Submerged Arc Welding Procedure: A Complete Step-by-Step Guide
Table of Contents
A fabricator once told me that the first time he watched submerged arc welding, he thought someone had switched the machine off — there was no visible arc, no spatter, and barely a sound, just a quiet crawl of wire into a gray mound of granules. Yet minutes later, when the leftover flux was vacuumed away, a perfectly smooth, glowing bead was sitting underneath. That hidden arc is exactly why the SAW procedure is both one of the most productive and one of the most misunderstood processes in heavy fabrication.
What Is Submerged Arc Welding and Why the Procedure Matters
Submerged arc welding, usually shortened to SAW, is a fusion welding process in which an arc is struck between a continuous, bare consumable wire electrode and the workpiece, and that arc is completely buried under a blanket of granular, fusible flux. The operator never sees the arc itself. The flux melts ahead of and around the arc, forming a liquid slag and a shielding gas envelope that keeps oxygen and nitrogen out of the weld pool. When the slag cools and chips away, what is left underneath is a smooth, uniform bead with a refinement that manual processes struggle to match at anything close to the same speed.
That buried arc is the whole trick, and it explains most of the process’s advantages. Because the arc is insulated by flux rather than open to the air, very little spatter is produced, virtually no weld fume escapes from the arc zone, and heat loss to the surroundings is reduced. The engineering literature on SAW consistently describes it as one of the highest-deposition processes available for steel, and in practice it has become the default choice for long, continuous seams in heavy fabrication.
Where SAW Is Used
You will find submerged arc welding wherever long seams and thick sections are joined in volume:
- Pressure vessels and boilers — longitudinal and circumferential seams in carbon and low-alloy plate.
- Pipe and tube — both longitudinal seam pipe and spiral-wound line pipe.
- Structural plate fabrication — bridge girders, built-up beams, heavy welded plate structures.
- Shipbuilding — hull panel seams and stiffener welds, often on panel lines.
- Tank and silo fabrication — storage tanks for oil, gas, water and grain, including field-erected tank seams.
- Wear-facing and cladding — stainless or hardfacing strip overlays on carbon steel substrates.
The unifying theme is material: mild steel plate, generally from about 5 mm upward, welded in the flat or horizontal position where the process performs at its best.
Why the Procedure Matters So Much
SAW is, by its nature, a mechanized process. The wire feeds through a machine head, the head travels on a tractor, boom, or column-and-arm rig, and the parameters are dialed in rather than judged by eye. That automation is exactly why a written procedure carries so much weight. In manual arc welding, a skilled welder compensates for a poor setup by adjusting technique on the fly. In SAW the machine will happily repeat the same mistake a hundred metres down a seam, and it will do it at high deposition rate.
So what does a proper SAW procedure actually consist of? At minimum, it covers:
- Joint design and preparation — groove type, bevel angle, root face, root gap, backing method, and cleaning requirements.
- Consumable selection — the wire and flux combination, matched to base metal and required deposit chemistry.
- Parameters — welding current, arc voltage, wire feed speed, travel speed, electrical stickout, and flux bed depth.
- Preheat and interpass temperature — dictated by material grade, thickness, and hydrogen control.
- Flux handling — storage, drying, screening, and reuse rules.
- Post-weld steps — slag removal, flux recovery, heat treatment if required, and inspection criteria.
Get any one of those wrong and the process still runs. It just runs badly.
Consumables and Classification Basics
SAW wires are bare, so the flux does more than shield; it also supplies deoxidizers, alloying elements, and arc-stabilizing ionizers. Because flux and wire act as a matched system, the industry classifies them together. Carbon steel SAW consumables fall under AWS A5.17/A5.17M, while low-alloy steel consumables are covered by AWS A5.23/A5.23M. You will see combinations written as, for example, an electrode designation such as EM12K paired with a flux designation such as F7A2-EM12K, where the “F” number describes the flux type and its usability and diffusible hydrogen level. Welders migrating from stick welding will recognize the F-number concept from SMAW rod classifications, and the Advantages of Stick Welding guide explains how that system works on the manual side.
Note: The wire, the flux, and the polarity form one qualified unit. Substituting a different flux brand for the same “F” classification can still shift deposit chemistry and mechanical properties, so always confirm the combination is covered by your qualification.
SAW also anchors a small family of related processes. Electroslag welding uses a conductive slag bath rather than a true arc to join very thick sections in a single pass. Strip cladding runs a wide flat strip electrode to lay down corrosion-resistant overlays, a routine practice in pressure vessel fabrication. Multi-wire, tandem, and dual-arc variants simply multiply the electrodes under one shared flux bed to push deposition even higher.
The productivity gains reported by fabricators with SAW come down to this combination of high deposition, zero stop-start, and consistent results over long seams. That is why the procedure itself, not the operator’s hand, is the thing worth getting right. For a fuller look at the technology, AWS-Schaefer’s SAW technology overview is a solid technical reference.
How the SAW Process Works: Arc, Flux and the Slag System
Submerged arc welding keeps its arc out of sight. A bare wire feeds continuously into a mound of granular flux, striking between its tip and the base metal beneath it. With the arc fully buried, there is no visible flame, little to no spatter, and no need for a shielding gas cylinder. The flux handles shielding itself in two stages: it melts ahead of the arc, then solidifies over the finished bead. The AWS practical guide to submerged arc welding describes it as acting at once as a consumable, a shield, and a metallurgical control.
The Liquid Slag Blanket
The flux granules directly above and ahead of the arc melt first. Rather than mixing with the molten steel, this liquid floats to the top and forms a continuous blanket over the pool. Its primary job is exclusion, blocking nitrogen and oxygen from the atmosphere to prevent porosity and brittle nitride formation. The blanket also insulates the pool, slowing its cooling just enough for dissolved gases to rise and escape before it sets.
Once cooled, the slag vitrifies into a glassy crust that lifts off in long strips and is vacuumed away with any unfused flux. Clean slag release is often a sign of a balanced procedure, while flux that clings to the bead can be an early warning for issues such as lack of fusion and porosity in SAW work.
What the Flux Does to the Weld Metal
SAW flux is a formulated blend, most often built from lime, silica, and fluorspar. It actively changes the weld metal in four practical ways:
- Shields and refines: As it melts, the flux produces a gaseous blanket over the arc and absorbs impurities into the slag, leaving a cleaner weld than many open-arc processes.
