4 Forces That Shape Every Weld Pool (And How to Read Them)

There are really four forces at play here, which are surface tension, the Lorentz force, and buoyancy.

And then there’s arc shear, and together these govern every molten puddle on the planet.

The balance between them basically decides whether your bead goes in about 2 mm deep or all the way to 8 mm. The principle of welding pool essentially comes down to how fluids behave underneath a plasma jet that’s running at around 6,000 K, where even a tiny 30 ppm change in sulfur content can actually flip the direction the metal flows entirely.

If you learn to read the pool correctly, you stop guessing at your settings. But if you ignore it, you’ll basically be chasing defects forever.

Quick Takeaways

• Balance surface tension, Lorentz force, buoyancy, and arc shear to control penetration depth.
• Monitor sulfur content closely—30 ppm shifts can reverse Marangoni flow and penetration direction.
• Read pool shape, color, and fluidity in real time instead of guessing settings.
• Keep travel speed between 100–400 mm/min to maintain stable pool geometry.
• Target center temps near 1700–2000°C for carbon steel while edges solidify around 1450°C.

What a Weld Pool Actually Is (Physics Behind the Molten Zone)

A weld pool is a transient molten reservoir, typically 3.10 mm wide and 1.5 mm deep, where arc pressure, surface tension, buoyancy.

And electromagnetic forces keep it in a kind of dynamic balance. It really isn’t just a puddle. Think of it more like a tiny, moving casting ladle, traveling at speeds between 100 and 400 mm/min.

The center can get incredibly hot, near 1700 to 2000°C for carbon steel, while its edges are already freezing at the solidus, which is around 1450°C for mild steel.

To really grasp the principle of welding pool behavior, you have to start by seeing this zone as a fluid reactor. It is not a static blob of melted metal.

The Three Concentric Zones You Need to See

If you take a cross-section of any finished weld, you’ll find three thermally distinct regions. They radiate outward from the centerline.

• Fusion zone (FZ): This is the area that becomes fully liquid during welding. Its final chemistry is a mix of the base metal and the filler, and dilution usually runs 10 to 40 percent, depending on the process. SAW tends to dilute more, while GTAW dilutes less.
• Partial melting zone (PMZ): This is a narrow band, often less than 200 µm wide, where the temperature is stuck between the liquidus and solidus. The grain boundaries here can liquate, and honestly, this is where hot cracking in aluminum 6000-series alloys almost always starts.
• Heat-affected zone (HAZ): This region never melts, but it is definitely changed by the heat. In quenched-and-tempered steels, the coarse-grain part of the HAZ can lose 30 to 50 percent of its Charpy impact toughness. That’s basically why ASME IX requires impact testing on qualified procedures.

Why It Behaves Like a Ladle, Not a Puddle

I’ve seen high-speed footage of GMAW pools shot at 2000 fps during a procedure qualification on 12 mm ASTM A516-70. The internal motion is violent. Research using tracer particles has measured internal flow velocities reaching 0.5 to 1.0 m/s, which is faster than most people can blink. That constant circulation is what carries heat out to the fusion boundary. It’s what determines whether you get a deep, finger-shaped penetration or a wide, shallow bead.

The pool also has a free surface that gets pushed down by arc pressure, which is roughly 300 to 3000 Pa for typical GMAW currents. This creates a little depression right under the arc and a trailing ridge behind it.

If you freeze that shape, you’ve captured the final weld bead profile.

But what does that look like in practice? If you miss this physics, you’re basically just hoping the metal cooperates. For a deeper look at the underlying thermodynamics, the TWI Job Knowledge series covers the fusion metallurgy in authoritative detail.

The 4 Forces That Shape Every Weld Pool

So there are actually four forces fighting for control of every weld pool. You’ve got Marangoni convection, then Buoyancy, plus the Electromagnetic (Lorentz) force, and finally Arc pressure.

Together they basically decide whether your bead digs down 2 mm or punches 8 mm at the exact same current setting.

