Metal Plate Welding: A Complete Beginner’s Guide

There’s a particular kind of silence in a workshop when a 10 mm steel plate won’t cooperate. The bead looks fine, the arc is steady, and then the joint cracks the next morning — because the weld wasn’t wrong so much as the plan behind it was. Most plate welding failures trace back to decisions made before the torch ever lit: which process, which joint design, how much heat, and how much restraint. This guide walks through those decisions in the order professional fabricators make them, so you can weld plate that holds up after it cools.

I’ll rewrite this section, tightening it to fit the 560–840 word range while keeping all the facts and links.

Let me count the current fragment’s approximate word count and plan the rewrite.

What Metal Plate Welding Actually Is (and Why It Fails Differently Than Tube)

At its most basic, metal plate welding joins flat pieces of metal by fusing the molten base metal together, often with added filler metal, to form a permanent, load-bearing joint. Unlike bolting or fastening, a proper weld becomes a single metallurgical structure whose strength depends less on how much metal you add than on how well the joint was prepared, how heat was managed, and how the metal was allowed to move as it cooled.

Vocabulary matters, because the gap between a passing weld and a failing one often lives in the terms. According to AWS A3.0, joints are described by how the plates are positioned, while welds are described by how they are applied. The common joint types:

Joint type	How the plates meet
Butt	Edge to edge, in the same plane
Tee	One member meets the face of another
Lap	Two overlapping members
Corner	Two members joined at their edges
Edge	Two parallel members joined along their edges

Weld types then fall into categories such as groove welds (complete joint penetration, or CJP, and partial joint penetration, or PJP), fillet welds, and plug or slot welds. The CJP versus PJP distinction dictates how much edge preparation is required and whether full fusion through the thickness is needed to satisfy a governing code such as AWS D1.1.

Why Plate Behaves Differently Than Tube

Weld a length of pipe and you’ll notice it wants to stay round. That’s geometry working for you: curved tubing is self-reinforcing and naturally resists the pull of contraction. A flat plate has no such advantage. Heat can sink in two directions across width and length, so welding stresses are far less restrained by shape and more likely to surface as visible movement.

That is why plate is notorious for angular distortion and bowing, while pipe is more forgiving. Every time the weld pool cools and shrinks, it drags the surrounding metal with it. On a flat sheet that pull has nowhere to go, so the plate twists, curves or lifts at the edges. On heavy structural plate those forces can be immense, and without intervention they can pull an otherwise sound weld out of alignment.

Thickness, Heat Source Control and Heat Input Control

Plate welding covers an enormous range, from delicate 24-gauge sheet metal up to structural plate several inches thick, with the approach changing drastically as mass increases. Thin stock holds very little heat and burns through easily. Thick stock behaves like a massive heat sink, demanding more energy to reach fusion without overworking the joint.

Heat source control and heat input control are not the same thing. The first is about the process itself, such as how concentrated the arc is. The second is the measurable energy going into the weld. As laid out by TWI, that energy drives shrinkage and is calculated from voltage, amperage and travel speed combined. You can have the right amperage and still wreck the joint if your travel speed is far too slow, because heat input skyrockets regardless.

The Six Forms of Distortion To Know

Movement isn’t random. TWI identifies six main forms of welding distortion on plate: longitudinal shrinkage, transverse shrinkage, angular distortion, bowing and dishing, buckling, and twisting. Most jobs won’t show all six, but angular distortion plus bowing on a flat plate is extremely common.

TWI’s typical shrinkage figures put this in perspective: fillet welds can shrink around 0.8 mm where the leg length doesn’t exceed three-quarters of plate thickness, butt welds around 1.5 to 3 mm for a 60° V joint, with longitudinal shrinkage around 0.8 mm per 3 m of fillet weld and around 3 mm per 3 m of butt weld. Small on paper, but across a large fabrication they add up fast.

On materials that need extra care, such as aluminium prepped with an aluminum cleaner for welding or specialized work like aluminum boat welding, movement shows up even more because of aluminium’s high thermal conductivity. The principle holds for any flat stock: no curved support means no built-in resistance to distortion.

The Core Mental Model

The takeaway is simple. A weld on plate isn’t just a bead of fused metal; it’s a metallurgical joint that has to absorb the full force of thermal expansion and contraction. Joint types, weld types, thickness, heat input and the six forms of distortion all feed into that truth. Get preparation and heat management right and the forces of cooling can be managed. Ignore them and the strongest-looking bead in the shop can reach room temperature bowed, twisted and out of spec.

Choosing Your Process: MIG vs TIG vs Stick vs Flux Core for Plate

Every plate job starts with a decision that costs nothing to make and everything to make wrong: which arc process you will run. Choose well and the joint fills in front of you. Choose badly and you spend an afternoon fighting burn-through on 3 mm sheet, or turning a 25 mm structural joint into a runaway puddle. The four processes a plate welder realistically reaches for are GMAW (MIG), FCAW (flux core), SMAW (stick) and GTAW (TIG). All of them melt base metal and filler with an arc. What differs is how the heat is delivered, how fast metal lands in the puddle, and how precisely you can hold the bead. Here is the honest comparison, using the process data published by TWI.

Process	Consumable	Manual thickness range	Deposition rate	Best fit
GMAW (MIG)	Solid wire, gas shielded	Thin sheet through heavy plate; short-circuit commonly limited to 0.25 in (6.4 mm) or less in structural work	2–5 lb/hr short-circuit; 5–12 lb/hr spray	General fabrication, production, thin to medium plate
FCAW (flux core)	Tubular wire with flux core, gas-shielded or self-shielded	About 1/8 in (3 mm) through several inches, with especially strong industrial use on medium and thick carbon-steel plate	8–25 lb/hr	Heavy structural plate, outdoor and field work
SMAW (stick)	Flux-coated stick electrode	About 1/8 in (3 mm) up to several inches using multiple passes	1.5–3 lb/hr	Site work, dirty or mill-scaled steel, thick plate
GTAW (TIG)	Non-consumable tungsten plus hand-fed filler r	Non-consumable tungsten plus hand-fed filler rod	Very thin sheet and foil through roughly 1/4 in (6 mm) in many manual applications	0.5–2 lb/hr	Stainless, aluminium, thin material, cosmetic bead

GMAW (MIG): the productivity workhorse

Solid wire feeds continuously through the torch and an argon/CO₂ blanket shields the puddle, which makes MIG the fastest general-purpose option in the shop and the easiest to teach. TWI’s manual range runs from thin sheet right through heavy plate, but short-circuit transfer, the low-heat stubbing arc you can hear spitting, is commonly limited to 0.25 in (6.4 mm) or less in structural applications. Practical reading: short-arc MIG for sheet and light gauge, spray transfer for 6 mm and above. It is also the most versatile of the four, handling steel, stainless and aluminium with only a gas and wire change, provided you have the right aluminium MIG setup for spoolgun work.