- Deoxidizes: Elements such as manganese, silicon, aluminium, and titanium scavenge residual oxygen to prevent blowholes and hard oxide inclusions.
- Alloys: Some formulations deliberately add molybdenum, chromium, or nickel to restore composition lost to arc burn-off. This alloy-tuning follows the same logic seen in Alloy Steel Welding Electrodes: Selection, Use & Safety, only delivered as a powder.
- Stabilizes the arc and shapes the bead: Fluorides and alkali compounds ionize easily to keep the arc stable, while the slag’s viscosity and surface tension control wetting and spread. The AWS Schaefer submerged arc technical overview notes that flux chemistry can drastically shift penetration, even if amps and travel speed stay the same. This buried-flux principle is also seen outside of steel, as outlined in Aluminum Flux Welding: How It Works and When to Use It.
Note: Flux grade, reuse, and moisture all matter. Grain size affects arc confinement and gas escape, so always use the grade your procedure qualified. Unfused flux can be screened and reused, but damp flux is the leading source of hydrogen and porosity. Keep it sealed, off the floor, and dried to the manufacturer’s data sheet before returning it to the hopper.
Polarity, Current Type and Arc Blow
Current type is just as critical as flux chemistry. Polarity determines where heat concentrates, but the exact balance of penetration and deposition will shift with flux type and wire diameter, so only settings written into a qualified WPS should be used.
| Current Setting | Typical Behaviour |
|---|---|
| DC electrode positive (reverse) | Deeper, narrower penetration, but more prone to magnetic arc blow |
| DC electrode negative (straight) | Higher deposition with a wider, flatter bead and generally shallower penetration |
| AC | Balanced penetration with far less arc blow, making it well suited for multi-wire and tandem heads |
Arc blow is the magnetic deflection of the buried arc caused by the current path itself. It’s most common on DC, near plate edges, corners, or joint ends. Where allowed, switching to AC is the most effective fix. Other practical steps include trimming excess current, correcting electrode angle, and adding extra flux coverage at starts and stops where the magnetic field is at its strongest.
Together, the buried arc, liquid slag blanket, and correct polarity create a tightly controlled, high-deposition weld. By the time the vacuum clears away the gray mound, all of that protection has already done its job.
Joint Design and Preparation for SAW
Every good submerged arc weld starts before the machine is switched on. Because SAW deposits metal so efficiently, the joint itself can be far leaner than a manual process would demand — but that same deep, fluid penetration makes fit-up and cleanliness far less forgiving of shortcuts. A joint that is too wide wastes filler metal and time; a joint that is dirty or loosely fitted shows up later as porosity, lack of fusion, or a repair. This is the stage where the numbers written in the WPS actually get built.
Why SAW Allows Narrower, Cheaper Joint Geometry
Compared with stick or MIG welding, SAW drives a concentrated arc deep into the plate with a blanket of flux holding the heat in. The practical result is that groove angles can be reduced and root faces increased, so fewer passes are needed to fill a given thickness. Where a manual butt joint might use a 60° included V, SAW joints commonly run at 45° or even less, and very thick sections are often welded as a square or narrow-groove joint with multiple passes from one or both sides. Less groove volume means less deposited weld metal, less flux consumed, and fewer passes — a direct cost saving on heavy plate. AWS’s practical guide to submerged arc welding covers how penetration characteristics shape groove selection.
| Joint type | Typical SAW use | Notes |
|---|---|---|
| Single-V butt | Plate to about 25–30 mm | Common workhorse; often welded from both sides after flipping |
| Double-V / double-bevel | Thicker plate | Balances distortion; smaller total fill than a single deep V |
| J-groove and U-groove | Heavy wall, pressure vessels | Narrower than a V, more edge prep cost, less fill |
| Square groove with backing | Ship plate, structural, tank work | Zero bevel; requires a backing strip or a sound root |
| Lap and T joints | Fillet welds, stiffeners | Extremely productive; SAW fillets are a staple of fabrication |
Bevel Angle, Root Face and Root Gap
Three dimensions control the outcome: the included bevel angle, the root face (the flat vertical land left at the edge of the bevel), and the root gap. Widen the gap and you raise the risk of excessive penetration and burn-through; close it too much and the arc cannot reach the bottom, leaving lack of fusion at the root. A generous root face paired with a modest angle keeps the first pass stable and gives subsequent passes a clean sidewall to tie into. Fit-up tolerance matters just as much as the nominal numbers — joint geometry, root opening, and alignment are listed as essential variables in procedure qualification under both AWS D1.1 and ASME Boiler and Pressure Vessel Code Section IX, which means a fit-up outside the qualified range technically puts the weld outside the qualified procedure.
Quick Note: Record the actual bevel angle, root face, root gap, and alignment tolerance in the WPS with a sketch or a tolerance range — not just “as prepared.” If fit-up drifts during fabrication, you need a documented allowance to know whether it is still code-compliant.
Backing: How One-Sided Welding Kills Back Gouging
The classic problem with single-sided joints is the root: unless the first pass fully fuses through, the back of the weld must be gouged out and re-welded, which on heavy plate is slow, expensive work. SAW solves this with backing. The main options are:
- Steel backing rings or strips — cheap, structural, common in pipe and vessel fabrication, then either left in place or removed.
- Flux backing (flux dams) — a layer of loose granular flux packed under the joint, held by a backing bar or tape, which shields and supports the molten root.
- Flux-copper backing — a water-cooled copper shoe topped with a thin flux layer; the copper shapes and cools the bead while the flux prevents arc action on the copper. This is the standard route to a full-penetration, one-sided weld with a smooth, often unmachined, reverse side.
A properly backed one-sided joint achieves full penetration without any back gouging, which is a large part of why SAW dominates tank and ship plate fabrication. Productivity write-ups on SAW repeatedly point to one-sided welding as the single biggest gain.
Surface Preparation and Cleaning
SAW is sensitive to anything that does not belong in the puddle. Oil, grease, paint, standing water, thick mill scale, and heavy rust all break down into gas or slag inclusions — and because the arc is covered, you often do not see the problem until the bead is uncovered. The standard requirements are straightforward:
- Remove oil, grease, paint, and moisture from the joint faces and at least 25 mm either side.
- Wire-brush, grind, or blast loose mill scale and rust back to sound metal; heavy mill scale must go.
