Flux Core Vertical Welding: Settings, Technique & Mistakes

The puddle is fine on the first inch, then quietly grows until it slides off the joint and leaves a fat, lumpy sausage with a huge crater at the bottom. Every fabricator has watched it happen, usually on the first vertical-up pass of the day when the machine still feels new. The frustrating part is that nothing about the job looks complicated: the wire is the same wire, the plate is the same plate, and the same settings that look perfect in flat work turn into a mess the moment gravity gets involved. This guide explains why vertical-up flux core behaves differently, exactly which settings to start from, and how to fix the six defects that show up most often on the plate.

Why Vertical-Up Flux Core Is a Skill of Its Own

Short answer: flux core vertical welding is simply FCAW done with the travel direction running upward, against gravity. The process itself does not change. What changes is what gravity does to the molten pool and to the slag floating on top of it. The rutile flux core inside the tubular wire decomposes into a stable, spray-like arc with real penetration, and that arc holds a puddle together far better than the short-circuiting or globular transfer you get from solid wire MIG. The result is deep penetration with none of the bouncing, spitting chaos that makes solid wire vertical work frustrating.

That is why a welder who runs a clean flat bead with a .045 flux core wire usually needs less practice to be productive going up than they would with stick. The arc is forgiving. Forgiving is not the same as easy.

Vertical Is a Position, and the Vertical Plane Has Two Directions

Welding has four basic positions: flat, horizontal, vertical and overhead. Vertical is not a direction of the electrode, it is the orientation of the joint. In the vertical plane there are two progressions: vertical-up, where you travel from the bottom of the joint toward the top, and vertical-down, where you start at the top and work down. Those terms describe travel direction, not electrode angle. You can hold the gun at any angle you like and still be welding vertical-down if you are moving downward.

The Physics You Are Actually Fighting

Gravity acts on two things at once: the molten metal and the slag layer. In flat and horizontal work the slag simply runs off or gets dragged aside. Held vertically, it stays on the face of the puddle and pushes down on it. Make the puddle too large, too hot or too liquid and the slag will not just flow over the side. It runs ahead of the arc, the puddle sags out of the joint, and you get the fat sausage bead with the crater at the bottom.

So the control variable in vertical-up is not wire feed speed alone. It is heat input and travel speed. A welder who thinks only in wire feed speed will fight gravity forever. A welder who watches travel speed and heat per inch will usually keep a small, tight, fast-moving puddle that the slag sits on without sliding off.

Is Flux Core Actually One of the Easiest Processes to Run Upward?

Honestly, yes. Machinery’s Handbook notes that all-position flux-cored electrodes used for vertical-up welding are simpler to operate than short-circuit GMAW or SMAW and are capable of greater weld quality. That is a real advantage, and it is why flux core is the default answer for vertical-up carbon steel in a lot of shops.

The catch is that the ease comes from the parameters, not from the process name. Start from the wire manufacturer’s data sheet rather than from memory or from whatever was left in the machine. The AWS Everyday Pocket Handbook for GMAW and FCAW is a free, citable reference for filler metal classifications, wire feed speeds and deposition data, and the AWS D1.1 structural welding code covers the code side of vertical work. The other half of the ease comes from cleaning. Flux core is tolerant of a little rust and mill scale, but trapped slag is not tolerated by anyone.

A Very Common Shop Scenario

A fabricator runs a 3/8 inch carbon steel butt joint flat all morning at 24 volts and 250 amps with .045 flux core, and it looks perfect. The same job gets set vertical for an install, the same wire and the same settings come out of the machine, and within four inches the bead is sagging, undercutting on the sidewall, and trapping slag. Nothing is wrong with the wire. The flat settings were simply never vertical settings, and the data sheet lists separate parameters for the position that was actually welded.

What This Guide Covers

The rest of this guide walks the process in order: how FCAW-S and FCAW-G actually differ, how to pick a wire and diameter for out-of-position work, then the operational core, which is polarity, wire feed speed, voltage and stickout. From there we cover torch angle, travel speed and weave patterns, then joint preparation and the root-to-cap pass strategy, which is where most vertical-up failures actually begin. The troubleshooting section works through the six defects that ruin vertical-up welds, followed by how flux core stacks up against stick and MIG, the safety and ventilation rules that apply, and a repeatable workflow you can use on your next job.

One small note on expectations. Vertical-up flux core rewards respect for the data sheet, the stickout, the size of the puddle and the cleaning between passes more than it rewards talent. Get those four right and the rest is practice. If you want a side-by-side on the alternative processes before you commit, our guide to why SMAW still wins in a lot of shops is worth a read, and all position rod and wire selection covers what the labels actually mean.

Flux Core Basics: FCAW-S vs FCAW-G and How the Wire Actually Works

Flux core arc welding is a hybrid of stick and MIG. The filler is a thin tube with granular flux packed inside it. An arc forms between the wire tip and the workpiece, melting tube wall, core and base metal in one pass. As the flux melts it blankets the puddle as slag, sealing molten metal off from the air, and it supplies elements that tie up oxygen and nitrogen in the weld metal. That protection is why the process exists, and why it handles vertical-up better than most welders expect.

Two Families, One Decisive Difference

The AWS filler metal classification system, documented in AWS A5.20 and the AWS Welding Handbook, splits flux core wires into two groups. S means self-shielded, G means gas-shielded. In vertical work, that single character matters more than anything else on the label.