FCAW (flux core): the thick-plate and field workhorse

Tubular wire carries flux inside its core, so the arc is shielded by a slag layer rather than by gas alone. Self-shielded wires need no cylinder at all, which is why flux core is the default for structural steel outdoors, on site and in the rain. TWI puts the range at about 1/8 in (3 mm) through several inches, with especially strong industrial use on medium and thick carbon-steel plate. At 8 to 25 lb per hour, FCAW has the highest deposition rate of the four. The trade-off is mess: more spatter, more slag, a rougher bead that needs cleanup, and a learning curve where wire angle and travel speed matter more than they do on MIG.

SMAW (stick): slow, but it goes anywhere

A flux-coated rod melts at the tip and the coating does the shielding, so there is no gas, no wire feeder and nothing to trip over in the mud. TWI puts the usable range at about 1/8 in (3 mm) up to several inches using multiple passes. Deposition runs 1.5 to 3 lb per hour, the slowest of the four. Yet stick is generally preferred for 6 mm to 12 mm structural plate, and above 12 mm in field conditions, because of its deep penetration and its tolerance of dirty, painted or mill-scaled surfaces. For someone who owns one machine and welds outside, that tolerance is worth more than speed. There is a solid case for it in why SMAW still wins.

GTAW (TIG): the cleanest bead, the lowest output

A non-consumable tungsten makes the arc, and filler rod is fed by hand into the puddle. That gives you total control over heat and a bead with no spatter at all. TWI’s range runs from very thin sheet and foil through about 1/4 in (6 mm) in many manual applications, with productivity falling off sharply on thicker sections. Deposition is only 0.5 to 2 lb per hour. TIG is the preferred process for stainless steel in food-grade and pressure applications and the standard choice for aluminium, where stick is not suitable. Research comparing MIG and TIG on 5 mm aluminium plate found MIG faster with generally higher penetration and stronger joints, while TIG remained the better choice for finer heat control, which is the trade in one sentence.

Pick in seconds: a decision framework

• Thin sheet, stainless, aluminium or a visible cosmetic bead: TIG, or MIG for aluminium production work.
• General fabrication, mixed thickness, clean indoor shop: MIG. This is the default answer when nothing else points clearly elsewhere.
• Plate over 12 mm, structural carbon steel: FCAW indoors, SMAW in the field.
• Dirty, painted or mill-scaled steel, wind, no gas supply: stick, no contest.
• Repeat production, thick section, consistent output required: FCAW or spray MIG, where operator fatigue is the real constraint.

One last check before you commit. If the joint falls under a structural code such as AWS D1.1, the process is not a free choice. The procedure and the welder have to be qualified, which usually means the process, filler, position and heat inputs you documented on your test coupons have to match the production weld.

Reading the Plate: Thickness, Grade and Material Chemistry Change Everything

Before you touch the machine, you need to know what you are actually welding. Two pieces of steel that look identical on the bench can behave completely differently because one is 6 mm and the other is 30 mm, or because one came off a mill with mill scale still on it and the other was shot-blasted. TWI’s arc process guidance puts it plainly: joint design, consumable choice and heat input all follow from the material and thickness, not the other way around.

The practical sequence: identify the material family, measure the thickness, check the surface condition, then decide joint design and process.

Know Your Material Family

Material	What changes for you
Carbon steel	The default. E-series electrodes (E6013, E7018) and ER70S-6 wire. Relatively forgiving if you control heat and clean the surface.
Low-alloy steel	Stronger, less forgiving of hydrogen. Needs low-hydrogen consumables and tighter preheat and interpass control. Filler selection gets more specific, as covered in the alloy steel welding electrode guide.
Stainless steel	Austenitic grades (304, 316) weld with heat input and interpass limits to avoid sensitization. Ferritic and martensitic grades behave more like carbon steel, with martensitic grades being the most crack-prone.
Aluminium	Heat races away from the arc. Different consumables, different gas, different cleaning. Our aluminium MIG guide and aluminium prep guide cover the specifics.
Galvanized or coated steel	Fume hazard, and a coating that can cause porosity. Needs its own approach.

Why Carbon Equivalent and Hydrogen Control Matter

Carbon equivalent is a single number that lumps a steel’s chemistry into one figure: how much carbon it effectively has once alloying elements are counted. It matters because hydrogen and a crack-sensitive microstructure go together. Higher-strength and higher-alloy steels have higher carbon equivalents, so their welds and heat-affected zones are less forgiving of hydrogen. They also have higher hardenability, so a hot HAZ can cool into hard, crack-prone martensite.

Under AWS D1.1-type practice, thick or higher-strength carbon steel is treated as hydrogen-sensitive. That changes what a procedure must specify: low-hydrogen consumables, preheat and interpass control become part of the procedure rather than optional extras. You can see the framework in the AWS codes and standards library, but the behaviour is what matters on the floor: on heavy plate, a hydrogen-cracked joint is one you will not know about until it fails.

Restraint makes it worse. A rigidly clamped joint cannot relieve contraction stress as the weld cools, so the same weld that passes on a loose bench plate can crack in a frame. More restraint, more hydrogen, more crack sensitivity. That is why fit-up and clamping in the joint preparation section are not just cosmetic.

Preheat and Interpass Temperature, in Plain Terms

Preheat means warming the plate before you start; interpass temperature is the ceiling the metal may reach between passes. Both limit how fast the weld area cools, which limits how hard and crack-prone the HAZ becomes. They are not the same thing. For thin mild steel, neither is needed. For thick, higher-strength, or heavily restrained carbon steel, both may be required, and the amount is set by the procedure, not by feel.

Stainless adds a different reason to control heat. Austenitic stainless exposed to too much heat in the wrong temperature window can become sensitized, where chromium carbides form at grain boundaries and the HAZ loses corrosion resistance. The practical rule: austenitic stainless needs lower heat input than carbon steel of the same thickness, and interpass temperature must be capped. Keep heat input low, keep the bead tight, let the part cool between passes.

Surface Condition: Mill Scale, Rust and Coating

Mill scale is the dark, flaky oxide left on plate straight from the rolling mill. It is hard, it floats off unpredictably during welding, and it traps gas. It is a well-documented contributor to undercut, because it lifts away while the arc passes and leaves a groove along the toe of the bead. Rust and dirt do the same thing differently: they introduce moisture and gas, which show up as porosity. Poor surface cleaning is one of the most commonly reported causes of weld defects in shop practice, which is why grinding and cleaning the joint faces is not an optional finishing step.

Galvanized steel deserves separate treatment. Welding through zinc coating releases zinc

oxide fume, and that fume is a leading cause of metal fume fever. OSHA’s welding fume fact sheet specifically flags cleaning surfaces of toxic coatings before welding as the control measure. In practice that means grinding the coating back well beyond the weld area, or at minimum treating it as a coated product with a dedicated procedure and fume extraction. Where coating must stay, do not simply ignore it.

Aluminium has its own surface issue: a tenacious oxide layer that must be removed mechanically or with a dedicated cleaner, and kept clean until the moment you weld. Our guide to prepping aluminium correctly walks through the chemistry and the tools.