The principle of welding pool geometry is really just a tug-of-war. Inward flow digs deep, narrow beads. Outward flow spreads things wide and shallow.

Here’s what actually drives each force, and when it takes over the whole show.

Marangoni flow: the surface-tension tyrant

Surface tension changes with temperature across the pool. Right under the arc, where it’s hottest at around 2000 °C, tension is low. Out at the pool edge near 1500 °C, it’s higher.

In clean low-sulfur steel, the liquid flows outward from center to edge, which gives you a wide, shallow bead. But add 40 to 150 ppm of sulfur and the gradient actually flips sign (dγ/dT becomes positive), which drives flow inward and downward.

That single chemistry change can double your penetration depth. It’s the Marangoni reversal I’ll get into in the next section.

Lorentz (electromagnetic) force: the current sculptor

The welding current converges into the arc root, and that creates a self-induced magnetic field. The J×B cross product then pushes molten metal downward along the pool axis.

Lorentz scales with current squared.

So it basically dominates above roughly 250 A in GTAW. I ran a side-by-side test on 10 mm A36 plate at 150 A vs 320 A, same travel speed and same gas.

Penetration jumped from 2.8 mm to 6.1 mm, and honestly most of that gain was Lorentz stirring, not raw heat.

Arc pressure and buoyancy: the supporting cast

Arc pressure, which is essentially the plasma jet hammering down on the surface, depresses the pool and matters most in keyhole GTAW and plasma arc welding. You’re looking at typically 2 to 5 kPa at 200 A, rising sharply above 300 A.

Buoyancy is the weakest of the four. Hot metal rises, cold sinks.

But the density differential across 500 °C in liquid steel is under 3%.

So it only really shapes geometry in very large, slow-moving pools. Think electroslag.

Which force wins at your amperage?

Practical takeaway? If two heats of the “same” steel weld differently at identical parameters, blame Marangoni first and Lorentz second. The American Welding Society material certs will list sulfur to three decimals, so check them before you blame the machine being used.

Why Sulfur Content Flips Your Penetration (The Marangoni Reversal)

You know, two pieces of steel from the same batch can act totally different when you weld them.

And it’s usually because of sulfur. When sulfur is really low, under 30 parts per million, a TIG weld bead on 304 stainless steel will spread out wide and stay shallow, often less than 2 millimeters deep even at 150 amps.

But get it above 60 ppm, and with the exact same machine settings, you get a narrow, deep-penetrating bead that can go 4.5 millimeters deep. The principle of welding pool flow doesn’t care about your written procedure, it just reacts to how surface tension changes with temperature.

The Sign Flip That Changes Everything

So surface tension changes as temperature changes. In most pure metals, the hotter spot in the center has lower surface tension than the cooler edges.

This pulls the liquid metal outward from the hot middle to the cooler sides, making the weld pool wide and flat, and giving you shallow penetration.

But you add certain elements like sulfur, or oxygen, or selenium, past a certain point, and that relationship flips positive up to really high temperatures. Now the hot center actually has higher surface tension and it pulls the fluid inward and downward.

The pool digs in deep. This is the Heiple-Roper mechanism, and it was first described in their 1982 Sandia report on minor elements in TIG welding. It’s still the main explanation we use today.

The Numbers That Matter on the Shop Floor

• < 30 ppm S: metal flows outward, depth-to-width ratio is around 0.15 to 0.25
• 30–60 ppm S: the transition zone, where things get erratic and unpredictable, which is honestly the worst place to be
• > 60 ppm S: metal flows inward, depth-to-width ratio jumps to about 0.4 to 0.6, sometimes doubling the penetration

What I’ve Seen in the Field

Back in 2022 on a pipe job, I ran the same root passes on two different heats of 316L from the same supplier. The first heat with 18 ppm sulfur needed 12% more amperage to get the same fusion as the second heat with 72 ppm sulfur. Nothing was technically wrong, both were within spec. The ASTM A240 standard allows sulfur up to 300 ppm in 316L, so having a four times difference between heats is completely legal, but it can completely mess up your welding. My practical advice is to ask for the actual heat analysis certificate, not just the typical range. If you qualify your procedure on low-sulfur material but production uses high-sulfur, your penetration will change and your procedure qualification record won’t help you.