- Grind plate edges where bevels were flame-cut, especially the root face, to remove the hard, oxidized cut surface.
- Keep damp flux dried and open bags sealed — moisture in the flux is a direct porosity source.
The same cleaning logic transfers across processes; if you also run aluminium work, the discipline described in aluminum surface prep applies, just with different chemistry. For the failure side, The Fabricator’s troubleshooting article on lack of fusion and porosity in SAW is a useful reference: contaminated joint surfaces and poor fit-up sit at the top of the cause list.
Heat Input, Thick Plate, Preheat and Flux Drying
Thick sections eat heat, but they also hold it. Because SAW runs at high current with slow, steady travel, heat input climbs quickly, and on heavy plate that means interpass temperature must be watched and preheat requirements — set by the code, the material’s carbon equivalent, and the thickness — followed exactly. Equally important is flux conditioning: store it dry, screen out the fines, and preheat or rebake it before use. Damp flux is the cheapest defect you will ever prevent.
The SAW Procedure Step by Step
Knowing the theory of submerged arc welding is one thing; running a clean bead is another. The procedure is a fixed sequence, and most problems trace back to a skipped or guessed step. Here is how a typical SAW pass runs, from paperwork to final inspection.
Step 1: Review the WPS and Confirm the Consumable Combination
Before anything is switched on, the operator checks the WPS: approved current range, voltage, wire feed speed, travel speed, electrode size, flux type, preheat and interpass temperature, joint geometry and position limitation. Work outside those ranges and the weld is technically unqualified, no matter how good it looks. Confirm the wire and flux pairing is the one listed in the WPS, since a different electrode or consumable grade can shift weld metal chemistry even when diameter and current settings match. The AWS practical guide to SAW is a useful pre-job refresher on how tightly consumables and parameters are paired.
Step 2: Machine Setup and Joint Preparation
Mount the wire spool, thread the liner and contact tip, check the conduit is clean and the drive rolls are correctly tensioned, then mount the flux hopper and vacuum recovery system. Align the electrode with the joint centre line — on a bevel joint, point the wire at the root face or joint centre as the WPS specifies, usually with a small drag angle of roughly 10–15°. Set the backing: a backing ring, a ceramic strip, or a flux-copper backing for one-sided welding. Finally, verify the joint is clean and dry: rust, oil, moisture and heavy mill scale all translate directly into porosity and inclusions.
Step 3: Set the Flux Bed and Electrical Stickout
Open the hopper and lay a layer of flux roughly 25–40 mm (1–1.5 in) deep ahead of the arc. That depth fully buries the arc, stops UV flash and holds the shielding gases in place. Too little flux and you will see the arc; too much and gas can get trapped, causing blowholes or an uneven bead. Then set the electrical stickout — the distance from the contact tip to the workpiece.
Quick Note: Stickout typically runs 25–50 mm. Because the wire itself carries resistance, increasing stickout lowers the deposited current for the same wire feed speed. If you lengthen the stickout mid-job to clear a fixture, expect less penetration and a wider, flatter bead — compensate deliberately rather than by accident.
Step 4: Initiate the Arc
SAW arcs do not start the way a stick electrode does. The usual methods are:
- Steel ball start: a steel sphere is placed in the joint, flux is piled over it, and wire feed starts. The ball shorts the wire to the plate, draws current and melts through, igniting the arc under the flux.
- Retracting-wire start: automated rigs touch the wire to the plate, then release it, striking the arc as the wire pulls away.
- Scratch start: the wire is dragged across the plate like a matchstick, mainly on simple manual-positioned rigs.
Whichever method you use, the arc must ignite under flux, never in open air — an exposed SAW arc burns at full intensity, an immediate UV flash hazard.
Step 5: Travel, Monitor and Stop
Once the arc is stable, the tractor moves off at the programmed travel speed. Watch three things continuously: flux bed depth ahead of the wire, the molten zone at the trailing edge, and the readouts for current and voltage. Adjust flux feed if the bed thins, and stop the wire if the contact tip begins to wander. To finish, stop wire feed and travel simultaneously, back off the voltage, then bury the crater under fresh flux so the final solidification happens shielded rather than exposed.
Step 6: Flux Recovery, Slag Removal and Inspection
Let the joint cool for a few minutes. Vacuum up the unreferenced flux, screen it for spatter and slag, and return the clean fraction to the hopper. Then break the slag: chip it with a hammer, or simply tap the bead edge — many high-basicity fluxes lift off in one continuous piece, a small daily pleasure. Inspect the exposed bead for surface porosity, undercut, waviness and crater cracks before any grinding or NDT is scheduled. As The Fabricator’s troubleshooting guidance notes, catching a parameter drift at the slag-removal stage is far cheaper than finding it after a radiograph.
| Stage | What to check | Typical fault if skipped |
|---|---|---|
| WPS review | Current, flux, preheat | Unqualified weld |
| Setup | Alignment, cleanliness, backing | Misalignment, porosity |
| Flux bed | 25–40 mm depth | UV flash, blowholes |
| Stickout | 25–50 mm, stable | Current drift, lack of fusion |
| Arc start | Ignition under flux | Crater porosity, spatter |
| Post-weld | Slag lift, bead inspection | Hidden defects reaching NDT |
Run the sequence in this order every time and the process becomes predictable. Skip a step and the arc will still make a weld — just not the weld your procedure promised.
Key Welding Parameters and How to Set Them
One of the nice things about submerged arc welding is that the machine holds variables steady once they are set. One of the dangerous things about SAW is exactly the same. Because the arc is buried you cannot see it responding, so a drift that would be obvious in stick or MIG welding can run for two metres unnoticed before anyone checks the bead. Setting parameters deliberately, then watching the few signals you do have (bead shape, slag colour, amp and volt readouts), is the whole craft. A practical guide to SAW in Welding Digest puts it plainly: the parameters are coupled, and changing one almost always moves several weld characteristics at once.
Current and Wire Feed Speed
In a typical constant-voltage SAW set-up you do not dial in amperage directly. You set wire feed speed, and the arc physics settle on a current matching it, influenced heavily by stickout. For a single solid wire of roughly 1.2 to 1.6 mm, the common working envelope is about 300 to 500 amps at 25 to 35 volts. Thinner wire runs toward the lower end; thicker wire and higher deposition work runs toward the top.