How Thickness Drives Joint Design and Pass Limits

Thickness decides the joint, and the joint decides everything else. Up to roughly 3 mm, a square butt with minimal root gap is fine. From about 3 to 12 mm, a single-V groove is typical. Above that you move to double-V, U or J preparations, because a deep narrow V means a long, hot, sagging weld. As a rule of thumb, butt welds on carbon steel over about 6 mm should be bevelled, and the bevel angle trades cost against control: 45 degrees is easy, 60 degrees deposits more metal but gives a flatter, more controllable root pass, and U or J preparations deposit even more but deliver the best access for thick sections.

Process limits follow the same logic. Short-circuit GMAW is commonly limited to around 6.4 mm (0.25 in) for structural work, and it is the transfer mode that suits thin material. Spray and pulsed GMAW extend that usable range. FCAW and SMAW then carry the heavy plate, where anything above 15 to 20 mm is normal territory.

Single-pass welding has real limits too. A single pass on thick plate means a long, wide, hot bead with deep penetration and high restraint, which is exactly the recipe for a cracked HAZ and a bowed plate. As a rough guide, a single V-groove butt on carbon steel becomes impractical somewhere around 12 to 15 mm; beyond that, multi-pass, multi-layer welding with controlled interpass temperature is the norm.

Aluminium: When Heat Runs Away

Aluminium’s thermal conductivity is several times that of carbon steel. Heat does not stay near the arc; it races along the plate and out through the edges. That single fact explains almost every technique difference: lower amperage settings, filler added quickly and deliberately, AC or pulsed current, and interpass temperature kept firmly in check. Thick aluminium sections often need a heat sink, meaning copper or aluminium backing bars placed under the joint to pull heat away and control the puddle.

Aluminium is also where aluminium flux core welding and aluminium flux welding have become genuinely useful, because the extra heat of a flux-cored wire helps compensate for the conductivity problem. If your plate is aluminium, plan the heat budget before you plan the weld.

Get these four things right first: what the material is, how thick it is, how clean the surface is, and how much the joint is restrained. Everything downstream, from joint design to consumable to preheat, follows from that read of the plate.

Joint Preparation: Bevel, Root Gap, Fit-Up and Tacking

Most plate weld failures are decided before the arc ever lights. A bead can look perfect and still crack, sink or distort, and the reason is usually a preparation decision: too much bevel, too little root opening, a fit-up that loaded the joint with restraint, or a tack buried inside the final pass. Prepare the joint properly and the welding becomes straightforward. Prepare it badly and no parameter tweaking will rescue the weld.

Pick the Groove That Matches the Plate

Groove design is really a decision about how much filler metal you intend to add and how deep the arc has to reach.

Groove type	Where it fits	Trade-off
Square butt	Thin plate, about 1/8 in (3 mm) or less	Lowest heat input, but the torch must bridge the gap
Single-V	General fabrication from about 1/4 in upward	Easy to set up, but the hottest, deepest groove
Double-V	Thicker plate, both sides accessible	About half the weld metal of a single-V
Single-U	Thick plate where a narrow groove saves metal	Saves filler, needs shaping roll or gouging machine
Single-J	Thick plate, one side accessible	Similar savings to U, different tooling
Double-U	Very thick plate, both sides accessible	Least filler in the family, highest tooling cost

Bevel Angle Is a Heat Budget

The groove cross-section sets the volume of filler metal, and filler metal sets both heat input and shrinkage. A wide 45° bevel on half-inch plate puts a lot of metal into the joint; a 60° bevel narrows the groove and cuts that volume. Going further, a J- or U-groove cuts the weld metal requirement again, because the groove is parallel-sided rather than tapered.

You can trade bevel for root opening instead. Structural joint design guidance suggests reducing the bevel angle and increasing root opening, on the order of 1/32 in to 1/16 in, to lower the total deposited metal. The rule of thumb from TWI’s arc welding engineering guidance is simple: minimise the preparation angle and the root gap, provided the weld can still be made satisfactorily. Detailed requirements live in AWS codes and standards such as AWS D1.1.

The clearest example is the double-V. It splits the groove between two sides, so it needs about half the weld metal of a single-V preparation, cutting heat input and distortion with it. That is why double-V is standard for thick plate when both sides are accessible.

Poor fit-up and over-welding both increase distortion for the same underlying reason: weld shrinkage is proportional to the amount of weld metal deposited.

Root Gap: The Fastest Lever You Have

Root opening is the one dimension you can change without a grinder or a shaping roll, so it is the fastest lever on heat input and cost. Closing the gap for a tidier joint is usually backwards: a tight groove risks a cold lap or incomplete root penetration, while a wider gap needs less filler and spreads heat over a longer travel. On thin plate the reverse applies, because too much gap lets the puddle drop through.

Process	Typical root opening	Watch for
GMAW, short circuit	1/8 in (3 mm) or less	Wider gaps push the arc into spray, risking burn-through
SMAW	1/8 in (3 mm) light plate, up to 1/4 in (6 mm) heavy	Small gaps cause stubbing and poor root fusion
FCAW and SAW	Around 1/4 in (6 mm), more for SAW	Extra gap adds heat and slows travel

These are starting points only; your procedure specification and code govern.

Fit-Up: Align the Joint, Don’t Force It

Fit-up is where restraint problems begin: hold alignment through the first passes without locking the joint so tightly that shrinkage has nowhere to go.

• Check alignment before tacking, not after. Run a straightedge across the joint and a feeler gauge or wedge at the ends and middle. Hi-lo misalignment beyond your procedure limit shows up later as a wandering bead. Clamp from the centre out, since clamping one end first pulls the joint out of true.
• Set the gap with spacers or a backing bar. Clamps hold pieces together, they do not hold a gap, and that gap must stay constant as the joint heats and moves.
• Do not over-restrain. Forcing a joint that will not fit adds stress as it cools, and restraint during welding raises residual stress and the risk of hydrogen cracking.

Tacking: Small Welds That Set Up the Big One

• Match the process and consumable you will use for the joint. A tack made with a different process or electrode can introduce defects the final weld cannot fix.
• Make tacks by a qualified welder to the qualified procedure, and keep them small so they do not trap heat. AWS D1.1 expects tacks to be made by qualified welders and inspected, with cracked tacks removed before welding continues.
• Keep tacks off the ends and space them along the joint. Tack right at the corner burns through thin plate and cracks the corner.
• Clean them out. Grind or gouge tacks back so the root pass fuses to clean parent metal. Tacks left proud under the root pass cause slag inclusions and lack of fusion that inspection will never catch.

Setting the Machine: Amps, Volts, Wire Feed, Travel Speed and Heat Input

Ask five plate welders for their settings and you get five answers: one quotes amperage, one quotes wire feed speed, one gives an eyeballed travel speed. What decides how the plate behaves is heat input, and once that idea clicks, every knob on the machine makes sense.

Heat Input: The One Number That Ties Everything Together

Heat input (arc energy) = (volts × amps × 60) ÷ travel speed in mm/min

Volts times amps is arc power in watts; times 60, joules per minute; divided by travel speed in mm/min, joules deposited per millimetre of weld. Divide by 1,000 for the kJ/mm figure you will see on a procedure or a welding code.