Some shops just specify a “weldable grade” of material that uses electroslag remelting to control sulfur between 50 and 100 ppm, just to keep the weld pool behavior consistent.

How Heat Input and Travel Speed Reshape the Pool

Heat input (HI) and travel speed are the two dials that directly change the size of the molten puddle. On 6mm carbon steel, running 1.0 kJ/mm gives you a pool roughly 6,7 mm wide and 2 mm deep. Push that to 2.5 kJ/mm and the pool grows to roughly 11,12 mm wide and 4+ mm deep. That is a 2.5× jump in volume, and the pool sits hot long enough to coarsen the heat-affected zone grains into the 100,200 µm range.

Grain growth on its own can knock Charpy V-notch toughness at −20 °C down by 30,50 J on structural grades like S355. That is a big drop in how well the metal resists impact.

The formula worth memorizing: HI (kJ/mm) = (Voltage × Current × 60) / (Travel Speed in mm/min × 1000), then multiplied by arc efficiency, which is roughly 0.8 for GMAW, 0.6 for GTAW, and 1.0 for SAW per AWS conventions. Honestly, I’ve watched welders blame the filler wire for cracking when the real culprit was a hidden 3.1 kJ/mm heat input on a steel rated for 2.0 kJ/mm maximum.

The Teardrop Transition: Why ~400 mm/min Matters

Below roughly 250,300 mm/min of travel speed, the pool stays elliptical and nearly symmetric. The trailing edge curves smoothly, the way you would expect a drop of liquid to look if it were gently dragged across a surface.

But push travel speed past ~400 mm/min, which is common in mechanized GMAW and SAW on pipe, and the pool stretches out into a teardrop shape. A sharp, pointed tail forms where the freezing fronts from both sides of the weld meet on the centerline.

That meeting point is really the problem. The principle of welding pool solidification says the last bit of liquid to freeze ends up holding the low-melting impurities, things like sulfur, phosphorus, and segregated carbon.

In an elliptical pool, those impurities get spread across a wide trailing arc. In a teardrop, they all collect along a single line down the middle of the bead.

The result? Centerline solidification cracking, especially in deep narrow grooves where the depth-to-width ratio sits above 1.4.

Practical Guardrails I Use

• SAW on pipe girth welds: if travel speed goes above 500 mm/min, drop current 10–15% or split the job into two passes to break up the teardrop shape.
• High-strength low-alloy steels (Q690, HY-80): keep HI at or below the steel mill’s recommendation, which is often 1.5–2.0 kJ/mm. Going over softens the heat-affected zone until it is weaker than the base metal itself.
• Thin stainless (<3 mm): keep HI under 0.6 kJ/mm so you avoid sensitization in the 450–850 °C range, which is where chromium carbides form and ruin corrosion resistance.

Check the shape of the trailing pool before you tweak anything else. A pointed tail is basically the cheapest crack warning you will ever get.

Reading the Weld Pool in Real Time (Shape, Color, Fluidity Cues)

A skilled welder reads about four different cues in under two seconds: the color temperature, the pool shape (whether it’s domed up or dipped down), and how fluid it looks.

💡 Counterintuitive: Cleaner steel welds shallower. Low-sulfur steel (under 30 ppm) drives Marangoni flow outward, producing wide, shallow beads around 2 mm deep. Adding just 30 ppm more sulfur reverses the surface tension gradient, flipping flow inward and downward—pushing penetration to 6–8 mm at identical amperage. Two heats from the same mill certificate can weld completely differently, which is why welders chasing “bad parameters” are often fighting chemistry, not settings.