Factor	FCAW-S (self-shielded)	FCAW-G (gas-shielded)
Shielding	Flux core decomposes and creates its own protective atmosphere. No gas cylinder required.	Needs an external shielding gas, typically 100% CO2 (E71T-1C) or a 75/25 argon-CO2 blend (E71T-1M).
Arc behaviour	Tighter, cooler arc with a compact puddle.	Hotter, slightly wider arc with a flatter, more fluid puddle and deeper penetration.
Slag	Heavy and thick, peels off in chunks as a distinct cap on the bead.	Light, thin and glassy. Can be nearly invisible on some settings.
Wind	Wind tolerant. The wire protects itself.	Not wind tolerant. Needs a shelter or a windy day will ruin the weld.
Surface tolerance	Handles rust, mill scale, oil and paint far better.	Wants reasonably clean steel for reliable results.

Gas-shielded wire of the same diameter runs hotter and slightly wider than self-shielded. That is a bonus in the flat position, where you want metal deposited fast. In vertical-up it is a complication, because a hotter, more fluid puddle held against gravity is exactly how you get a run. Self-shielded wire trades some heat for forgiveness, which is why many vertical-up procedures use it.

The Slag Is Working for You

Beginners treat slag as dirt to knock off. On a self-shielded wire running vertical-up it is closer to a live instrument. The slag forms a soft cap over the molten puddle, and that cap tells you how your travel speed and heat are matching up. Hobart Brothers says it directly: the slag indicates how fast you are able to travel while welding upward. A stiff, distinct rope sitting neatly on the bead means heat and speed are balanced. Slag that slumps ahead of the puddle means you are moving too fast for the puddle to support. A puddle running down the plate with slag following it in a molten mess means you are travelling too slow and adding too much heat.

The cost of that system is weight. Self-shielded wire makes noticeably more slag than gas-shielded, and that slag is abrasive. Expect more chipping between passes, more risk of trapping slag in a hot pass, and faster wear on nozzle, contact tip and liner. In return you get an arc that shrugs off wind, a puddle that stays put, and a free visual check on your own speed.

What “All-Position” Actually Tells You

When a wire label says all-position, the manufacturer has qualified it for flat, horizontal, vertical and overhead work within a stated parameter range. It is the label that tells you the wire can be run in a vertical plane at all, and it is why the E71T-1 class of rutile flux core wire is the default for vertical-up carbon steel in North America. The self-shielded version carries the 11 suffix (E71T-11), while the gas-shielded versions are E71T-1C for CO2 and E71T-1M for the 75/25 blend.

Position limits are not a suggestion. An all-position wire still has to be welded inside its rated amperage, voltage and travel speed windows, which is why a shop that cannot hit those numbers on a vertical pass still ends up with a bad bead. Our all-position welding rod guide breaks down how position designations work across processes.

Where FCAW Gets in the Way

Honesty about the process helps more than hype. Flux core throws more spatter than solid wire, much of it from slag breaking apart. It wants a stronger machine than a small MIG unit, because a tubular wire and a heavier arc chew through consumables, and the abrasive slag will eat a liner and a contact tip far faster than clean MIG wire does. Gas-shielded flux core is not a weather-or-not process: the Suggested Specification for FCAW of Carbon Steel states that FCAW-G welding should not be performed in a draught or wind unless the weld is protected by a shelter that reduces wind velocity to about 8 km/h (5 mph).

For that reason, self-shielded flux core is often the smarter pick on a site job, where the nearest gas bottle is a long walk away and the weather has other plans. If you are weighing it against stick, the case for stick welding is worth a look, since both share the no-gas advantage but differ a lot in deposition rate and control.

Once you choose self-shielded or gas-shielded, everything downstream changes: polarity, stickout length, voltage window and slag handling. That is the foundation for the settings in the next section.

Choosing the Right Wire and Diameter for Vertical-Up Work

Get the wire wrong and no amount of technique saves the pass. Get it right and vertical-up starts to feel forgiving. So begin with the decision that matters most: what is on the spool, and how thick is it?

Most North American shops running vertical-up on carbon steel land on the same classification. As Machinery’s Handbook notes, E71T-1 is the electrode most commonly used in the United States for vertical up welding on carbon steels. Pick up a spool at a US farm supply store or big-box welding counter, and that is almost certainly what is in your hands.

Decoding E71T-1 on the label

Once you know the shorthand, the AWS classification is easy to read.

• E means arc welding with an electrode.
• 71 is tensile strength in thousands of psi, so 70,000 psi minimum weld metal tensile strength. Deposited E71T-1 typically lands higher: about 84,000 to 93,000 psi with a 75% argon/25% CO2 blend, and roughly 84,000 to 87,000 psi with 100% CO2, depending on the shielding gas used.
• T means tubular wire, as opposed to solid rod or metal-cored (MC).
• -1 is the digit that matters: a general-purpose, all-position wire, which is what makes E71T-1 the default for vertical-up.

The trailing letter tells you how the wire is shielded. E71T-11 is self-shielded, the flux core decomposing to make its own protective atmosphere, so no gas cylinder is needed. E71T-1C and E71T-1M are gas-shielded, C for 100% CO2 and M for a 75/25 argon-CO2 blend. Same 70,000 psi class, same all-position capability. What changes is shielding, arc character, and machine setup, not the mechanical properties of the deposit.

Why all-position wire earns its keep out of position

Not every wire on the shelf says all position. Many are rated flat and horizontal only, and they say so on the data sheet. An all-position wire is formulated to hold a stable arc and a workable puddle when that puddle is climbing a vertical joint.