• Amps set how hard the arc works and how fast filler melts into the puddle.
• Volts control arc length, shaping the bead and the spatter.
• Travel speed controls how much heat the joint absorbs. Hardest to hold by hand, biggest effect on the result.

Scenario	Settings	Heat input
6 mm plate, steady MIG bead	24 V, 200 A, 200 mm/min	1.44 kJ/mm
Same weld, faster travel	24 V, 200 A, 300 mm/min	0.96 kJ/mm
3 mm plate, TIG cap pass	12 V, 90 A, 100 mm/min	0.65 kJ/mm
25 mm plate, slow root fill	24 V, 300 A, 80 mm/min	5.40 kJ/mm

MIG: You Set Wire Feed Speed, Amps Follow

On a constant-voltage MIG machine, wire feed speed is the control and amperage is the consequence: more wire, more aggressive arc. Feed too little and the wire melts back to the contact tip, the arc sounds hollow and the puddle starves. Feed too much and you get stubbing, heavy spatter and a wide, flat bead that burns through thin material.

Voltage is arc length control. Too low and the wire shorts into the puddle; too high and the arc stretches, the bead goes wide and domed and spatter rises. Keep arc length about equal to the wire diameter, hold CTWD (contact tip to work distance) at about 1/2 in (12 mm), and keep stickout, the exposed wire between tip and gas cup, near 1/4 in (6 mm). Longer stickout preheats the wire, lowering deposition and adding spatter; shorter and you hit the work with the nozzle.

Push the torch for flat and horizontal fillet work, pull it for vertical-up and overhead, at 5 to 15 degrees. Pushing gives better penetration and a flatter bead but throws spatter forward. Pulling controls the puddle better, which is why it wins on vertical-up.

Transfer mode sets the usable envelope. Short-circuit GMAW suits thin material, spray GMAW handles thicker plate at higher current, and pulsed GMAW cycles the current so transfer happens in a controlled spray rather than dipping into the puddle, pushing MIG past short-circuit limits. TWI’s overview of arc processes covers the trade-offs.

Stick (SMAW): Diameter Sets the Current, and You Always Pull

With stick, amperage comes from electrode diameter, not the machine dial most people reach for first.

Electrode	Typical amperage range
1/16 in (1.6 mm)	20 to 80 A
3/32 in (2.4 mm)	60 to 130 A
1/8 in (3.2 mm)	80 to 160 A
5/32 in (4.0 mm)	110 to 200 A
3/16 in (4.8 mm)	130 to 250 A

Below its range you get poor fusion and a weak, cold-looking bead. Above it the tip smears, the coating splits, spatter flies and you undercut the toe of the weld. The electrode is always dragged, never pushed, at roughly 10 to 15 degrees; pushing causes stubbing and erratic fusion. The case for stick welding helps if you are still choosing a process, and the all-position rod guide covers electrode selection by position.

TIG (GTAW): Precise Control, Limited Reach on Thick Plate

TIG gives the cleanest, most controllable bead of any arc process, so it dominates thin stainless, aluminium and cosmetic work. The torch runs DCEN (direct current electrode negative) for steel and stainless. Aluminium demands AC, because the reversing half of the cycle scrapes the oxide layer away; without that cleaning action the puddle will not wet properly.

Tungsten follows the job: thoriated and lanthanated types work well for DCEN steel and stainless, while pure tungsten, usually zirconiated, suits AC aluminium. Amperage is set by eye and feel, roughly 100 A per 0.1 in (2.5 mm) of material thickness on steel. Feed filler rod from the side of the puddle, dip the tip into the molten metal rather than into the arc, and hold the torch at 10 to 15 degrees.

Where TIG falls down is thick plate. There is no cheap way to add heat quickly, so the arc crawls, the bead melts into a wide pool and the plate soaks. Thick plate work goes hybrid: GMAW for the fill passes, TIG only for the cap. The aerospace TIG welding guide covers the process at the demanding end of the spectrum.

Travel Speed: The Variable Everybody Ignores

Most welders set amps and volts, then move the torch at whatever speed their arm happens to travel. That is where hidden damage comes from. A slow bead burns straight through thin plate; on thick plate it soaks heat into the joint, widens and softens the heat-affected zone, and in restrained high-strength or high-carbon steel it invites hydrogen cracking. Push the speed up and heat input drops, the bead narrows and the HAZ tightens; go too fast and you get poor fusion, undercut at the edges and a bead that barely fuses into the groove.

Aluminium: Heat Sink and Interpass Control

Aluminium pulls heat from the puddle far faster than steel, so the puddle shrinks ahead of the torch and the weld comes out porous and cold. A heat sink, usually a steel or copper backing bar, slows that loss and holds the profile. Interpass temperature needs watching too, since excess heat in the HAZ reduces strength in aluminium alloys. The complete aluminium MIG guide covers settings, cleaning and gas in more depth.

Get the machine dialled in on a scrap of the real material, write the numbers down and adjust from there. Consistent settings beat theoretical ones every time.

Welding Different Metals on Plate: Steel, Stainless, Aluminium and Coated Stock

The biggest trap with plate work is assuming one set of parameters fits every metal. Switching from mild steel to stainless, aluminium or coated stock means reworking your heat input, filler, shielding gas and prep work to match its metallurgy.

Carbon steel: the forgiving baseline

For most plate projects, carbon steel is the place to start. For MIG, ER70S-6 wire in the 0.030 to 0.035 in range is a reliable general-purpose option, whilst for stick, E6013 is great for clean plate and E6011 is better for dirty, rusty or outdoor work. If you’re still getting to grips with rod classifications, the all-position welding rod guide is a solid breakdown of the numbering system.

Gas choice is just as important. For MIG, 75/25 argon-CO2 is the most common mix, with argon stabilising the arc and CO2 boosting penetration and wetting. For flux-cored work, FCAW is self-shielded with 100 percent CO2 or flux-cored wire, removing the need for a gas bottle entirely. Depending on your needs, suppliers such as AllGas Welding Supply can blend or fill the correct mix for the job.

Stainless steel: heat and contamination control

Austenitic grades like 304 and 316 demand a more delicate approach. For TIG, 100 percent argon or an argon/helium mix is used, whilst MIG on stainless typically calls for a tri-mix shielding gas of argon, helium and a small percentage of CO2. The helium addition helps widen the arc and improve penetration on thicker sections.

Heat input is the most critical lever. Stainless transfers heat far more willingly than carbon steel, so most procedures keep interpass temperatures below 350 °F (175 °C) for austenitics to avoid sensitization, grain growth and distortion. Alongside this, contamination must be prevented at all costs. Any contact with a carbon steel brush, grinding wheel or bench can seed free iron into the weld, causing it to rust prematurely. Dedicated “inox” consumables and brushes are a must, and filler must match the base metal, such as E308L or ER308L for 304, and 316L for 316, as specified under the AWS codes and standards.

