How to Weld 304 Stainless Steel Thin Sheet
Weld thin 304 with clean, closely fitted edges, stable shielding and enough energy to fuse the joint without lingering over it. TIG offers precise manual control; suitable MIG waveforms and laser welding can suit repeated production. For the roughly 0.8–1.5 mm sheet discussed here, gap control, root protection and welding sequence matter as much as the machine setting.
Choose TIG, MIG or laser for the joint you actually have
For a short visible corner, start by considering TIG (gas tungsten arc welding). For repeated wire-fed seams, compare an appropriate MIG (gas metal arc welding) mode. Laser becomes attractive when narrow welds and limited distortion justify control of the joint and beam delivery. None of these choices fixes poor fit-up.
On a small screen, scroll tables sideways. You can also focus a table and use the arrow keys.
| Process | A useful reason to choose it | What must remain controlled |
|---|---|---|
| DC TIG, with optional pulse | Short seams, changing manual access and careful control of filler addition. | Arc length, dwell, edge alignment and operator consistency. Pulse is an aid, not a requirement for every thin-sheet weld. |
| Short-circuit or controlled short-circuit MIG | A wire-fed route for thin sections with an appropriate machine program. | Wire/gas/program compatibility and fusion. A bead that sits on the surface can hide an unfused interface. |
| Pulsed MIG | Repeated seams where controlled transfer and reduced spatter are useful. | Deposition volume and the lowest stable operating range. A stainless pulse program is not automatically suitable for every submillimeter edge. |
| Fiber laser | Concentrated heating, narrow seams and repeatable production paths. | Gap, seam position, focus, shielding and the complete laser safety arrangement. Added wire changes the process that needs validation. |
Process background: Miller’s thin-sheet guidance and IPG’s laser-welding overview.
Prepare the sheet before adjusting the settings
304 is an austenitic stainless steel. Compared with carbon steel, this family conducts heat away less readily and expands more during heating. The resulting uneven contraction makes thin panels prone to distortion. Thin edges also provide little support for an oversized molten pool. TWI explains these physical differences.
- Confirm material and finish. Check the material specification, actual thickness and coating or protective film. 304L has less carbon, which reduces sensitization risk in relevant welded service; it does not remove the need for good shielding. See what changes between 304 and 304L.
- Clean both sides and the interface. Remove oil, moisture, adhesive and cut-edge oxide with suitable methods. Use stainless-dedicated abrasives and clean tooling. A brush alone may leave tightly attached oxide on a cut edge.
- Check fit-up after clamping. Measure the gap and mismatch along the seam, including between tacks. Use the joint detail’s permitted gap; there is no one gap limit for TIG, MIG and laser.
- Support and tack the real geometry. Support free edges near the seam. Use a planned tack sequence to keep the gap uniform, and remove defective tacks before they become part of the finished weld.
Schematic top views; the gaps are exaggerated. Preparation and tacking principles: Nickel Institute/IMOA fabrication guide, Sections 12.1.1–12.1.2.
Set up TIG, then understand what pulse changes
Use DCEN and a stable, clean arc
Direct current electrode negative (DCEN) is the usual TIG polarity for 304. Pure argon is a common shielding choice. Select the tungsten type, diameter and preparation from the torch and power-source instructions for the current range. Keep the electrode clean and use a short, stable arc with a clear view of both edges.
Check the gas path, cup, torch angle and drafts. A gas lens can improve coverage. Increasing gas flow without checking the delivery arrangement can create turbulence rather than better protection. Keep the torch in position during the required post-flow.
Choose filler by the joint requirement
ER308L is a common filler for 304-to-304. Hobart’s 308/308L data sheet lists Type 304 among its applications. The final choice still depends on the other material, service and specification. Use a rod size that lets you add the needed metal without repeatedly cooling the small pool with oversized additions. See the consumable data.
A tightly fitted butt or outside-corner joint can sometimes be welded autogenously, meaning without filler. That route must still achieve the specified fusion and weld size. Filler may be needed for contour, missing joint volume or weld chemistry; a closed-looking seam does not settle that decision.
Read all four pulse controls together
Peak current sets the high-current level; background current sets the lower level. Peak-time percentage specifies how much of each cycle is spent at peak. Pulses per second (PPS) specify how often the cycle repeats. Background may be displayed in amperes or as a percentage of peak, so check the machine’s convention.
Increasing PPS alone does not necessarily reduce average current or heat input. Travel, arc voltage, peak duration and both current levels still matter. Start with the equipment maker’s guidance for the actual setup and test the joint before adopting a production setting.