There is a second reason, straight from the Suggested Specification for FCAW of Carbon Steel: choosing the correct all-position electrode diameter allows the high current capability of the wire to be fully used to attain maximum deposition rate, and it minimizes weaving. That matters, because every weave in vertical-up is another chance for slag to run ahead of the puddle, the arc to wander off the sidewall, or undercut to appear.

Diameter is a vertical decision, not a thickness decision

This is where fabricators get burned. The instinct is a fat wire for thick plate because it fills faster, then the vertical pass runs away. Per the Weldcote E71T-1 technical data, smaller diameters, usually 1/16 inch (1.6 mm) and smaller, are used for welding in all positions, while larger diameters are used in the flat position and for horizontal fillets.

So the decision framework looks like this:

Diameter	Typical Plate / Fit-Up	Control in Vertical-Up	Machine Capacity	Spatter
0.035 in	1/8 in and under, light-gauge or repair work	Excellent. Small, easy-to-hold puddle	Often 130-150 A	Moderate
0.045 in	1/4 to 3/8 in carbon steel, general fabrication	Very good. The sweet spot for most vertical-up work	Roughly 150-250 A	Moderate
0.052 in	3/8 to 1/2 in plate, heavier joints	Good, but needs a steady hand	Needs 200 A or more	Heavier
1/16 in	1/2 in and up, structural work where deposition rate matters	Fair. All-position capable, widest puddle of the group	220 A and above	Heaviest

The trend is the point: bigger wire means bigger puddle, more heat, and faster-moving slag that resists your control. Small wire gives you a puddle you can watch and steer.

How fast you can go without weaving

One number from ESAB is worth keeping in mind when someone asks why they should not weave a fillet. ESAB reports that all-positional rutile flux cored wires can weld a 3/16 in (4 mm) throat fillet weld at welding speeds up to 7 ipm (18 cm/min) without weaving. That argues for stringer beads on vertical-up fillets: the travel speed you need for control and a narrower bead, without the defects weaving drags in.

Rules of thumb

• Match diameter to plate, not to the flat pass. Up to about 1/4 in, 0.035 or 0.045 in is comfortable. From 1/4 to 1/2 in, 0.045 or 0.052 in works. Above 1/2 in, go to 0.052 or 1/16 in and consider a split-pass strategy.
• Never exceed your machine. If the wire needs 220 A and your power source tops out at 200 A, you fight the machine all pass. Size the diameter to sit inside your welder’s amperage range, with headroom for CV voltage behavior at higher amperage.
• Split the pass on thick plate. Run the root with smaller wire, where penetration and puddle control matter most, then switch to larger wire for fill and cap, where deposition rate is the priority. One spool change, far less rework.
• Use the data sheet, not the spool label, for ranges. The classification tells you what the wire is, the manufacturer’s sheet tells you what it wants. When a vertical-up pass goes wrong, the gap between the two usually holds the answer.

Pick the wire from the data sheet, size it for the plate in front of you, and remember that out of position the smallest practical diameter usually gives you the most control. The Hobart guide to vertical-up with self-shielded flux core is a free reference for the process details, and your machine’s manual carries the amperage and stickout chart that matches your setup.

Machine Setup: Polarity, Wire Feed Speed, Voltage and Stickout

Get the setup wrong and nothing else matters. A vertical puddle that will not hold usually fails because the machine is still set for flat work. The fixes are small: re-pole, back off the feed, drop the volts, add stickout.

On a flat fillet the puddle sits on the plate and the arc pushes it along. In vertical-up it is molten metal clinging to a vertical wall under a slag blanket that wants to run downhill. Four controls matter: wire feed, voltage, travel speed and stickout. You are always dialing down compared with flat.

Polarity first, because getting it wrong wastes the whole pass

Polarity decides which way current flows, and for flux core it is a wire requirement, not a preference. Gas-shielded rutile wires in the E71T-1C and E71T-1M family run DCEP (electrode positive, torch lead positive, work lead negative). Self-shielded wires run the opposite way, and Hobart’s Fabshield is a DCEN wire per the company’s technical note on vertical-up self-shielded work.

First attempts fall apart when a welder swaps a gas-shielded crate for a self-shielded spool and the machine arcs short, spits and throws slag: the wire changed, the polarity did not. A polarity switch or studs you swap leads on means physically re-poling the box, not tapping a pendant button.

Vertical-up starting parameters for E71T-1 wire

These are the ESAB-derived vertical-up ranges published in the flux core arc welding settings reference. Start at the optimum column; treat the range as your window.

Wire diameter	Voltage	Amperage	Optimum
0.035 in	16–23 V	150 A	23 V / 150 A
0.045 in	22–26 V	150–250 A	25 V / 200 A
0.052 in	21–26 V	210 A	25 V / 210 A
1/16 in	21–27 V	220 A	25 V / 220 A

Optimum voltage is 25 V for every diameter from .045 up, and amperage climbs with diameter: .045 inch wants 200 A, 1/16 inch wants 220 A. If the machine cannot hold 220 A at 25 V without bogging down, the wire will not feed at the set speed and the arc wanders. A 200 A shop should step down to .045 or work 1/16 inch at the bottom of its range.

Self-shielded wire runs a different window. FCAW-S is not the same process with the gas turned off; the flux core makes its own shielding atmosphere, which changes the arc’s electrical characteristics and feel. Reference ranges for .045 inch self-shielded wire in vertical-up:

Wire	Voltage	Amperage	Wire feed	Stickout
0.045 in E71T-11 (self-shielded)	20–24 V	160–240 A	180–350 IPM	0.75–1.00 in

Compare with the gas-shielded .045 row: lower voltage ceiling, lower amperage ceiling, much longer stickout. Run a self-shielded wire at the same 25 V and 200 A and you are asking for heat and a wide arc, a fast way to grow a runaway puddle.