Current is the penetration driver. Raise it and the arc digs deeper, penetration increases and deposition climbs, because more wire is consumed per minute. Too much current for the plate thickness and you risk burn-through, excessive dilution with the base metal, and a coarse, fast-freezing bead.
Voltage: The Bead Shaper
Voltage mainly controls arc width, bead width and profile. Higher voltage widens the arc, spreads the bead and flattens its profile, which is useful for covering a wide joint in fewer passes. Too much voltage, though, and the arc starts to wander under the flux, slag can flow ahead of it, and you get sidewall undercut with a glassy, high-standing bead edge. Too little voltage produces a narrow, tall, convex bead with poor wetting at the toes, and often lack of fusion at the joint walls.
Travel Speed, Stickout, Flux Depth and Electrode Angle
Travel speed sets how much energy and metal go into each millimetre of weld. Push it too fast and you get thin, ropey beads and the classic lack-of-fusion failure mode that consumables troubleshooting articles point to first. Run too slow and you pile up heat input and get a wide, washed-out bead.
The flux bed generally wants about 25 to 40 mm (1 to 1.5 inches) of granular coverage. Too shallow and the arc breaks out of the blanket, spattering and oxidising; too deep and the arc struggles to strike, and the bead may pick up unmelted flux, showing up as worm tracking. Electrode angle matters less than in manual processes, but a slight drag angle of about 10 to 15 degrees from vertical is the usual starting point, and keeping the wire centred in the joint is non-negotiable.
Quick Note: Electrical stickout (the length of wire from the contact tip to the arc) acts as a series resistor. Lengthen it and resistance rises, so the machine pushes more current through the same wire feed speed, increasing penetration and deposition. Shorten it and amperage drops. This is why stickout must be held constant during a run: it silently changes your amperage while the display settings look unchanged.
| Parameter | Increase it and you get | Watch for |
|---|---|---|
| Current (via wire feed) | More penetration, more deposition | Burn-through, high dilution |
| Voltage | Wider, flatter bead | Undercut, arc wander |
| Travel speed | Thinner bead, less heat per mm | Lack of fusion at speed |
| Stickout | Higher effective amperage | Tip burn-back if too short |
| Flux depth | Better shielding, quieter arc | Worm tracking if excessive |
Calculating Heat Input
Heat input is the number you record on every run if you want repeatable metallurgy:
Heat input (kJ/mm) = (Voltage × Amperage × 60 × efficiency) ÷ (travel speed in mm/min × 1000)
SAW is the most thermally efficient arc process, with efficiency commonly taken around 0.9 to 1.0, though your governing code may specify a different factor. Higher heat input softens the HAZ and can coarsen grain structure, which is why thin sections and fine-grain steels need parameter restraint, and why interpass temperature has to be tracked alongside it.
Multi-Wire, Tandem and Strip Electrodes
Once a single wire is dialled in, productivity can be pushed much further. Multi-wire systems share one arc zone; tandem machines use two independently controlled arcs, typically a leading wire set for deep penetration and a trailing wire set for cap and profile control, with separate voltage and wire feed on each. Strip electrodes replace round wire with a flat band, giving enormous deposition rates for cladding and surfacing work. As outlined in productivity coverage in The Fabricator, these arrangements can double or triple metal deposition, but every added wire adds variables, so qualification trials matter more. The same discipline applies to consumables: wire and flux must be qualified as a matched pair, the same way alloy steel welding electrodes are selected strictly by classification rather than convenience.
One honest limitation: SAW parameters assume flat or horizontal work. If the job needs vertical or overhead capability, this entire parameter set is irrelevant, and an all-position welding rod with a different process is the right answer.
The practical takeaway: pick current for penetration, voltage for profile, travel speed for heat per millimetre, then lock stickout, flux depth and angle and write every value down. A parameter sheet that is measured rather than guessed is the difference between a procedure and a habit.
Welding Procedure Specification (WPS) and Qualification Essentials
Every reliable submerged arc weld starts on paper before it starts on steel. The welding procedure specification, or WPS, is the controlled document that tells the operator exactly how a joint must be made: which consumable, which current, which temperature ranges, and which travel speeds were proven to produce an acceptable weld. In heavy fabrication, an unqualified SAW parameter set is not a paperwork problem, it is an accountability problem, because pressure vessels, structural plate and pipeline work are all built to codes that demand documented proof.
What a SAW WPS Must Contain
A complete WPS identifies the process (SAW per AWS A5.17 or A5.23), the base metal groupings, the filler metal classification and diameter, the flux classification, current type and polarity (DC straight, DC reverse or AC), shielding or backing conditions, and the essential variable ranges within which the welder may operate. For SAW specifically, the practical ranges that matter most are:
- Welding current and range (for example 300–500 A for 1.2–1.6 mm wire)
- Arc voltage range and travel speed limits
- Heat input limits, often the controlling variable on thick plate
- Preheat and interpass temperature minimums and maximums
- Electrode diameter, stickout and orientation
- Single-pass versus multi-pass technique, including whether backing is required
Heat input deserves special attention in SAW. Because deposition rates are high, it is easy to exceed the energy the metallurgy allows. A WPS will typically cap heat input to protect toughness in the heat-affected zone, and the operator calculates it from amperage, voltage and travel speed. Guidance on the practical side of these settings is covered in our piece on the Alamo Welding & Boiler Works services and boiler expertise, which shows how fabrication shops translate code requirements into day-to-day shop practice.
Qualification Codes: ASME Section IX, AWS D1.1 and API 1104
Three codes dominate SAW qualification, and which one applies depends on the product:
| Code | Typical application | Qualification focus |
|---|---|---|
| ASME Boiler and Pressure Vessel Code Section IX | Pressure vessels, boilers, power piping | Qualifies the procedure (PQR) and the welder/performance separately |
| AWS D1.1 | Structural steel fabrication | WPS + PQR with prequalified or test-based procedure qualification |
| API 1104 | Pipeline welding | Procedure and welder qualification through test welds and inspection |
Under ASME Section IX, the procedure is qualified through a Procedure Qualification Record (PQR). The PQR is the evidence file: a test coupon is welded using the proposed parameters, then examined visually and nondestructively, and finally mechanically tested with tensile and bend tests to prove the joint actually performs. The WPS is then written from the ranges demonstrated on that successful PQR, not the other way around. AWS D1.1 follows the same underlying logic, and research on SAW-specific qualification confirms that the process’s variable sensitivity makes that test-record link especially important (NSRP SAW qualification study). The general framework for these requirements is set out in AWS guidance such as A Practical Guide to Submerged Arc Welding.
