount of 316L filler when the joint has a gap over 0.1 mm, or when the alloy is prone to hot cracking. Filler also helps if the seam will be machined flush afterward.
How much distortion should we expect?
On a 2 mm panel with a 1 m seam, expect 0.3 mm to 1 mm of bowing depending on speed and fixture. Clamping and a copper backing bar reduce it.
Distortion is why the flange face is machined after welding. Do not expect the welded shell to hold a sealing flatness on its own.
Can the welded tank be stress-relieved?
Solution annealing of 304 or 316L needs roughly 1040 °C to 1120 °C followed by rapid cooling. Most cabinet shells are too large for a vacuum furnace at that size, and the quench can re-introduce distortion.
In practice, most builders control distortion with the weld sequence and fixture instead, then machine the critical faces after cooling.
What file format and detail do you need for a quote?
Send the 3D model plus a 2D drawing with the welded joint callouts, the post-weld machining tolerances, and the material grade. Note which surfaces must stay unmachined.
We return a quotation and a DFM analysis within 12 hours. Uploads are kept confidential and an NDA is available on request.
Welded shell needs machining to tolerance?
Send your drawing and we will review the joint design, the post-weld datums, and the flange flatness callout before quoting.
12-hour quote100% inspection±0.005 mm