Stickout is the setting people forget

Stickout, or electrode extension, is the distance from the contact tip to the work. ESAB is explicit about the trade: shortening the stickout raises current and lengthens the arc, lengthening it lowers current and shortens the arc. Same kind of lever as voltage, pushing the opposite way.

Self-shielded wires run long stickout, roughly 3/4 to 1 inch, because the extension lets the flux core decompose and set up its own atmosphere before the arc reaches the puddle. Gas-shielded wires run short, roughly 1/2 to 3/4 inch, because a long extension pulls the arc off the work and steals shielding coverage. Stretch a gas-shielded wire out in a vertical position and you get oxidation, spatter and a wandering arc.

You will run more stickout than in flat work: the gun is angled so the arc points down into the joint and you can see the root, and that angle pushes the wire out. Set stickout for the angle you actually hold.

The “dial down 10 to 20 percent” rule

A qualified flat procedure should not run vertical on the flat numbers. GoWelding.org guidance on 3G vertical flux core recommends reducing wire feed speed and voltage by roughly 10 to 20 percent from flat settings. That is a field adjustment, not a substitute for a qualified vertical WPS under AWS D1.1, but it beats running flat values and wondering why the bead ran. Lowering both keeps the puddle controllable; still fat, add travel speed before you add heat.

Inductance and the CC/CV power source

On a constant voltage machine the inductance setting controls how fast current rises and falls around the arc. Turning it up makes the arc more stable and forgiving, but the machine then resists your inputs, and a vertical puddle needs fast correction. Turning it down makes the arc responsive and corrects a wandering pool quicker. Most shops settle on a low to mid setting for vertical-up FCAW.

On constant current you set amperage and the wire feed follows the melt rate you want; on constant voltage, wire feed sets amperage and voltage sets arc length. CC/CV boxes run stick and wire from one machine and add arc start, hot start and pre-flow for a repeatable first second. Confirm the box is in CV before welding wire; in CC it feels stiff and fights the small adjustments vertical-up requires. The airgas welding equipment guide covers the hardware inside the cabinet.

Torch Angle, Travel Speed and Weave Techniques

Vertical-up welding is mostly geometry. Once your machine settings are dialed in, what separates a clean bead from a lumpy one is where you point the torch, how fast you travel, and the pattern you lay the wire down in.

Forehand (Push) vs Backhand (Drag)

Forehand, or push, angles the torch back toward you so you push the arc into the joint as you travel upward. Backhand, or drag, trails the torch behind, pulling the arc along behind the puddle.

The Suggested Specification for FCAW of Carbon Steel spells out the tradeoff: forehand gives better visibility of the joint and a flatter puddle, while backhand delivers better penetration. The same document puts the torch angle at 10 to 20 degrees on either side of vertical.

For most vertical-up work, a 5 to 15 degree push angle is the better starting point, and GoWelding.org recommends the same figure for vertical flux-cored arc welding:

• You can see the root. Pushing gives you a line of sight into the joint so you can confirm the arc is landing where you think. Dragging hides the root behind your hand.
• The arc is driven into the joint. A push angle concentrates arc force downward into the root gap, helping on butt welds and vertical fillets where filler metal must penetrate, not sit on the surface.
• Control is easier. Backhand penetration comes with a narrower, harder-to-see puddle that tends to run.

Backhand still has its place on thick plate needing maximum penetration, or on a vertical fillet whose toes need extra drive. You are trading puddle control for depth.

Wire Placement and Stickout

The wire tip should sit in the leading third of the puddle, the position GoWelding.org recommends. That means the wire sits closer to the top of the weld you are building than to the bottom of the puddle you just laid, so heat lands in the joint ahead of the puddle and molten metal stays behind the arc.

Stickout matters too. For .045 and .052 inch wire, GoWelding.org recommends roughly 1/2 to 3/4 inch. Hold the contact tip a consistent distance from the work and point the wire angle at the leading edge of the puddle.

Weave Patterns and Heat Input

Three patterns cover most vertical-up work, and each puts a different amount of heat into the joint:

• Stringer beads. No weaving. The wire stays in one spot as you move straight up the joint. Lowest heat, narrowest bead.
• Split weave. The weave breaks into a V, with the split at the top of the puddle rather than the bottom.
• Triangle weave. A continuous V pattern, up one side, across, and down the other.

ESAB quantifies the difference: full width weaving lands at roughly 2.5 to 3.5 kJ/mm, split weave at 1.5 to 2.5 kJ/mm, and stringer beads at 1.0 to 1.5 kJ/mm. ESAB also recommends split weave and stringer bead techniques where optimal sub-zero weld metal toughness is required, such as offshore fabrication, because narrower, more controlled passes limit the heat left in the plate.

For most carbon steel vertical-up work, stringer beads or a tight split weave are the safest default. Full width weaving is aggressive: more heat, a slower arc, more time for the puddle to run. Use it when the section is heavy enough to absorb it and the WPS calls for it.

Travel Speed Is Your Puddle Control Tool

Typical vertical-up travel speeds run 4 to 7 IPM for a .045 inch root pass and 8 to 12 IPM for cap passes. Cap passes move faster because you are shaping the surface of an already solid weld rather than laying down new penetration.