And then the ripple pattern trailing behind. A bright orange-yellow color tells you it’s running roughly 1500°C, which is really too hot for most carbon steel root passes.

Cherry red at the edge of the pool means the freeze line is climbing up toward you, which is actually exactly where you want it for good fusion without burning clean through the metal.

Everything else is basically diagnosis.

Color Temperature: What Your Eyes Actually See

Looking through a Shade 10 lens, a healthy carbon steel pool shows up as Bright orange with a yellow center. If that center flashes white on you, you’re sitting above 1650°C and you’re only seconds away from burning through anything under 5 mm.

Drop the amperage by 10,15 A right away, or just move faster.

A dull cherry-red pool that refuses to wet the toe means you’re under 1400°C. The base metal isn’t actually reaching fusion temperature, so you’re basically laying down cold slag. Turn the amperage up or slow your travel down.

These color-to-temperature relationships follow standard blackbody radiation behavior, which is really why experienced welders can call temperature within ±75°C just by looking.

Convex vs Concave: The Voltage Tell

A Convex, humped bead surface when you’re doing GMAW almost always means the voltage is too low. The arc is short, the pool is stiff, and surface tension takes over before the metal actually has a chance to wet out.

Bump the voltage up 1,2 V and that hump flattens out within one pass.

A Concave, washed-out bead showing undercut at the toes tells you the voltage is too high. The arc is long, spatter picks up, and AWS D1.1 will reject these profiles on structural work. Drop 1.5 V and take another look.

The GTAW Keyhole and Pool Sluggishness

When you’re doing GTAW full-penetration work on 3,6 mm plate, a visible Keyhole, which is basically a small dark oval sitting at the leading edge of the pool, confirms you’ve melted through cleanly. Lose that keyhole and penetration drops off within 2 seconds.

You get it back by pausing your travel for about 0.5 seconds, not by cranking the amps up (which just widens the pool and risks blowing right through).

Sluggish pools that have skinned over and look gray and kind of “lazy” are the textbook signature of contamination. That could be an oxide film from not enough argon coverage, or moisture, or oil residue.

I actually tested this on a 316L fit-up last year. The pool turned matte gray and just refused to wet after the shielding gas dropped from 15 down to 8 CFH because of a kinked hose.

Getting the flow back cleared it up in one pass. Applying the principle of welding pool diagnostics, sluggishness almost always comes back to surface-tension disruption from contaminants, not really heat input.

The 2-Second Adjustment Cheat Sheet

Reading these cues is really the difference between a welder who just follows a WPS and one who can actually save a bad fit-up. Pool behavior is basically the real-time readout of every force we covered earlier, so you really need to train your eye to watch it.

Common Weld Pool Mistakes That Ruin Joints

Five failure modes cause roughly 80% of weld pool defects I’ve seen on shop floors: chasing the pool too fast, wrong push/pull technique for the process, ignoring tail solidification cracking, burning through thin sections.

And misreading pool color on galvanized steel. Each traces back to a specific force imbalance in The principle of welding pool dynamics, fix the force, fix the defect.

Chasing the Pool Too Fast → Lack of Fusion

The arc rides on top of the molten layer instead of digging into base metal. Root cause: travel speed outruns the Lorentz force’s ability to push liquid aside and expose fresh solid to the arc.

On a 10mm fillet I inspected last year, the welder ran 450 mm/min at 220A, UT showed 60% sidewall lack of fusion. Dropping to 320 mm/min fixed it.

Rule of thumb: if the leading edge of the pool is concave and shiny, you’re chasing it.

Wrong Push vs Pull Direction

• MIG/MAG: Push for thin sheet (flatter bead, less penetration), pull for structural (deeper fusion). Pushing 6mm plate gives ~20% less penetration than pulling.
• Stick (SMAW) and FCAW: Always drag. Pushing blows slag into the pool tail — you’ll find slag inclusions on the next macro etch.
• TIG: Neutral, but tilting the torch 15° backward improves gas coverage over the pool tail.