Essential Variables: What Triggers Requalification
Essential variables are the settings that, if changed beyond the qualified limits, invalidate the qualification. For SAW, common requalification triggers include:
- Changing electrode diameter or going from solid wire to strip cladding
- Switching current type or polarity (DCEN to DCEP, or DC to AC)
- Substituting a flux with a different basicity or alloy system without qualification
- Exceeding qualified heat input, preheat or interpass ranges
- Changing joint geometry outside the qualified range, such as going from double-sided to single-sided with a backing
Note: Consumable and flux combinations must be qualified together. A WPS qualified for EM12 wire with a basic flux is not automatically valid for the same wire with a neutral or acidic flux, because the weld metal chemistry and penetration profile change. Treat every wire-and-flux pairing as its own qualified system until documentation says otherwise.
Documentation Practices That Hold Up in an Audit
Good shops keep the WPS, PQR, supporting test reports, consumable certificates and traceability records together, and they keep them current. Manufacturer datasheets for flux and wire should be filed with the qualification so that a substitution is caught before welding starts. Operator qualification records must match the specific WPS being used, and any essential-variable deviation on the floor is recorded as an engineering change rather than quietly absorbed.
This discipline pays off most in high-consequence work. It is the same documentation mindset used in Alloy 20 welding procedures, filler metals and best practices, where material cost and service conditions leave no room for improvised parameters. Get the qualification right, and the rest of the SAW procedure, from flux bed depth to final inspection, has something solid to be measured against.
Common SAW Defects, Causes and Troubleshooting
Because the arc in submerged arc welding sits under a blanket of flux, defects rarely announce themselves while you can still fix them. Most SAW problems surface at the slag-lift stage or on the ultrasonic report, hours after the cause has passed. Troubleshooting is therefore less about reacting to what you see and more about reading small clues: the shape of the slag, the colour of the bead, the sound of the arc, and the condition of the flux coming back through the recovery hose.
Porosity and Blowholes: The Hidden Gas Problem
Porosity is the most reported SAW defect, and in most cases it traces to three culprits: damp flux, a contaminated base metal, or excessive travel speed. Flux is hygroscopic and silently picks up moisture from the air, and that moisture becomes hydrogen in the arc. Welding over oil, rust, thick mill scale or primer adds another gas source. Travel too fast gives trapped gas less time to float out of a rapidly freezing pool, which shows up as fine porosity along the bead centre or as large, irregular blowholes.
The fixes are unglamorous: bake and screen the flux, degrease and grind the joint, and slow travel until the bead profile normalises. Consumables specialists at The Fabricator’s Consumables Corner emphasise that porosity in SAW is usually a flux and surface-preparation problem before it is a machine problem.
Lack of Fusion and Incomplete Penetration
Lack of fusion is the defect most likely to pass visual inspection and fail the customer. It typically results from inadequate heat input for the plate thickness, poor fit-up (an excessive root gap or a proud misalignment), or a weld head that has wandered off the joint centreline so the arc digs into one sidewall instead of the root. Incorrect polarity compounds it: running DCEN with a flux formulated for DCEP gives shallow, wandering penetration.
Check travel speed and current together, realign the head and carriage tracks, and confirm polarity and stickout against the WPS before adding amps. On thick sections, a second pass or a multi-wire setup is usually smarter than pushing one wire past its limit.
Undercut, Worm Tracking and Uneven Beads
Undercut appears at the toe lines and comes from excessive voltage or travel speed relative to the current: the arc cuts a groove the filler is not refilling. Lower the voltage or slow down, add a little current, or reduce the electrode angle so the pool wets into the toe.
Worm tracking is a tunnel of trapped slag or gas just under the bead surface, often seen when the flux is too fine or high in bonded particles, when the flux bed is very deep, or when current exceeds what the flux chemistry supports. Sieving or replacing the flux and reducing bed depth usually clears it.
Uneven or wandering beads point to unstable wire feed, a worn contact tip changing stickout, dirty conduit, or inconsistent carriage speed. None are metallurgical problems, which is why they are so easy to overlook.
Centerline Cracking and Hydrogen Control
Weld metal cracking relates to high restraint, the wrong flux/wire combination, and excessive diffusible hydrogen. Centerline (solidification) cracking shows up in deep, narrow, high-current beads on thick restrained joints, especially where the bead width-to-depth ratio is poor. Hydrogen-induced cracking appears after cooling, often overnight, in hardenable alloy steels.
Remedies include switching to a lower-hydrogen, correctly basic flux, applying preheat, resequencing passes to reduce restraint, and widening the bead by lowering current slightly or adding a weave-free second pass. Consumable pairing is critical: a flux qualified for one wire diameter and steel class is not automatically valid for another. The same selection discipline described in Alloy Steel Welding Electrodes: Selection, Use & Safety applies to SAW wire-and-flux sets.
Slag Entrapment and Surface Contamination
Incomplete slag removal usually starts as a parameter problem. A bead with excessive reinforcement and heavy toe undercut traps slag in the crevices, and the next pass welds into slag instead of metal, producing layered inclusions visible on radiography. Chipping technique matters too; lifting slag at a shallow angle instead of driving it forward leaves residue on the toes.
Surface contamination faults, from oil and rust to grinding-wheel deposits, follow the usual logic: the joint must be genuinely clean. Fabricators who maintain strict prep habits, such as those covered in Aluminum Cleaner for Welding: How to Prep Aluminum Right, tend to carry the same routine into steel SAW work, and their porosity rates drop accordingly.
Quick Note: Never return recovered flux to the hopper without screening. Reuse is normal and economical in SAW, but slag chips, spatter and fine dust carried over from the previous bead are a leading cause of worm tracking and inclusions. Bake and sieve doubtful recovered flux, or discard it; flux is trivial next to the cost of cutting out a defective seam.