The rule tying it together: in vertical-up, travel speed is your puddle control. Slower travel means more heat per inch and a bigger, faster-running puddle that may sag or drop off. Faster travel concentrates the heat, keeps the puddle small and fluid, and gives a bead with lower reinforcement.

Most vertical-up problems blamed on settings are really travel speed problems. When the puddle starts growing, the instinct is to add wire. Wrong instinct. Speed up the travel first.

Ergonomics: Your Body Is Part of the Process

Vertical-up is physically demanding in ways flat work is not.

• Stand so the fume plume rises past you. Offset your stance so the plume passes to one side. Ventilation is covered in the safety section of this guide, but body position alone cuts your exposure.
• Set your helmet angle for the work. Look slightly down the joint, not into the arc. If you cannot see the toe of the bead, you cannot fix it.
• Take rest breaks. Standing still in full FR gear while radiating heat is taxing, and a break every 20 to 30 minutes does more for consistency at the end of a pass than grinding through.
• Keep the free hand purposeful. A chipping hammer and wire brush within arm’s reach beat setting the torch down and hunting for them.

Check the Toe of the Weld

At the end of a vertical-up pass, look at the toe, where the bead meets the base metal at the sides. A sound bead has a tight, slightly concave toe that blends smoothly into the plate. A convex, bulged or undercut toe means the arc was wandering off the sidewall or running too long, and it shows up long before the puddle gives you trouble.

Fixing the toe at the end of a pass is far easier than fixing a run that already dropped onto the plate. Make the check part of the habit.

Related reading: if you ever move from carbon steel to aluminum with the same torch discipline, our guides to aluminum flux core welding and aluminum MIG welding cover the different shielding and heat control demands of those alloys.

Joint Preparation and the Root-to-Cap Pass Strategy

Most vertical-up flux core failures are not settings failures. They start on the bench, at fit-up. By then the joint has already decided how big your puddle wants to be, how the slag will flow, and whether you fight the metal or work with it.

Getting the Geometry Right

Four variables control a vertical-up groove: bevel angle, root opening, root face, and fit-up. Change one and the puddle changes size. A wide root opening sounds good but is not: more molten metal means more weight hanging on the root, more burn-through risk on the bottom pass, and more room for slag to fall in ahead of the puddle and get trapped. A tight opening makes a smaller, easier puddle, but too tight the wire cannot reach the sidewalls, giving you lack of fusion in the root.

Root face is the land you leave on the bottom piece. It keeps the pool from dropping straight through onto the root pass and gives you a bridge to land the root bead against. Typical groove preparations call for roughly 1/8 inch of root face on a 3/8 inch carbon steel butt joint.

The Code and the WPS Come First

Nothing outranks the welding procedure specification on your job. Under AWS D1.1, prequalified FCAW joint details for carbon steel include square butt with backing, single-V with backing, single and double-V grooves welded from both sides, and multi-pass fillet welds, most qualified for vertical-up. Two points bite in the field. The groove details cover only what the standard lists, so a joint that changes an essential variable needs a qualified WPS. And changing an essential variable, such as switching a pass from uphill to downhill in a vertical joint, requires requalification.

Prequalified procedures still carry preheat and interpass requirements, taken from D1.1 Table 5.11. For non-prequalified procedures, use the interpass temperature limits in the notes to Table 6.10. Carbon steel FCAW generally carries modest preheat at the thicknesses most shops handle vertically, but the number on your WPS governs.

The Pass Sequence, Step by Step

1. Root pass. Small puddle, short arc, travel speed slightly on the faster side. Hold a push angle of roughly 5 to 15 degrees so you can see into the joint and drive the arc into the root. The goal is sidewall fusion, not a big deposit.
2. Clean the root. Chip and brush. This is the step most people skip, and it is why the next step exists.
3. Hot pass. A quick, hot pass with higher heat and faster travel than the root. One job: burn out slag pockets trapped in the root toes and merge the toes cleanly.
4. Fill passes. Work upward in multiple passes, each cleaned before the next. Keep the weave narrow and the puddle small.
5. Cap pass. Slightly faster travel and a controlled, narrow weave to keep reinforcement low. Slowing down for a heavy cap in vertical-up is how you get a runaway bead.

Cleaning Between Passes Is Not Optional

GoWelding.org identifies trapped flux core slag as the single most common between-pass defect and the dominant cause of inclusions in vertical work. Knock off every visible slag island with a chipping hammer or the edge of a flat bar, paying attention to the root toes where slag hides in the pockets at each side of the bead, then run a wire brush over the toes and tie-ins to remove the last scale and film. Skipping the brush is why inclusions keep showing up in inspection.

Back Gouging the Underside

Before welding the underside of a groove, the root must be gouged or chipped back to sound metal so the second side can fuse into the first. The Suggested Specification for FCAW of Carbon Steel requires the root to be gouged or chipped to sound metal before depositing weld metal on the underside of a welding groove, unless the WPS states otherwise. Air carbon arc gouging is the usual method, and it works in vertical position when you keep the gouge moving and let each bead cool enough to chip out cleanly. A light back gouge also relieves restraint as the joint closes, which helps with distortion.

Tacking, Restraint and Distortion

Tack welds in a vertical joint are made in the same position and with the same process as the joint itself, so they are vertical-up welds that need cleaning too. Use short tacks, staggered on both sides rather than clustered, and chip them before the root pass. Skip tacks if the joint holds itself. Distortion shows up as the plate edges pulling toward the center of the weld, so clamp the plates to each other across the joint, back them with strongbacks, and weld in a balanced sequence. Never clamp hard enough to squeeze the root opening shut at the bottom.