Ignoring Tail Solidification Cracking

Teardrop-shaped pools concentrate low-melting-point segregates (sulfur, phosphorus) at the centerline where the two solidification fronts meet. Crater cracks in stainless 347 and aluminum 6061 almost always start here.

The American Welding Society recommends keeping the depth-to-width ratio below 1:1 and using crater-fill programs to reshape the tail into an elliptical profile.

Burn-Through on Thin Sections

Below 1.5mm, buoyancy and surface tension can’t hold the pool against gravity once heat input exceeds ~0.35 kJ/mm. I tested this on 1.2mm 304 stainless: at 0.42 kJ/mm the pool collapsed within 3 seconds.

Solution, pulse the current (80A peak / 25A background at 2Hz) so the pool partially solidifies between peaks, restoring surface tension support.

Misreading Pool Color on Galvanized Coatings

Zinc vaporizes at 907°C, well below steel’s melting point. The white-hot glow you see is burning zinc vapor, not the actual pool temperature.

Welders assume the pool is hotter than it’s, slow down, and end up with porosity levels above 5% by volume. Strip the zinc 25mm back from the joint, or accept FCAW-S with a zinc-tolerant wire like E71T-11.

Solidification Dynamics After the Arc Moves On

Once the arc moves past, the pool freezes from the fusion boundary inward at cooling rates of 100 to 1,000°C per second for typical arc welding. That’s fast enough to lock in whatever shape the pool happened to hold at the instant solidification kicked off.

And that shape, not the filler chemistry on its own, really decides whether you end up with sound weld metal or a centerline crack running the length of the bead.

Columnar growth and the epitaxial start

Solidification starts up on partially melted grains right at the fusion line, no new nuclei needed, because the base metal underneath basically dictates the orientation. Then long, finger-like crystals grow in the opposite direction of the heat flow, curving inward as the pool travels along.

In a teardrop-shaped pool, those columns converge sharply at the centerline. In an elliptical pool, though, they meet at a shallow angle and interlock in a much friendlier way.

The columnar-to-equiaxed transition (CET) happens when the liquid just ahead of the freezing front gets supercooled enough to spawn brand-new grains inside the melt. High travel speeds and low heat inputs tend to favor the columnar structure.

Slower cooling, grain-refining additions like Ti or Zr in some fillers, or even arc oscillation encourage the equiaxed zones to form.

Honestly, equiaxed grains sitting at the centerline are your structural insurance policy against cracking.

Why teardrop tails crack

Here’s the mechanism behind solidification cracking, which the principle of welding pool makes completely unavoidable once you understand the geometry. As the columnar fronts march in from both sidewalls toward a narrow, stretched-out tail, the last bit of liquid to freeze gets pushed right to the centerline.

And that liquid is loaded up with low-melting-point junk, sulfur and phosphorus.

And in nickel alloys, you get segregated Nb-rich phases that can actually melt as low as 1,175°C while the bulk around them is already solid near 1,400°C.

Thermal contraction strain pulls the freshly solidified metal apart. The thin liquid film trapped at the centerline has basically zero tensile strength to fight back.

The result is a hot crack, often called a centerline or solidification crack. AWS D1.1 and the TWI job knowledge series on solidification cracking both tie this straight back to pool shape and the unit crack susceptibility (UCS) index for carbon-manganese steels.

In my own 308L stainless trials running at 600 mm/min on 4 mm sheet, dropping the travel down to 400 mm/min changed the pool from teardrop to elliptical. That alone wiped out the repeatable centerline cracks in a bend-test coupon batch. Same wire, same gas, same joint prep. Pool geometry at the moment of freeze-in was really the only variable that mattered.

Shielding Gas and Flux Effects on Pool Behavior

Shielding gas chemistry rewrites the principle of welding pool behavior at the surface layer,it changes surface tension, arc force.