Aluminium: oxide, heat and dedicated hardware

Aluminium is a completely different beast. For both MIG and TIG, 100 percent argon is mandatory. AC TIG is the benchmark for breaking down its stubborn oxide film, whilst for production work, a spoolgun is often used as outlined in the aluminum MIG welding guide to prevent wire feeding issues.

Hardware cannot be shared. Aluminium wire is incompatible with steel contact tips and steel liners, so dedicated aluminium contact tips and a dedicated aluminium gun liner are required. Preparation is just as strict: mechanically abrade, chemically clean and wipe with a clean solvent. The aluminum cleaner preparation guide breaks down these steps in more detail.

The largest hurdle is its heat sink effect. Aluminium’s thermal conductivity is roughly five times that of steel, so heat races away from the puddle. Standard practice is to use a back-to-back plate as a heat sink, stick to a controlled interpass temperature and maintain a controlled travel speed. Flux-based options are also covered within aluminum flux welding and aluminum flux core welding, however these are far less common than gas-shielded methods.

Coated and galvanized plate

Welding galvanized plate introduces a fume hazard. Vaporising the zinc coating creates zinc oxide fume, which can cause metal fume fever. For this reason, ventilation or local fume extraction is non-negotiable, in line with the guidance found in OSHA 1910.252.

Grinding the galvanizing back to bright metal before welding is the most effective mitigation. Where this isn’t possible, strict extraction and consumable selection become vital. For stick welding over these areas, low-hydrogen electrodes such as E7018 are the usual recommendation. More specific fume control details can be found within the OSHA welding fact sheet.

Special cases: cast iron and dissimilar metals

Not every plate weld is a match between two identical pieces. Cast iron is highly brittle due to its high carbon content, requiring preheat, nickel-iron consumables and a slow, controlled cooling rate to avoid cracking. Dissimilar joints, such as stainless to carbon steel or aluminium to steel, often require buttering, transition pieces or a fully qualified procedure to control dilution between the two metals. Both of these are specialist jobs. For alloy filler considerations, the alloy steel welding electrode guide is a helpful starting point, but for these scenarios, a qualified procedure should always take priority over guesswork.

Quick comparison

Material	MIG	TIG	Stick	Key gotcha
Carbon steel	ER70S-6, 75/25 Ar-CO2	ER70S-2 as needed	E6013 / E6011 / E7018	Rust and mill scale cause porosity and undercut
Stainless	Tri-mix, ER308L/316L	100% argon or argon/helium	E308L-16	Low heat input required and avoid iron contamination
Aluminium	100% argon, spoolgun, dedicated tips	AC, 100% argon	Not recommended	Oxide removal and a heat sink are mandatory
Galvanized / coated	Grind back coating first	Grind back coating first	E7018, fume extraction	Zinc fume hazard and poor surface adhesion
Cast iron	Specialist only	Specialist only	Nickel-iron rods, preheat	Cracking from high carbon and fast cooling

Distortion and Warping: Why Plate Moves and How to Fight It

Weld a 6 mm fillet along the edge of a 3 mm plate and it curls like a leaf: the bead looked perfect, yet the part came off the bench bent with its machined dimension gone.

The cause is physics. A weld heats a narrow strip until it expands, then wants to shrink back on cooling, but it is tied to cold, rigid steel on both sides. The contraction wins and the plate bends. That is why plate welding has a distortion story of its own.

Why Plate Distorts Differently

Tube’s closed round shape is self-reinforcing. Flat plate is not: it sheds heat in two directions and can bow, twist or buckle toward its weakest axis. TWI identifies six main forms of distortion in welded joints:

• Longitudinal shrinkage — the weld shortens along its length
• Transverse shrinkage — the joint draws across the weld
• Angular distortion — the plates rotate relative to each other
• Bowing and dishing — the plate curves lengthwise or across width
• Buckling — thin plate wrinkles or ripples
• Twisting — the joint spirals out of plane

Angular distortion ruins flatness fastest, because contraction through the plate thickness is not uniform: the top near the weld shrinks before the bottom. In a single-V butt weld the first run shrinks the plates and they rotate; the second uses that deposit as a fulcrum and rotates them further. Every extra pass adds an increment.

T-joints behave the same: a single fillet pulls the upstanding leg inward, and a matching fillet on the far side cancels much of that angular distortion, but the movement does not vanish. It moves into the base plate, which now bows instead.

Pass Size and Pass Count Matter More Than Most People Expect

Here is the counterintuitive result from TWI’s distortion research: completing a joint with a few large deposits produces more longitudinal and transverse shrinkage than many small passes. One heavy pass dumps enough heat to buckle the plate; dozens of small passes spread the heat and let the plate relax between them. Multi-pass welding also cuts angular distortion per pass, because earlier deposits restrain the later ones.

Multi-layer, multi-pass research measured distortion of roughly 0.5 to 2 mm in some initial passes, with the root pass showing the greatest shrinkage because it has no restraint at all and later passes held back by metal already deposited around them. It also found that angular distortion grows roughly linearly with the number of passes at a given thickness, another argument for fewer, well-planned passes. The detail sits in the Materials journal literature and TWI’s arc processes and welding engineering work.

The Prevention Hierarchy

The Australian Welding Institute’s distortion guidance offers a hierarchy worth memorising, because it moves you from fighting distortion to preventing it:

1. Eliminate welding by forming. Formed parts do not shrink.
2. Minimise weld metal. Prefer a double-V groove to a single-V.
3. Do not over-weld. A 6 mm fillet where 4 mm would do is extra distortion, not strength.
4. Prefer intermittent welding. Spaced beads let the plate relax.
5. Place welds near the neutral axis. Heat there bends less.
6. Balance welding about the joint centre line.

ESAB echoes the double-V logic: because metal goes on both sides, a double-V groove needs roughly half the weld metal of a single-V groove, meaning less heat, less shrinkage and less rotation. ESAB also prefers fewer beads over many small ones, which sits against the multi-pass research, so tune pass size and count rather than minimising either blindly.

Welding Sequence and Fixturing

Even a good joint moves according to bead order.

• Weld from the centre outward. The ends anchor the joint; the middle moves most.
• Alternate sides wherever the design allows.
• Use the backstep technique: weld a short length, step back 25 to 30 mm, weld forward again.
• Clamp and fixture deliberately. Restraint fights shrinkage and can drive cracking, so aim for balance.
• Build back-to-back. Two plates face to face cancel most of the movement.

The Honest Trade-Off

Every control measure has a price: double-V grooves cost more to prepare, intermittent welding takes longer, fixtures cost money and floor space, balanced sequencing slows the job. Consumable suppliers argue the same way: a controlled part that meets tolerance beats a fast one that gets straightened, re-machined or scrapped. Start with joint design, then sequence, and save fixturing and intermittent welding for parts that truly need flatness.