Worked Example: 3/8 Inch Carbon Steel Vertical Butt Weld

A common shop job: 3/8 inch carbon steel butt joint, vertical-up, full penetration, single-V groove, 60 degree bevel both sides, 1/8 inch root face, 1/8 inch root opening.

Pass	Wire	Approximate settings	Notes
Root	0.045 in E71T-1	22 to 26 V, 150 to 250 A, 5 to 7 IPM travel	5 to 15 degree push, small puddle
Hot pass	0.045 in E71T-1	Higher heat, 7 to 8 IPM travel	Burns out slag pockets in the root toes
Fill	0.045 in E71T-1	Same as root, narrow weave	Clean fully between every pass
Cap	0.045 in E71T-1	8 to 12 IPM travel, 5 to 10 IPM wire feed	Faster travel keeps reinforcement low

Those numbers are a starting point, not a procedure. Set them on a coupon of the same thickness and adjust. If the root runs hot and sags, drop wire feed and voltage a notch and lift travel speed. If the root is cold with lack of fusion at the sidewalls, add heat, slow down slightly, and check the root opening first. If you are still struggling, the problem is probably fit-up rather than settings. Remember the sequence: correct geometry, clean joint, sound tacks, a small root puddle, a hot pass to clear trapped slag, fill passes cleaned between every one, and a cap run with a controlled weave.

The Six Defects That Ruin Vertical-Up Flux Core Welds and How to Fix Them

Most vertical-up flux core problems are not exotic. They are six recurring defects that show up on every shop floor, and once you can read them in the bead, the fix is usually one setting change and one habit change. Diagnose before you dial. Change five parameters at once and you lose track of which one helped.

Symptom	Likely cause	Corrective action
Sagging, running bead, crater at the bottom	Wire feed too high, voltage too high, travel speed too slow	Lower wire feed, trim voltage, speed up travel, hold a smaller puddle
Undercut along the toe	Arc too long, travel speed too high, wire wandering off the sidewall	Shorten stickout, reset torch angle, pause at the toes
Cold starts, dull gray bead, stacked look	Arc too short (stubbing), wire feed too slow, puddle left behind	Add heat slightly, keep moving, re-establish the puddle
Porosity, wormholes, pinholes	Rust, oil, moisture, wind stripping shielding, trapped slag	Clean to bright metal, shelter the joint, verify gas flow and dew point
Heavy spatter, rough bead	Voltage too high, wrong polarity, long arc, abrasive slag	Confirm polarity, trim stickout, lower voltage, expect some spatter
Slag trapped in the bead or under the cap	Weave too wide, wire not in the leading third, slag running ahead	Narrow the weave, hold the wire forward, clean between passes
Nozzle and liner blocking up fast	Slag spatter, long stickout, contact tip too close to the work	Clean the cup, check the tip, replace liner and tip on schedule

1. Sagging and Running Welds

The puddle grows, runs down the joint and leaves a lumpy sausage with a crater at the bottom. Three causes drive nearly every case: too much wire feed, too much voltage, and travel speed too slow for the heat you are putting in. Fix them in that order. Turn the wire feeder down first, since wire feed sets the deposition rate and therefore the size of the puddle, trim voltage next, then speed up travel. Aim for a puddle that sits just ahead of the arc and moves with you rather than away from you.

2. Undercut

Undercut is a groove along the toe where the arc strayed off the sidewall. It appears when the arc gets long, when travel speed is too high, or when the wire wanders off the sidewall as the puddle shrinks. Bring the stickout in to shorten the arc, reset your torch angle so the wire is aimed into the joint, and pause briefly at each toe so the fillet ties in. ESAB’s FCAW data warns that an arc length that is too long gives insufficient penetration and creates a real risk of slag traps and burnback to the contact tip, so treat a wandering arc as a geometry problem, not just a heat problem.

3. Lack of Fusion and Cold Starts

A cold start happens when you stop, the puddle freezes, and you restart on solid metal. Related symptoms are a dull, glassy bead and a narrow stacked look with no fusion at the toes. The causes are an arc that is too short, which makes the wire stub, a wire feed too slow for the current you are running, and a puddle that gets walked away from. Add a little heat, keep moving, and re-establish the puddle before you return to your normal travel speed.

4. Porosity and Wormholes

Porosity usually starts before the arc is struck. OSHA requires that loose or thick scale, rust, moisture, grease or other foreign material which would prevent proper welding or produce objectionable fumes be removed, so clean to bright metal rather than to “mostly clean” (OSHA welding fact sheet). On gas-shielded wire, wind stripping the shielding is the classic culprit. Self-shielded wire carries its own shielding and does not care about wind, which is why many fabricators keep it in service year round; for a stick comparison, why SMAW still wins on windy sites is a handy reference. On self-shielded wire the usual culprit is trapped slag from a poorly cleaned previous pass. Check gas flow, and note that the Suggested Specification for FCAW of Carbon Steel requires shielding gas with a dew point of -40 °C (-40 °F) or lower. A regulator and flowmeter check is worth doing before blaming the wire; our welding gas equipment guide covers the setup.

5. Excessive Spatter

Spatter comes from four places: too much voltage, wrong polarity, an arc that is too long, and the abrasive slag itself. Verify polarity before anything else, because a wire running on the wrong setting will spit the moment the arc strikes. Trim the stickout and lower the voltage a notch. Expect some spatter regardless. Hobart notes that a self-shielded wire such as Fabshield XLR-8 has a thicker slag system, which is why proper torch positioning matters so much when welding vertical up with self-shielded flux-cored wire. Switch consumables often? Recheck the classification and polarity every time, not once per year.