SAW Troubleshooting Chart: Symptom, Cause, Fix
| Defect | Frequent causes | What to change first |
|---|---|---|
| Porosity / blowholes | Damp flux, oily or rusty plate, travel too fast, shallow cover | Bake flux, clean joint, slow travel, check bed depth |
| Lack of fusion / cold lap | Low heat input, wide gap, head off-centre, wrong polarity | Raise current or slow travel, realign head, verify polarity |
| Undercut | Voltage or travel too high for current, wrong angle | Reduce voltage or speed, correct angle, add current |
| Centerline cracking | High restraint, mismatched flux and wire, hydrogen, narrow bead | Low-hydrogen flux, preheat, resequence, widen bead |
| Worm tracking | Over-fine or bonded flux, moisture, deep bed, current too high | Sieve or replace flux, dry it, reduce depth |
| Uneven / wandering bead | Worn contact tip, dirty conduit, unstable drive, speed variation | Replace tip, clean liner, check drive rolls and carriage |
| Slag inclusions | Heavy reinforcement, toe undercut, incomplete slag removal | Adjust parameters, chip slag properly, dress between passes |
The parameter logic behind these fixes is consistent across fluxes and is set out in technical references such as the AWS-Schaefer submerged arc welding overview and the AWS Welding Digest practical guide to SAW.
A Simple Fault-Finding Routine
When a defect appears, work in this order rather than jumping straight to the amperage knob. First, check the consumables: flux condition, wire diameter, and the flux/wire combination against the WPS. Second, check the machine: calibration, polarity, actual current and voltage at the arc, and stickout. Third, check the joint: fit-up, alignment, cleanliness, and preheat temperature. Only then adjust parameters, one variable at a time, and run a trial bead. This order matters because roughly three quarters of field SAW defects resolve at steps one and two. On jobs where a cut-out is severe, such as clad vessels and corrosive-service components like Alloy 20 fabrications, that disciplined sequence is what keeps a repair from becoming a rejection.
Safety, Fume Control and Health Hazards of SAW
Submerged arc welding has a reputation for being the friendliest process on the shop floor: no visible flash, no spatter, barely a sound. That comfort is partly earned, but it is also a trap, because most SAW hazards are invisible by design. The moment the arc is exposed, during an arc strike, a tack, a flux shortfall or routine maintenance, the process behaves like any other high-current arc process. OSHA identifies metal fume and ultraviolet radiation exposure among the core health hazards of welding, cutting and brazing operations, and neither disappears because the arc is buried under granules.
Ultraviolet Radiation and Arc Flash
Under a proper flux bed, ultraviolet and infrared radiation stay contained. The problem is that the bed is not always proper. Arc strikes, the first few seconds before flux flow settles, a bridged hopper, or an operator leaning in to check a tie-in can expose the arc, and at SAW currents of several hundred amps that exposed arc produces the same intense UV as any open-arc process: photokeratitis, commonly called arc eye, and flash burns on exposed skin that behave like a severe sunburn.
The shop-level risk is that SAW stations often run unattended while other workers walk past, so screening the bay is a layout decision as much as a PPE one. Full-shade helmets, opaque leathers and skin coverage stay mandatory even when nothing appears to be glowing. The same enclosure and eye-protection logic applies in other high-energy processes; Alpha Laser Welding: Systems, Uses, and Safety Guide walks through it in detail.
Welding Fume: Manganese, Oxides and the Cancer Evidence
A buried arc still makes fume. The wire, the flux minerals and the base metal all vaporize at the arc and recondense into ultrafine particles that can carry manganese, iron oxides and, when stainless or chrome-bearing steels are involved, hexavalent chromium and nickel compounds. Manganese draws the most attention for its chronic neurological effects, but the wider evidence has shifted the conversation. OSHA’s welding fume fact sheet states that prolonged exposure to welding fume may cause lung damage and several types of cancer, and in 2017 the International Agency for Research on Cancer concluded that mild steel welding fume can cause lung cancer and possibly kidney cancer in humans, a reclassification that surprised many shops that had long treated mild steel fume as harmless. The EU-OSHA OSHWiki entry on welding fume dust and aerosols reaches the same practical order of priority: engineering controls first, not after symptoms appear.
Three controls matter most inside a SAW cell:
- Source capture. Extraction hoods or nozzles positioned at the flux front and at the slag-removal station, where residual fume escapes while the bead is chipped and brushed.
- Flux chemistry. Basic and neutral fluxes generally emit less fume than highly acidic, fast-freezing types, so consumable selection is a fume-control decision as well as a metallurgical one.
- Respiratory protection. P100 half-face respirators, or supplied-air rigs for long weld-outs, where engineering controls cannot hold exposure down. This is the fallback layer, not the first resort.
Fume composition also tracks the alloying elements in the filler wire, which is worth reviewing when a job moves onto Cr-Mo or high-strength low-alloy steel; Alloy Steel Welding Electrodes: Selection, Use & Safety explains how those additions change both metallurgy and fume.
Flux Dust: Silica and Fluoride Exposure
Before it ever reaches the arc, the flux is its own exposure. Pouring, screening or scooping dry granular flux generates respirable dust from mineral silicates, and chronic inhalation of respirable crystalline silica is a well-established cause of silicosis. Some fluxes also carry fluoride-bearing ingredients that release irritating fluorides when damp or heated, which is one reason flux drying and sealed storage are procedure items rather than housekeeping preferences.
Note: Treat flux as a dusty chemical, not as inert sand. Pour it rather than dump it, wear eye protection and a dust mask when changing drums or cleaning the recovery hopper, and keep bags sealed and lifted off damp concrete floors. Damp flux does not just weld badly, it spatters and pops at arc start.
Thermal, Electrical and Mechanical Hazards
SAW leaves a lot of stored heat: a heavy deposit stays hot long after the bead looks finished, and fresh slag will burn straight through gloves and boots. Electrically, SAW machines run high current and often high open-circuit voltage, with heavy cables dragged across fabrication bay floors, so cable integrity, work-clamp placement and dry footing are non-negotiable. The moving parts are easy to overlook as well: tractors and positioners travel with real force, wire spools are heavy, and flux recovery vacuums pull hard enough to catch a sleeve. Gouging work in the same bays adds its own flash and fume load, covered in Air Arc Welding and Carbon Arc Gouging: Process, Safety, Applications, and Best Practices.
Noise, Confined Spaces and Hot Work
The welding itself is quiet, which tempts people to skip hearing protection. The support work is not: needle guns, chipping hammers, grinders and vacuum recovery units running for hours create exactly the steady, moderate-level noise that gradually damages hearing. Vessel and tank shops, the heart of the SAW market, add confined-space entry on top. Fume accumulates inside enclosed shells even though SAW uses no shielding gas, so atmospheric testing, forced ventilation, an attendant at the access point and a written rescue plan belong in the procedure, not in a binder. Hot work permits and fire watch apply too, since arc starts throw sparks and slag stays hot for a long time.