What you see	Likely cause	Next time
Plates out of flat	Angular distortion	Double-V, or fillets both sides
Bowed lengthwise	Longitudinal shrinkage	Intermittent welding, backstep
Bowed across the width	Transverse shrinkage	Centre outward, alternate sides
Wrinkles in thin plate	Buckling under compression	Lower heat input, add backing bars
Twists out of plane	Asymmetric shrinkage	Symmetric layout, balanced passes

Distortion is the price of putting heat into a flat object and expecting it to stay flat. Budget for it: design the joint so heat enters symmetrically, keep the weld metal lean, and plan the sequence before you strike arc.

Safety: Fumes, Ventilation, Fire and the Rules That Are Not Optional

Plate welding brings hazards together all at once: extreme heat, a constant fume plume, and a large fire load from offcuts and cylinders. For jobs this size, the rules aren’t suggestions. They’re what keep you out of the hospital.

What Welding Fume Actually Is

Welding fume isn’t campfire smoke. It’s a fine metal aerosol, mostly made up of particles under one micron, mixed with gases such as carbon monoxide, nitrogen oxides and, depending on the flux, ozone or hydrogen fluoride. These particles reach deep into the lungs and stay there. As NIOSH’s welding research notes, the base metal and consumable dictate what goes into that fume, which is why stainless and galvanised work carry a very different hazard profile from mild steel.

Acute symptoms can include burning eyes, a raw nose and throat, coughing, chest tightness, headache, dizziness or nausea. Don’t push through them. Stop welding, get to fresh air, and seek medical attention if they persist.

Ventilation: Capture Before It Reaches You

General ventilation can dilute fumes, but for plate work, local exhaust ventilation is far more effective. A hood or extraction arm placed near the arc captures the plume at its source before it ever reaches your breathing zone.

Two points can’t be overlooked. First, welding outdoors is no guarantee of safety. Wind can blow fume directly into your face, and still pockets can form between walls or equipment. Second, you cannot rely on a mask alone to weld in a confined space. Hot work there requires a permit, gas monitoring, a rescue plan, and adequate ventilation before striking an arc.

Fire Prevention and the 35-Foot Rule

Fires from plate work don’t always happen at the moment of arcing. Slag and spatter can smoulder long after the torch is off. OSHA 29 CFR 1910.252 requires that, where practicable, all combustibles be relocated at least 35 feet (10.7 m) away from the work site. If they can’t be moved, they must be protected, and a dedicated fire watch must remain on duty for at least half an hour after hot work ends.

Extra caution is needed for used containers. Drums, barrels, or tanks must be thoroughly cleaned, with all hollow spaces vented and, where advised, purged with inert gas before any hot work begins. Even a small amount of residual solvent can flash over an arc with devastating results.

• Assign a dedicated watch: No other duties, and someone who knows how to use the correct extinguisher.
• Inspect twice: Check the area before striking the arc, and again at the end of the half-hour watch. Never use water on a magnesium or sodium fire.

Eye Protection and Filter Shades

The arc’s ultraviolet and infrared radiation is more than blinding in the moment. It can cause arc eye (photokeratitis), often showing up hours later. For every plate job, you’ll need a proper helmet or goggles with side shields, safety glasses underneath, flame-resistant clothing, and gloves. The correct filter shade is just as vital.

Based off of OSHA 29 CFR 1910.252 and its welding safety fact sheet, use the following as a minimum:

Process or electrode size	Minimum filter shade
SMAW (stick), 1/16 to 5/32 in electrodes	10
Gas-shielded arc welding of ferrous material (MIG/TIG)	12
SMAW (stick), 3/16 to 1/4 in electrodes	12
SMAW (stick), 5/16 to 3/8 in electrodes	14

Auto-darkening helmets are ideal, but only if they’re set for the process you’re running, not left at a generic setting.

Hexavalent Chromium, Manganese, and Long-Term Risks

Some fumes become far more hazardous over time. Welding stainless or chrome-plated materials can generate hexavalent chromium, a known carcinogen. For this reason, OSHA sets its permissible exposure limit at 5 µg/m³ as an 8-hour time-weighted average, making local exhaust non-negotiable for this work. For more detail, OSHA’s hexavalent chromium guidance outlines the long-term disease risks, which can appear decades after heavy exposure.

Manganese is a more common everyday concern. NIOSH recommends a limit of 1 mg/m³ as an 8-hour TWA with a 3 mg/m³ short-term limit, while OSHA’s PEL is a 5 mg/m³ ceiling and ACGIH’s TLV is 0.02 mg/m³ for the respirable fraction and 0.1 mg/m³ for the inhalable fraction. Prolonged exposure above roughly 1 mg/m³ can lead to manganism, a Parkinsonian syndrome. These exposures are drastically increased in confined spaces, where there is nowhere for the fume to go.

Cylinder Safety Under 1910.253

Gas cylinders demand just as much respect as the arc itself. A damaged cylinder is a projectile, and a small leak can become a fire in seconds. The requirements set by 29 CFR 1910.253 are strict for a reason:

• Acetylene: Never use above 15 psig. Over this limit, acetylene can decompose and self-ignite inside the cylinder.
• Storage: Oxygen cylinders must be kept at least 20 feet (6.1 m) from fuel-gas cylinders or combustibles, or separated by a suitable barrier.
• Positioning: Fuel-gas cylinders must always be stored valve-end-up.

For torch work, the Air Acetylene Welding guide is useful for regulators and hoses, whilst the welding gas supply guide covers cylinder types and fittings. No matter the job, perform a soapy-water leak check on every new cylinder before bringing it near an open flame.

Common Plate Welding Mistakes, Defects and How to Troubleshoot Them

Most plate weld defects announce themselves: a bead that dips at the toe, a root that glows orange before you finish the pass. Trouble starts when a defect looks acceptable, gets covered, and is discovered later as a crack. Reading defects as they form is the difference between a repairable mistake and a scrapped assembly.

Machine-vision research on multi-layer multi-pass welding confirms that lack of fusion, porosity and burn-through remain the dominant defect families in thick-section plate welding, generated in testing by deliberately varying welding parameters (machine-vision study of weld defect detection). Those three, plus undercut, overlap and cracking, cover nearly everything you will meet.

Undercut: A Groove Along the Weld Toe

Undercut is a groove melted into the base metal at the toe of the bead. It is driven by excessive travel angle, meaning you drag the puddle off the groove edge; amperage too high for the joint, which melts the edge faster than filler can fill it; and inadequate surface preparation. Mill scale is a specific contributor, because its oxide layer melts at a different temperature than the steel beneath it, so the edge melts unevenly and the bead cannot bridge cleanly.

Lower the amperage slightly, flatten the travel angle, increase stickout so the arc is less concentrated, and grind the groove face down to bright metal. Undercut is also a fatigue crack initiation site, which is why structural codes treat it as more than cosmetic.

Porosity: Holes in the Weld Metal

Porosity shows up as round or elongated voids, either on the surface or revealed when you break a bead. Causes cluster around cleanliness and shielding. Contaminated base metal, paint, oil, rust and moisture all release gas as they vaporise, and moisture in the shielding gas supply or a damp gas line does the same thing. Low or incorrect flow weakens the protective envelope, and on outdoor plate work wind deflects it entirely.