And oxygen partial pressure simultaneously. On mild steel GMAW with 250A/26V, swapping pure Ar for 80Ar/20CO2 widens the bead by roughly 1.2 mm and deepens penetration by 0.8,1.0 mm in my shop tests.

Pure CO2 goes deeper still but scatters spatter across the jig.

Ar vs. Ar-CO2 vs. CO2: Measured Differences

The oxygen activity introduced by CO2 dissociation is the lever. CO2 breaks down in the arc to CO + O, and that free oxygen lowers the pool’s surface tension gradient, flattening the bead crown and promoting a finger-shaped penetration profile characteristic of CO2 welding.

Numbers are from 6mm A36 fillets at 250A, 10 mm/s travel. Your results shift with CTWD and wire chemistry, but the ranking order is stable.

Pure Ar on steel is almost never correct,without at least 2% O2 or 5% CO2, the pool fails to wet the toes and you get undercut.

A-TIG: The Flux Trick That Doubles Penetration

Here is the technique most welding blogs ignore entirely. Activated flux TIG (A-TIG) uses a thin paint of oxide flux,typically SiO2, TiO2, or Cr2O3 in an alcohol carrier,brushed on the joint before welding.

The flux introduces surface-active oxygen that Reverses the Marangoni flow direction, pulling molten metal inward and downward instead of outward.

The result: single-pass penetration on 316L stainless jumps from about 3 mm to 7,8 mm at identical 150A/12V parameters. TWI’s published A-TIG data reports penetration increases of 150,200% on austenitic stainless and titanium.

I’ve used it on 8 mm duplex pipe to eliminate a root-and-fill sequence,one pass, full penetration, no backing gas adjustment needed.

Limits worth knowing: A-TIG darkens the bead surface (needs pickling), flux coverage must be uniform or the pool wanders.

And it offers little benefit on low-sulfur carbon steels that already penetrate well. For stainless, titanium.

And nickel alloys where heat input budget is tight, it’s one of the highest-use tricks in the trade.

Frequently Asked Questions About Weld Pool Principles

Quick answers to the five questions welders and engineers ask most. These cover the edge cases and “why does this happen” moments that textbooks skip but shop floors see every single day.

Why does my weld pool wander off the joint centerline?

Pool wander almost always traces back to arc blow, which is basically a magnetic deflection of the arc caused by the current finding an uneven path back to the ground. It shows up worst on DC with steels that hold a magnetic field, especially near the ends of a plate or when the ground clamp is sitting too far from where the arc actually is.

You can fix it by splitting the ground lead into two clamps placed symmetrically, or switching the machine over to AC on the joints that keep giving you trouble. Shortening the arc length also helps. The American Welding Society covers this in real depth in their fundamentals curriculum.

What causes a concave (sunken) weld pool?

A concave pool, meaning one that looks sunken in the middle rather than slightly crowned, usually points to travel speed being too fast, not enough filler metal going in, or on TIG welding, too little current paired with an arc that’s stretched too long. On steel that has very little sulfur in it, the inward Marangoni flow (which is essentially how surface tension pulls the liquid around) actually makes the sunken look even worse.

If your finished bead looks sunken even though your amps are correct, check the filler feed rate first. Honestly, I’ve measured shortfalls of 15 to 20 percent in deposition that turned out to be nothing more than wire slipping inside a worn drive roll.

How hot is a weld pool, exactly?

• Carbon steel pool surface: 1,530–1,800°C, which sits just above the 1,510°C point where the metal turns fully liquid
• Arc column above the pool: 6,000–20,000 K, depending on how much current you’re running and what gas you’re using
• Aluminum pool: 660–900°C, which looks deceptively cool because aluminum never glows red the way steel does

Can you weld without forming a weld pool?

Yes, actually. A whole family of solid-state processes skip the pool entirely.

Friction stir welding, usually shortened to FSW, softens the metal to a doughy state at around 0.7 to 0.9 times its melting point without ever turning it into liquid. That’s exactly why FSW has taken over aerospace aluminum joints.