Building the Safety Layer into the Procedure
The best SAW safety programs write controls into the procedure document instead of bolting them on afterward. Specify ventilation and extraction points, required PPE and filter shades, flux handling and drying rules, guarding for tractors and positioners, and confined-space arrangements for internal weld-outs. Then verify the same way you verify a weld: trained operators, audited bays, kept records. Done that way, the calmest-looking process in the shop stays the safest one in it, and a practical guide to submerged arc welding is a useful companion when you are drafting those shop-level rules for the first time.
SAW vs. Other Welding Processes: When It Wins and When It Doesn’t
No welding process answers every job. Submerged arc welding produces some of the cleanest, most consistent and most economical welds in heavy fabrication, but it comes with hard boundaries: a fixed position, a fixed work envelope, and a heat input that is either exactly what the job needs or far too much for it. Knowing where SAW ends and another process begins is what separates a smart production line from a frustrated one.
Head-to-Head: SAW, GMAW/FCAW, SMAW and Electroslag
A practical comparison across the factors that actually decide a job:
| Factor | SAW | GMAW / FCAW | SMAW | Electroslag |
|---|---|---|---|---|
| Deposition rate | Highest of the group on thick plate | Moderate | Lowest; stub and spatter losses | Very high, single pass |
| Welding positions | Flat and horizontal only | All positions | All positions | Vertical only |
| Typical thickness | About 6 mm to very heavy plate | Thin sheet to medium plate | Any thickness, especially repair | Usually 25 mm and heavier |
| Portability | Low; tractor, hopper, flux recovery | High | Very high | Fixed clamping gear |
| Shielding | Granular flux | Shielding gas, plus flux in FCAW | Electrode coating | Slag pool |
| Heat input | High | Low to moderate | Moderate | Very high |
On deposition rate and travel speed alone, SAW is hard to beat on thick plate: the wire is continuous, there is no electrode stub to throw away, very little spatter is lost, and the arc runs without interruption for long stretches. That combination is what the productivity case for submerged arc welding rests on, and why SAW dominates pressure vessel courses, longitudinal and circumferential pipe seams, tank shells, heavy structural plate and shipbuilding panels. The AWS practical guide to submerged arc welding makes the same point from the consumables side: the process keeps operators welding instead of changing electrodes.
Against SMAW the trade-off is portability. Stick welding still wins in the field, on repairs, overhead and vertical joints, and anywhere an extension cord is not an option, a reality our Advantages of Stick Welding guide covers in depth. Against GMAW and FCAW, SAW wins on volume and bead quality but loses on flexibility: wire processes can follow a curved joint, move outdoors with gas cylinders, and handle out-of-position work. An Aluminum Flux Core Welding setup shows how flux-cored wire trades some of that positional freedom for higher deposition than plain GMAW.
Why SAW Is Locked to Flat and Horizontal
The restriction is physical, not a matter of operator skill. Granular flux holds the arc only when gravity keeps the mound sitting on the joint. Tilt the work and the flux slides away, exposing the arc, so SAW is essentially limited to flat and horizontal positions. Joints that must be welded out of position need a process designed for it; our All Position Welding Rod Guide outlines the electrodes built for that work. Layout therefore becomes part of the procedure: heavy assemblies are rotated, turned, or run on a turning roll so the joint comes to the machine rather than the reverse.
Heat Input: The Double-Edged Sword
Note: High heat input is an asset on thick, hydrogen-sensitive plate and a liability on thin or microalloyed material. The same amperage that gives deep fusion on a 40 mm shell course will burn through a 4 mm sheet and coarse the grain in a fine-grained HSLA plate.
SAW concentrates a large amount of energy into the joint, which produces deep penetration and a wide, smooth bead. The cost of that is a coarse heat-affected zone. In fine-grained, microalloyed or TMCP steels, repeated high-heat cycles grow prior-austenite grains and can knock toughness down, particularly in the subcritically reheated zone; research on heat input control in submerged arc welding examines exactly this relationship. The countermeasures are well known: lower current, faster travel, smaller-diameter wire, multiple passes instead of one wide bead, and interpass temperature control. Never run more heat than the joint design requires, and calculate it: heat input in kJ/mm is roughly (voltage × amps × 60 × efficiency) ÷ (travel speed in mm/min × 1000). Then record it against the WPS limits. For broader thermal context, the ScienceDirect overview of submerged arc welding is a useful reference.
Portability, Cost and Flux Handling
A real SAW station is a small plant: power source, wire reel, flux hopper, tractor or boom, and a vacuum system to recover unused flux. That footprint means SAW is a shop process. It does not climb a scaffold, and it does not fix a cracked bracket on a jobsite. Flux also imposes a recurring material cost and discipline that gas and stick processes do not: it must be screened, kept dry, and stored correctly, because damp or contaminated flux is a direct route to porosity. As the AWS Schaefer technical overview notes, flux conditioning is part of the process, not an optional extra.
When SAW Doesn’t Win
Choose something else when the joint is out of position, when the plate is thin, when the work is scattered across a construction site, when the alloy is sensitive to grain coarsening, or when lot sizes are too small to justify setup and flux handling. Choose SAW when you have long seams, thick plate, a rotatable fixture, and a run long enough to keep the arc burning. Match the process to the joint and SAW looks effortless; force it into the wrong job and it fails in the most expensive way possible. For related process-geometry lessons, Aluminum Flux Welding shows how flux-based shielding behaves when material and position rules change.
Best Practices, Common Mistakes and Getting Reliable SAW Results
Ask ten SAW operators what separates a clean, compliant bead from a weekend of grinding and chipping, and you will hear the same answer: the arc never fixed anything, it only revealed what the setup already decided. Reliable submerged arc results come from a short list of disciplined habits, repeated on every shift. Below is the checklist we recommend hanging near the machine, followed by the mistakes that cause most rework.
The Pre-Weld and In-Process Checklist
- Store flux dry and screen it before reuse. Leftover flux should be collected, screened to remove fines and crushed slag chips, then kept in sealed containers away from humidity. Reusing unscreened or damp flux is one of the fastest routes to porosity and erratic arcs.