Set flow for the nozzle size and the process, check the hose and fittings for leaks, and use a windscreen in any breeze. On coated or dirty plate, grind back to clean metal before welding to remove the gas source.

Lack of Fusion and Incomplete Penetration

These two are related but not identical. Incomplete penetration means the weld did not reach the bottom of the groove, so the joint is not fully joined. Lack of fusion means the weld metal did not melt and merge with the base metal, typically at the groove walls.

Shared root causes: insufficient groove angle, amperage too low for the thickness, travel speed too high, and an unclean or oxidised root face. A 45-degree bevel that was never actually cut, a bead moved too quickly to wet the sidewall, and a root pass laid over mill scale all produce the same result. Gouge back to sound metal and re-weld with more heat, slower travel, a clean surface and, if needed, a wider bevel.

Overlap, Burn-Through and Spatter

Overlap happens when travel speed is too slow for the amperage. The molten puddle has nowhere to go, so it rolls over the previous bead and leaves a lumpy step. Beyond looking untidy, the overlap is an unfused, irregular region and a genuine crack initiation site. Raise travel speed, lower amperage, or reduce bead size.

Burn-through is the opposite error on thin plate: too much heat, and the puddle drops through the joint before you can control it. On 3 mm and 4 mm material, add a backing bar or copper strip, back-step to let the root cool, and cut amperage and wire feed together.

Spatter comes from wrong polarity, wrong amperage for the wire diameter, dirty or coated plate, and wind. In MIG welding it usually signals a DCEP setting when the machine and wire expect another polarity, or an arc length that is too long. Clean the plate, check polarity against the wire spool, and re-establish a tight arc length.

Cracks: Two Very Different Mechanisms

Hot cracking happens while the weld is still hot. It is driven by high restraint combined with susceptible material chemistry, particularly elevated sulphur and phosphorus, which form low-melting films between grains that tear as the joint contracts. The crack opens at or beside the weld toe, often in the first pass, and it is a solidification problem.

Cold, or hydrogen-assisted, cracking appears hours or days later. It shows up in hardenable and high-carbon steels, initiates at a stress raiser such as an undercut, an overlap, an arc strike or a notch, and propagates into the heat-affected zone. That is the reason low-hydrogen consumables and preheat exist.

Out-of-plane distortion plays a role too. As the joint pulls during cooling, bending the plate out of plane concentrates tensile stress at the root, and a micro-notch becomes a long crack. A crack running across the plate into the heat-affected zone rather than along the weld is often the signature of this combination.

Dimensional Error

Not a metallurgical defect, but the one that most often ruins the job. It traces to poor fit-up, too much deposited weld metal, or a poor welding sequence. Once the plate is out of flat, it cannot be pulled back without adding more heat and more distortion. Measure twice, clamp properly, and plan the sequence before the first arc strikes.

Quick Diagnostic Table

Defect	What you see	Likely cause	First fix
Undercut	Groove at the weld toe	Steep travel angle, high amps, mill scale	Flatten angle, lower amps, grind clean
Porosity	Round holes in the bead	Dirt, moisture, low flow, wind	Clean steel, check flow, screen wind
Incomplete penetration	Root not joined through	Narrow bevel, low amps, fast travel	Gouge out, widen groove, add heat
Lack of fusion	Sidewall not wetted	Dirty root face, too little heat	Grind bright metal, slow travel
Overlap	Lumpy step over the bead	Travel too slow for the amps	Raise travel speed or lower amps
Burn-through	Hole under the puddle	Too much heat on thin plate	Backing bar, lower amps, back-step
Spatter	Balls of metal around the bead	Wrong polarity, amps, dirt, wind	Check polarity, clean plate, shield arc
Hot crack	Open crack at the toe	Restraint, sulphur, phosphorus	Cut heat input, reduce restraint
Cold crack	Crack appearing hours later	Hydrogen plus a stress raiser	Remove defect, low-hydrogen filler, preheat
Dimensional error	Plate bowed or out of square	Bad fit-up, excess metal, poor sequence	Refit, re-weld with a planned sequence

Why You Must Remove a Defect, Not Cover It

This is the rule that gets broken most often, usually because a second pass is faster than a gouge. AWS codes require that defective weld metal be removed before new weld metal is deposited (AWS codes and standards). A crack covered by a new pass is still a crack, and a porous region buried under fresh weld metal stays an invisible stress raiser and a potential corrosion initiation site.

One habit ties this together: keep a scrap coupon of the same thickness beside the work and run a test bead whenever you change wire, gas, position or machine setting. A four-minute test weld shows the problem long before you have burned through a real assembly. If you are working with aluminium or coated stock rather than plain carbon steel, the prep and parameter rules shift again, so read the guides on aluminium cleaner for welding and aluminium MIG welding before setting parameters there.

Inspection, Testing, Codes and Planning Your First Plate Project

A weld is only as good as the evidence that it is good. Judgment starts when the slag comes off, and most plate welding inspection is visual because most plate welding defects are visible to trained eyes. Codes exist because a weld can look clean and still be wrong.

Visual inspection: the primary method

Visual inspection (VT) needs nothing beyond trained eyes, a straightedge and a weld gauge. The attributes checked are defined in AWS A3.0, Standard Terms and Definitions for Welding and Allied Processes; acceptance criteria come from the governing code.

• Profile and reinforcement: bead face shape and its blend into the plate.
• Size and length: fillet leg, groove throat, reinforcement height, effective length.
• Undercut: groove melted into base metal at the toe, a stress raiser.
• Porosity: gas pockets, round voids or dark spots in the bead.
• Cracks: any linear discontinuity, cold or hot.
• Adjacent areas: arc burns, spatter, undercut at weld ends, poor tie-in.

The five standard nondestructive testing methods

When visual inspection is not enough, five standard NDT methods remain, each with its own trade-off in access, cost and sensitivity.

Method	Finds	Needs	Best for
Visual (VT)	Profile, size, undercut, porosity, cracks	Trained eye, gauges	Every weld, first
Magnetic particle (MT)	Surface and near-surface flaws	Ferromagnetic material, magnetizer	Carbon and low-alloy steel
Dye penetrant (PT)	Fine surface-breaking defects	Clean, non-porous surface	Stainless, aluminium
Radiographic (RT)	Internal volumetric flaws	Radiation controls, two-side access	Pressure and critical joints
Ultrasonic (UT)	Flaws deeper in section	One-side access, skilled operator	Thick, volume-critical welds

Magnetic particle testing shows a magnetic field jumping a surface or near-surface flaw in ferromagnetic material. Fast and sensitive, it fails on austenitic stainless, aluminium and dirty surfaces.

Dye penetrant testing uses capillary action: penetrant seeps into a flaw, excess is wiped off, a developer draws it back out. It cannot find flaws buried in the metal.

Radiographic testing images density variations on film or a detector, finding porosity, slag and lack of fusion. It needs radiation safety controls and a qualified reader.

Ultrasonic testing reads echoes from sound sent into the weld. The most sensitive method for internal flaws in thick sections, it needs one-side access and a skilled operator.