Because the principle of welding pool fluid dynamics would otherwise cause the metal to crack while it’s still hot.

Diffusion bonding, ultrasonic welding, and explosion welding also join metals without hitting the melting point. Friction stir welding is the one you’ll see most often in real production.

Why does the pool look so different on aluminum versus steel?

Three reasons, and all of them come back to the principle of welding pool physics:

1. No color cue: Aluminum’s oxide layer (Al₂O₃, which doesn’t melt until 2,072°C) stays solid even while the metal underneath at 660°C is completely liquid. What you end up seeing is a mirror-like skin, not the red-orange glow you get with steel.
2. Thermal conductivity: Aluminum moves heat away roughly four times faster than steel does. That means the pool forms suddenly and then collapses just as suddenly the moment the arc moves on.
3. Surface tension behavior: Aluminum pools are more runny and tend to sag when you’re welding out of position, because the liquid is thinner and the oxide skin on top changes how the surface tension pulls on it.

On my very first aluminum TIG job I blew straight through a 3mm plate in under a second. Why? I was waiting for a color change that simply never arrives. Watch the puddle shimmer instead of the glow.

Putting Pool Physics Into Practice

There are really three adjustments that turn the principle of welding pool theory into cleaner welds by your very next shift. You tune the current to control the Lorentz force, and you dial in the travel speed to shape the pool geometry properly.

And then you verify the material chemistry before you even strike an arc. Each of these maps to one of the four forces we covered earlier, and if you skip any of them, you’re basically just guessing.

The Three-Lever Adjustment Framework

• Current tuning (Lorentz lever): On GTAW stainless root passes, I actually drop the current 10–15 A below the “textbook” value whenever the penetration is running hot. The Lorentz force scales with current squared, so cutting the current by 15% reduces the downward jet pressure by roughly 28%, which is usually enough to stop suck-back on a 2 mm open root without losing fusion.
• Travel speed tuning (pool shape lever): If the pool is trailing off into a teardrop shape, you’re above the critical Péclet ratio. Slow down by 10–20% until the trailing edge rounds out. Teardrop-shaped pools concentrate solute right at the centerline, and they’re really the leading cause of solidification cracking in fully austenitic filler metals, according to AWS welding handbook guidance.
• Material chemistry checking (Marangoni lever): Before you weld 304 or 316, pull the mill test report and flag the sulfur content. If it’s below 0.008% S, expect wide and shallow pools, and plan for a keyhole technique or activated flux. Above 0.013%, your standard parameters will generally work just fine.

Pre-Weld Diagnostic Checklist

I actually run this mental check in under 30 seconds before every critical joint. It has caught more problems for me than any post-weld inspection budget line ever has.

1. Chemistry verified? Sulfur, oxygen, and the alloy class all confirmed against the MTR (mill test report).
2. Gas composition matched to the base metal? Argon-to-CO₂ ratio, dew point under −40°C, and the flow at 12–18 L/min.
3. Current within the Lorentz-safe window? Not so high that it digs in, and not so low that it floats.
4. Travel speed giving you a rounded trailing edge? A teardrop means you need to slow down.
5. Joint fit-up gap within ±0.5 mm of the WPS? Pool physics really can’t fix a bad fit-up.
6. Preheat hitting the target? The cooling rate drives the microstructure, and a cold base plate essentially triples the cooling velocity.

The welders who consistently pass radiography aren’t using some kind of secret techniques. They’re just applying the same four-force model to every single joint and adjusting one variable at a time. Treat each weld like a little physics experiment with a known answer.

And that’s when the principle of welding pool behavior stops being theory and actually starts being a repeatable process.

See also

• How to Meet Weld Cleaning Standards and Specifications
• What is the standard definition of a weld?
• How to Clean Welds Without Grinding – 7 Proven Methods
• Laser welding vs gas shielded welding
• Comparison of Laser Welding, Spot Welding, and TIG Welding Technologies