6. Slag Inclusion

Trapped slag is the defect that most often traces back to weaving and cleaning rather than to amperage. It shows up when the weave is too wide, when the wire is not held in the leading third of the puddle, or when slag runs ahead of the weld pool. ESAB makes the same point from the other direction, advising a backhand angle because a forehand weld can deliver reasonable appearance but often poor penetration, with slag running ahead of the pool. Keep the weave narrow, hold the wire forward, and clean every single pass.

7. Slag Spatter Damage to the Gun

Flux core slag is abrasive, and it will eat your gun if you let it. Cratered nozzles, burned contact tips and grooved liners all point to a long stickout and a tip sitting too close to the work. Keep the nozzle extended so it shields the gas cup, check the tip and cup often, and plan on liners wearing out faster on flux core than on stick or MIG.

When the Problem Is Not the Settings

Sometimes the dial is not the culprit. If a vertical bead still sags at the lowest wire feed and voltage you can achieve, the root opening is too wide or the fit-up is pushing the plates apart. If wire feed fluctuates, look at the feeder rollers, drive cable and gun liner rather than the machine dial. If voltage dips when you pull the trigger, check the power source, the ground clamp and the lead length. And whatever the symptom, the qualified procedure on the job outranks any table here, including the code requirements in AWS D1.1.

Flux Core vs Stick vs MIG for Vertical-Up Welding

If you remember one number from this comparison, make it this one. Machinery’s Handbook reports that vertical-up welds made with all-position flux-cored electrodes reach deposition rates of 10 to 14 pounds per hour. Vertical-up welding with GMAW (MIG) or SMAW (stick) may deposit an average of just 2 to 4 pounds per hour, and pulsed GMAW sits in between at 3 to 6 pounds per hour.

That is not a marginal difference. On a 40-pound fillet, flux core finishes the vertical in a fraction of the time the other two processes need, and it matters far less whether the wind is blowing.

Translating Deposition Rate Into Actual Clock Time

Let’s make that real: a structural vertical fillet with a 5/16-inch leg, roughly 6.25 inches of weld metal per linear inch of joint, and a total joint length of 20 feet.

• Weld metal needed: about 125 pounds
• At 12 lb/h (flux core): roughly 10.4 hours of arc time
• At 4 lb/h (stick or MIG): roughly 31 hours
• At 4.5 lb/h (pulsed MIG): roughly 28 hours

Add the interpass cleaning, chipping, and grinder time that eats 30 to 50 percent of any vertical clock, and the gap becomes a full shift on a single 20-foot joint. That is why fabricators reach for flux core on vertical-up structural work even when the wire price stings.

Where Flux Core Wins Outright

Machinery’s Handbook also lists where a flux-cored electrode is the better choice, and it lines up neatly with vertical work:

• Base metal surfaces contaminated with mill scale, rust, oil, or paint
• Fillet weld sizes larger than 3/8 in (9.6 mm)
• Welding in the vertical-up or overhead positions

Those are the jobs that frustrate welders most. Flux core tolerates dirty joints because the core flux does much of the cleaning as it decomposes, where solid wire MIG wants bright metal and will punch porosity into a scaled surface. Grind a vertical joint for twenty minutes and the rust is still there; flux core lets you skip that step.

The handbook also notes that keeping the weave to a minimum reduces the skill level required and improves the chances of consistent sidewall fusion with minimum porosity. On a vertical-up, less weaving means a smaller, more controllable puddle, the point of every torch-angle and travel-speed discussion in this guide.

Flux Core vs Stick (SMAW)

Stick welding has real virtues, and there are plenty of good reasons to still choose SMAW. Electrodes cost almost nothing, no shielding gas to buy, no wire feeder to maintain, and all you need is a stick welder. On a field repair with a generator and a toolbox, that simplicity beats anything flux core offers.

But on vertical-up carbon steel, stick is slow and messy. It needs more passes to build thick plate, throws more spatter, and molten slag drips down the joint and onto the welder. Hobart Brothers puts it plainly when discussing the Fabshield XLR-8: self-shielded flux core vertical up is a slow progression, but still faster and more efficient than a stick electrode. Stick makes you chase the puddle with each bead; flux core climbs the joint in fewer passes.

Process	Vertical-up deposition	What it costs you
FCAW, all-position wire	10 to 14 lb/h	Spatter, premium wire, gas for FCAW-G, a CV power source with real amperage, liner wear from abrasive slag
Pulsed GMAW	3 to 6 lb/h	Clean metal and steady shielding, plus a pulse-capable machine
GMAW or SMAW	2 to 4 lb/h	More passes, a slower climb, more clean-up

Solid wire MIG is harder to run upward: you are stuck with short-circuiting or globular transfer, so the puddle is large and quick rather than the tight, forgiving arc a rutile core gives you.

Outdoors, FCAW-G will not tolerate wind. SMAW and FCAW-S will, and that alone can pick your process on a site with a breeze.

AWS D1.1 sets a maximum prequalified size for a single-pass fillet weld in the vertical position and constrains how vertical passes progress. With a prequalified procedure, passes generally must run upward, with narrow exceptions: small weld sizes, undercut repair, and FCAW-specific allowances.