AWS D1.1 and the rest of the code landscape

AWS D1.1, the Structural Welding Code, is the most widely referenced structural welding code in the world. It sets requirements for welder qualification, qualified welding procedures, inspection and structural acceptance, and its examination provisions decide which NDT method a joint needs. Method names come from AWS A3.0, so use the standard’s terminology exactly. AWS publishes more than 350 standards on welding practices and procedures.

Two ideas carry the weight. A qualified welding procedure specification (WPS) lists, in ranges, the variables that produce an acceptable weld under a code, and welder qualification proves a welder can meet that code with a given process in a stated range. Those documents turn a good-looking weld into a code-compliant one, and they are how you get insurance, contracts and approvals.

Material-specific work brings other codes: AWS D1.6 for stainless, AWS D1.2 for aluminium, and the ASME codes for pressure and boiler work. If you are unsure which governs, ask before you start.

Planning your first plate project

Here is an ordered plan that works. It is deliberately boring, and that is the point.

1. Match joint to process: square butt on thin sheet with MIG, single-V on 10 mm plate with stick or FCAW, TIG fillet on stainless.
2. Write parameters down before you light the arc: amps, volts, wire feed, travel speed, gas, filler, stickout.
3. Test coupons in the real material. Aluminium needs a dedicated setup: aluminium MIG welding differs in gun setup and consumables, and preparing aluminium correctly comes first.
4. Tack small, clean and staggered, cleaning each tack before welding over it.
5. Control interpass temperature by watching the plate, not the clock. On thick steel it separates a sound joint from a hydrogen-cracked one.
6. Clean between every pass and remove slag and spatter fully.
7. Plan the sequence first: centre outward, alternate sides, backstep technique.
8. Inspect after each pass. Groove geometry and distortion shift as the weld builds up, so track the seam line and profile pass by pass, which multi-pass research also supports.
9. Repeat the identical procedure on a second coupon. Two good welds in a row means you have a procedure; one means you got lucky.

The eight decisions that matter most

Eight choices account for nearly every plate welding problem worth having: the process that matches material and thickness, joint design, base material grade and condition, surface preparation, heat input in kJ/mm, interpass temperature and preheat, sequence and restraint, and inspection with documentation. Get them right and the weld will usually be sound and repeatable. Get one wrong and nothing else will rescue it.

Where to go next

Do I need to be certified? For garage work, no. For structural, pressure-related or commercial work, yes: the code and the customer set the qualification.

Same settings for stainless or aluminium? No. Different filler, gas, often polarity and a different contamination discipline. The aluminium welding course overview covers the cost of learning it properly.

Which NDT method? Start with visual, then magnetic particle for carbon steel surface flaws, dye penetrant for stainless and aluminium, ultrasonic for internal flaws in thick sections, radiographic where the code demands internal flaw detection.

Before your next plate job, read the procedure you are about to follow, fix the one of the eight decisions you are still guessing at, then see why stick welding still wins in so many situations.

Frequently Asked Questions

What is the best welding process for metal plate?

For most general shop work on 3 mm to 12 mm plate, MIG (GMAW) is the best all-round choice because it combines good penetration, high deposition rates and a short learning curve. Stick (SMAW) is the better answer for heavy plate over about 12 mm, outdoor or field work, and joints that are dirty or mill-scaled. TIG (GTAW) wins for thin sheet, stainless steel, aluminium and any joint where appearance and precise heat control matter, but it is impractical on thick plate.

How thick can I weld with a MIG welder?

Short-circuit MIG is commonly limited to around 6.4 mm (0.25 in) for structural work. Beyond that, spray transfer, pulsed MIG or flux core are the practical options, and thick sections normally need beveling, a root gap and multiple passes rather than one big bead. A machine limited to 200 amps of MIG output will run out of heat well before you run out of plate thickness.

Do I need to bevel the edges of plate before welding?

On thin plate, square butt joints work fine. Once the plate is thick enough that a single pass cannot reach the bottom of the joint, you need a groove — usually a single-V at 45 to 60 degrees or a U or J preparation. A double-V groove needs roughly half the weld metal of a single-V, which cuts both heat input and distortion, and a small root opening of about 1/8 in (3 mm) on half-inch plate helps the root pass reach the bottom.

What causes cracks in welded plate?

Two different mechanisms are usually responsible. Hot cracking happens while the weld is still hot and is driven by high restraint plus susceptible material chemistry. Cold or hydrogen-assisted cracking shows up hours or days later in hardenable and high-carbon steels, and is prevented by controlling hydrogen, preheating and limiting restraint. Cracked weld metal must be removed and re-welded, never covered over with a new pass.

How do I stop my plate from warping while I weld it?

Distortion is driven by weld shrinkage, so the answer is less weld metal, better balance and better restraint. Weld from the centre outward, alternate sides between passes, use a double-V in preference to a single-V, avoid oversized fillet welds, and clamp the work or weld two plates back-to-back to hold the flat. Accept the trade-off: these measures add setup time and material cost, but they beat trying to straighten a warped weldment afterwards.

Is it safe to weld in a garage or small workshop?

It is safe if you ventilate properly, protect against fire and control the fume. OSHA requires local exhaust or general ventilation to keep fume below the permissible exposure limits, and warns that welding outdoors does not by itself guarantee adequate ventilation. Keep combustibles 35 feet (10.7 m) away where practicable, use a fire watch for at least half an hour after finishing, and never weld in a confined space without ventilation.

What is the correct welding helmet shade for plate work?

OSHA’s guidance table recommends shade 10 for stick welding with 1/16 to 5/32 in electrodes, shade 12 for gas-shielded arc welding of ferrous material, shade 12 for SMAW at 3/16 to 1/4 in and shade 14 at 5/16 to 3/8 in. Start at the higher (darker) shade and move lighter if you cannot see the puddle clearly, since an over-dark filter hides defects you need to spot.

Should I use MIG or stick for outdoor metal plate work?

Stick, or self-shielded flux core, is usually the better outdoor answer. Shielded MIG depends on a clean, still environment because any breeze deflects the shielding gas away from the molten puddle and you will get porosity. Stick electrodes and self-shielded flux core wire carry their own protection and handle dirty, rusted or mill-scaled plate better, though both produce more spatter to clean up.

How do I know if my plate weld is good without cutting it open?

Start with visual inspection, which is the standard first check on any procedure: look at profile, size, length, and any undercut, porosity or cracks. If the weld is critical, non-destructive testing adds confidence — magnetic particle testing for surface and near-surface flaws in carbon steel, dye penetrant for clean non-porous surfaces, and ultrasonic or radiographic testing for internal flaws in thicker sections. For structural work, AWS D1.1 governs who is allowed to inspect and accept the result.

What is the single biggest beginner mistake when welding plate?

Failing to clean and prepare the joint before welding. Mill scale, rust, paint, oil and moisture all cause undercut, porosity and lack of fusion, and no amount of parameter adjustment will fix a contaminated surface. Grind or wire-wheel the joint back to bright metal, remove all moisture, and clean between every pass.