A documented 304 example shows why context matters
Miller reports an autogenous outside-corner weld on 22-gauge 304 using 40 A peak, background at 20% of peak, 175 PPS and 75% peak time. The background is therefore 8 A. These are the conditions of its demonstration, not an amperage table for every thin 304 joint.
The idealized waveform averages 32 A: (40 × 0.75) + (8 × 0.25). This calculation explains why “40 A pulsed” is incomplete information; it is not a heat-input or weld-quality calculation.
Miller’s published demonstration, Figure 3. Gauge designation is retained as reported; the example is not extended to different joints, backing or equipment.
Protect the root as well as the face
When a full-penetration weld exposes its underside to air, the torch shield above the sheet does not automatically protect that root. For a corrosion-sensitive or hygienic seam, plan the backing-gas arrangement and acceptance condition before welding.
Argon backing gas can displace air around the hot root. Check leaks, delivery, the outlet and the required purge quality before starting. Keep the purge vented; excess pressure can disturb a thin molten root. Maintain protection while the root remains hot enough to oxidize.
Rough, dark backside oxidation is often called sugaring. The oxide and the chromium-depleted surface underneath can impair corrosion resistance. A clean top bead does not establish the condition below it.
A non-penetrating fillet may not expose a molten backside at all. Purge requirements therefore follow the joint, service and specified root condition, rather than the grade name alone.
Control heat accumulation along the whole panel
One successful tack does not show how a long seam will behave. Later welds start in a warmer part, and shrinkage can change the gap as work progresses.
- Keep the required weld small. Avoid unnecessary reinforcement and wide weaving. Reduce deposited volume only within the drawing’s weld-size and strength requirements.
- Distribute the sequence. Balanced tacking and a planned skip or back-step sequence can spread contraction. If the drawing requires a continuous sealed seam, intermittent welds are not an acceptable final substitute.
- Use backing deliberately. A clean copper chill bar can support the sheet and remove heat where access permits. Avoid rubbing copper onto the stainless or melting the backing into the joint; copper contamination can cause cracking.
- Control the starting condition. Follow the procedure’s temperature and cooling requirements. Compare early and late seams in the cycle, then measure flatness and angle after cooling and unclamping.
See Miller’s sequence/backing guidance and the fabrication guide’s copper-backing precautions.
What changes when you use MIG or laser?
The preparation and inspection goals remain, but the controls are different. An arc-current setting cannot be converted directly into laser watts, and a wire-feed speed does not transfer between unlike MIG programs.
For MIG, match the wire, gas and transfer mode
Select a stainless wire and diameter supported by the machine’s program and the joint. ER308L or ER308LSi may be suitable for 304, subject to the procedure. Keep the liner and feed components clean and dedicated to stainless work.
Gas selection follows the transfer mode. Miller describes helium-rich mixtures for conventional stainless MIG and argon-rich mixtures for pulsed MIG. Do not copy the pure-argon TIG setup into MIG by default.
Hold contact-tip-to-work distance and travel steady. Check fusion on a sectioned trial joint before accepting a low-energy setting because its face looks neat.
For laser, validate beam placement and fit-up
Concentrated energy can produce a narrow heat-affected zone and limit distortion. That advantage depends on delivering the beam consistently to the joint. Record the source, optics, focus, travel, shielding and any wobble or wire feed.
Use the machine builder’s tested starting setup for the specified material and joint. Compare the nominal fit-up with the allowed gap and position limits. Check underfill, penetration and starts/stops; a visually narrow bead can still miss an edge.
Wobble spreads the beam across a wider path. It does not supply missing metal or automatically validate a larger gap. If filler is added, assess the resulting geometry and fusion again.
Process basis: IPG’s laser-welding explanation. Gas delivery is covered further in the laser shielding-gas guide.