- Dry the flux to the manufacturer’s recommendation and check for clumps before loading the hopper. Moisture in the bed translates directly into hydrogen in the deposit.
- Read the WPS before the first pass. Verify that current, voltage, wire feed, travel speed, stickout, heat input limits, preheat and interpass ranges are all within the qualified envelope, and confirm the wire–flux combination is the one actually qualified.
- Monitor stickout and travel speed. These are the easiest variables to control and the most often neglected. A drifting contact tip-to-work distance silently changes your current, and travel speed drift changes heat input.
- Maintain the contact tip, conduit and liner. A worn tip or a fouled liner causes feed surges, arc instability and stubbing.
- Calibrate the machine. Ammeter and voltmeter readings should be checked on a schedule so the numbers you record are real numbers.
- Run parameter trials on scrap when the joint, thickness or flux lot changes, rather than discovering the settings on the production weldment.
- Record heat input on the weld log for every pass or pass sequence, which is how you demonstrate compliance with the qualified WPS during audit.
- Check interpass temperature and keep preheat in place for thick or hydrogen-sensitive steel. As the AWS practical guide to submerged arc welding emphasizes, SAW’s high productivity does not exempt it from hydrogen control. For alloy and carbon steels that need careful filler selection, our alloy steel welding electrodes guide covers the surrounding decisions.
- Inspect after slag lift: check bead profile, look for undercut, lack of fusion indicators, worm tracks and centerline cracking before the next pass hides them.
- Train the operator, not just the machine. SAW looks hands-off until something drifts; the operator’s job is reading the bead, the sound of the arc and the flux bed, not holding a torch steady.
Quick Note: Keep a simple one-page shift log: flux lot, screening/drying date, wire lot, calibrated readings, stickout, travel speed, preheat and interpass values, and calculated heat input per pass. That single sheet resolves most audit questions before they become nonconformance reports.
The Four Mistakes That Cause Most Rework
| Mistake | What goes wrong | Correction |
|---|---|---|
| Reusing contaminated flux | Fines and slag chips cause arc wandering, inclusions and porosity | Screen and dry every reclaimed batch; discard contaminated lots |
| Ignoring travel speed | Too fast gives shallow penetration and cold lap; too slow overheats and widens the bead | Set and hold speed to the WPS; verify with bead profile |
| Mismatched flux and wire | Wrong basicity or alloy system leads to cracking, poor wetting or off-composition weld metal | Use only the qualified consumable pair, per AWS A5.17/A5.23 classification |
| Skipping preheat | Hydrogen cracking in thick or restrained joints, especially low-alloy steel | Measure, record and hold preheat and interpass temperatures |
Other recurring offenders: allowing stickout to drift as the conduit wears, running a flux bed that is too shallow so the arc breaks out, stopping without adequate crater fill, and failing to keep oil, moisture and mill scale out of the joint. The Fabricator’s troubleshooting notes on lack of fusion and porosity keep coming back to the same root causes: contaminated surfaces, wrong settings and mismatched consumables.
Productivity gains are real, but they only hold when the process is controlled. The reported productivity improvements from SAW come from shops that standardized setup, tracked parameters and trained operators to read the bead. If your shop is building that skill base from scratch, or expanding into other processes, a structured aluminium welding course shows how formal training compresses the learning curve, and sourcing consumables from a supplier you trust, such as our Airgas welding supplies overview, keeps flux and wire consistent lot to lot.
Reliable SAW is not complicated: keep the flux clean and dry, keep the consumables qualified, keep the variables inside their ranges, and keep the numbers written down. Do that consistently, and the hidden arc will keep handing you the same smooth, glowing bead, shift after shift.
Frequently Asked Questions
What is a submerged arc welding procedure?
A submerged arc welding (SAW) procedure is the documented set of parameters — electrode and flux classification, current type and range, voltage, wire feed speed, travel speed, stickout, joint design, preheat and inspection requirements — used to produce a sound weld. It is normally written as a Welding Procedure Specification (WPS) qualified under a code such as AWS D1.1 or ASME Section IX.
Why can’t you see the arc in SAW?
The arc is buried beneath a layer of granular fusible flux that melts and forms a protective liquid slag over the weld pool. This blanket shields the molten metal from air and blocks most of the arc’s ultraviolet light, which is why the process appears quiet and dark during operation.
What are typical SAW welding parameters?
For a single 1.2–1.6 mm wire, common settings are roughly 300–500 amps and 25–35 volts, with travel speed tuned to the plate thickness and joint. The exact values depend on flux type, wire diameter, electrode connection and required penetration, so they should always follow the qualified WPS.
What causes porosity in submerged arc welding?
The most common causes are damp or contaminated flux, oily or rusty base metal, and travel speed that is too fast for the parameters. Correcting flux storage and drying, cleaning the joint and adjusting parameters usually eliminates the porosity.
Can SAW be used out of position?
No, in practice SAW is limited to flat and horizontal positions because the granular flux cannot be held over the puddle against gravity. Out-of-position work is better handled by processes such as FCAW, GMAW or SMAW.
Which codes govern SAW procedure qualification?
Common codes include AWS D1.1 for structural steel, ASME Boiler and Pressure Vessel Code Section IX for pressure-retaining work, and API 1104 for pipeline girth welds. Each code defines the essential variables, test coupons and acceptance criteria for qualification.
What flux and wire combinations should be used?
Wire and flux must be matched as a qualified pair under AWS A5.17 (carbon steel) or A5.5 (low-alloy), because flux chemistry controls weld metal composition, bead shape and mechanical properties. Always verify the manufacturer’s recommended combinations and your WPS.
Is submerged arc welding safe for the welder?
SAW is relatively low in visible fume and UV exposure during normal operation, but hazards remain: flux dust, manganese and other fume components, electrical risk, burns and noise. OSHA advises ventilation, local exhaust and appropriate PPE for all welding operations.
How do you improve SAW productivity?
Productivity gains come from multi-wire or tandem setups, strip electrodes, optimized travel speed, automatic flux recovery and reuse, and tight parameter control that reduces defect rates and rework. Mechanized column-and-boom or gantry systems also keep parameters consistent over long seams.
Nicholas Harper
Editorial TeamCovering welding processes, equipment selection, and shop-floor practice for WeldZen — with every claim checked against the governing code.
Related Studies
Get The Next Benchmark First
New field studies and calibration sheets, once a week. No spam.