Diagnose the defect before changing more settings
Stop and inspect the joint when the pool or result changes. Establish a clean, repeatable setup, then change one variable at a time on representative scrap. The checks below are a troubleshooting order; the accepted repair and final limits belong to the procedure and drawing.
| Symptom | Check first | Confirm after adjustment |
|---|---|---|
| Burn-through or edges pulling back | Gap opening, unsupported edges, excessive dwell or heat building in the part. Correct fit-up before simply lowering current or power. | The joint still fuses at the coldest condition and does not collapse later in the seam. |
| Bead present, interface unfused | Arc/beam placement, contamination, travel, deposited volume and insufficient local fusion. | Representative cross-sections show the required fusion and weld size. |
| Pores or pinholes | Oil or moisture in the joint, gas leaks, drafts and contaminated filler. For laser, also check the beam delivery and process stability. | The specified inspection finds acceptable soundness; cosmetic blending alone is insufficient. |
| Dark or rough root | Missing backing protection, purge leaks, premature shutoff or excessive hot exposure. | Root condition and post-weld cleaning meet the service requirement. |
| Panel bends after release | Excess weld volume, unbalanced sequence, cumulative heating and a fixture that masks movement. | Cooled, unclamped dimensions meet the drawing through the production cycle. |
| Crack at a stop or along the weld | Hold the part. Examine the crack, crater fill, restraint, contamination and filler/base-metal combination before planning repair. | The approved repair and required examination establish the crack has been removed and the cause addressed. |
Stainless defect background: fabrication guide, Section 12.1.6. Appearance alone cannot establish the internal condition.
Remove damaging oxide, then finish for the service
Heat tint is an oxidation signal. It is not a reliable thermometer, and a silver-looking face is not a penetration test.
Cleaning removes contamination. Descaling or pickling removes oxide and, with the appropriate process, the chromium-depleted layer beneath it. Chemical passivation treats an already clean surface; it should not be used as a shortcut over heavy weld scale.
ASTM A380/A380M-25 addresses cleaning, descaling, pickling and passivation practices. ASTM A967/A967M-25 covers chemical passivation treatments and alternative verification tests. Select the specified treatment and acceptance checks for the actual component.
Agree on the required finish, roughness and allowable metal removal before blending the weld. Repeated grinding can leave an attractive surface while reducing the useful section. Chemical treatment also needs suitable handling, rinsing and a layout that does not trap the solution.
Inspect the root, the section and the released shape
Before production, define what the weld must do: carry load, seal, resist the process fluid, meet a finish requirement, or several of these. Then choose evidence that addresses that function.
- Face, root and dimensions: examine accessible surfaces, weld profile, oxidation and cooled dimensions after unclamping.
- Fusion and section: section representative trial joints at starts, steady travel and stops as appropriate. A cut shows only that location; it does not prove the entire seam.
- Functional performance: add the mechanical, leak or nondestructive tests required by the design. A leak test alone does not establish strength or corrosion life.
ISO 5817:2023 covers specified fusion-welded joints but excludes beam welding. ISO 13919-1:2019 addresses electron- and laser-beam welds in steel and certain other alloys. Both scopes start at 0.5 mm thickness. These workmanship quality levels do not, by themselves, establish fitness for a particular product or qualify a welding procedure.
Record the material, joint tolerances, fixture, machine setup, filler/gas, sequence and inspection results in the applicable welding procedure specification (WPS) and work instructions. Confirm the required procedure and personnel qualification route for the product, rather than assuming a successful sample is enough.
Plan fume and beam controls before a trial
Stainless welding still needs fume control, even when little smoke is visible. Assess the material, consumables and any coatings; capture fume at source and provide suitable respiratory protection where the exposure assessment requires it. Position extraction so it captures the plume without disrupting shielding. HSE explains the control hierarchy.
For a Class 4 laser source, direct and reflected beams present serious eye and skin hazards. Establish an assessed laser-controlled area, engineered containment, appropriate interlocks and trained operation, with wavelength-appropriate protection for those exposed. Eyewear alone does not contain the beam. See the FDA hazard classification and IPG’s enclosed-system safety example.
Share the material specification, thickness, joint drawing, permitted gap, seam length and required finish with Oceanplayer Laser. Include the distortion, penetration or leak requirement so the trial can be assessed against the part’s purpose.
Technical references
- Nickel Institute/IMOA: Practical Guidelines for the Fabrication of High Performance Austenitic Stainless Steels. Original guide hosted by SCI; Sections 12 and 14 distinguish general fabrication practice from high-alloy-specific requirements.
- Miller: Successfully Welding Sheet Metal with MIG and TIG. Fit-up, backing and sequence.
- Miller: Pulsed TIG Welding Stainless Steel. Source of the 22-gauge 304 demonstration.
- Hobart 308/308L data sheet. Filler classification and listed base-metal applications.
- TWI: Avoiding Heat Tint. Root oxidation, corrosion and backing-gas principles.
Process, safety and standard-scope references are also linked beside the relevant guidance. The governing drawing, service specification and adopted standard edition determine production requirements.