Safety, Ventilation and the Regulatory Side of Vertical-Up Welding

With flux core, vertical-up welding puts you square in the fume column. Unlike flat passes, the plume rises straight up the joint, placing your head directly over it. Add radiant heat, spatter, and hot slag to the equation, and the exposure is far higher than the job looks.

The Fume Problem

Welding fume is the airborne mix of metal oxide particles created when the arc vaporizes the base metal, filler metal, and any coating or mill scale. Elements such as zinc, chromium, iron oxide, manganese, and copper form fine particles that follow the rising column of hot air. Self-shielded flux core compounds this, as its decomposing flux and slag release their own particulates and gases. For welders indoors with poor extraction, this can lead to the classic flu-like symptoms of metal fume fever. To better understand how shielding affects fume generation, it’s worth reading about how gas-shielded welding works.

General and Local Ventilation

The baseline for most shops comes from OSHA’s Subpart Q ventilation standards. A few rules are worth committing to memory:

• Minimum rate: At least 2,000 cubic feet per minute (cfm) of mechanical ventilation per welder.
• When required: In spaces under 10,000 cubic feet per welder, ceilings under 16 feet, or areas where partitions block cross ventilation.
• Alternatives: Local exhaust ventilation or airline respirators for each welder.
• Scaling: Higher arc voltage and larger electrode diameters increase fume generation, calling for more air.

For day-to-day work, local exhaust is almost always the superior option. It captures fume at the source, rather than diluting an entire room. OSHA requires a control velocity of 100 linear feet per minute at the welding zone, and the airflow needed increases dramatically as distance from the arc grows.

Distance from arc	Minimum hood airflow
4 to 6 inches	150 CFM
6 to 8 inches	275 CFM
10 to 12 inches	600 CFM

For vertical-up, the easiest fix is positioning the hood below and to the side of the arc to catch the rising column.

Confined Spaces and Chromium

Welding inside a tank, vessel, hopper, or pipe triggers both Subpart Q and 29 CFR 1910.134. A permit, atmospheric testing, ventilation, and an outside attendant with a retrieval system are all required. If local exhaust isn’t practical, supplied-air airline respirators with escape bottles must be used. Under no circumstances is oxygen to be used for ventilation. It’s also wise to check filler metal labels, such as those found for all-position welding rods, as they can warn against use without adequate ventilation.

Working with stainless, chrome-moly, galvanized, or plated materials can generate hexavalent chromium, a known carcinogen. OSHA’s permissible exposure limit is 5 micrograms per cubic meter as an 8-hour time-weighted average, with an action level of 2.5 micrograms per cubic meter. NIOSH recommends reducing all welding emissions to the lowest feasible concentration using engineering controls. For practical guidance, see OSHA’s small entity compliance guide and the OSHA welding fact sheet.

Heat, Burns, and PPE

Vertical-up is physically draining. Restricted helmet vision, raised shoulders, and back-radiated heat all build towards heat stress. Hydrate before you’re thirsty, take breaks in the shade, and watch for headache, dizziness, confusion, or a lack of sweating. Any of these are medical emergencies.

Burns are just as common. Slag from passes above can drip down onto gloves, boots, or forearms, and molten slag stuck to cotton can cause a nasty second-degree burn. Keep sleeves down, use heat-rated leather gloves, and never look down the joint while chipping. On some jobs, a stick electrode can be a lighter option for fume and slag exposure.

At a minimum, your PPE should consist of flame-resistant clothing, heatrated leather gloves, safety boots with a metatarsal guard, an auto-darkening helmet with the correct shade for your amperage, and safety glasses or a face shield worn under the helmet for grinding and chipping.

For fume, engineering controls must always come first and respiratory protection second. A particulate respirator or an airline respirator can offer added protection, but the goal is to capture the plume before it ever reaches your helmet. To evaluate your setup, take a look at our welding equipment guide for extraction and PPE options.

A Practical Workflow, Field Tips and Conclusion

Everything above collapses into one repeatable sequence. Run it without improvising and you will beat the welder still dialing in beside you. Order matters more than any step: vertical-up flux core fails early and visibly when the sequence gets shuffled.

The Step-by-Step Sequence

1. Read the WPS and the code first. If the job falls under AWS D1.1, the qualified procedure beats preference. Note the qualified positions, thicknesses, diameters and pass direction. D1.1 treats certain changes as essential variables needing requalification: a position not covered by its tables, a change of diameter or thickness, and switching a vertical pass from uphill to downhill or the reverse.
2. Confirm wire, diameter, polarity and stickout. Self-shielded or gas-shielded? Which diameter is qualified? Gas-shielded rutile wires generally run DCEP; most self-shielded wires run DCEN. Set stickout at the manufacturer figure: roughly 3/4 to 1 inch for self-shielded, 1/2 to 3/4 inch for gas-shielded.
3. Set the machine from the data sheet, then dial down. Use the wire maker’s vertical-up ranges or the AWS Everyday Pocket Handbook for GMAW and FCAW, which lists wire feed speeds, melting rates and deposition rates for carbon and stainless consumables. Then cut amperage and voltage from your flat numbers by 10 to 20 percent.
4. Run a coupon and prep the joint. A few minutes on scrap of the same thickness tells you what the machine does today, so adjust there, not on the joint. Grind back to bright metal, and tack to the root opening you intend to weld: misplaced tacks are a leading cause of burn-through.
5. Run a small root puddle, then clean between passes. Hold a 5 to 15 degree push angle and keep the wire in the leading third. Read the slag as your speed gauge: with self-shielded wire it shows how fast you can safely travel, as Hobart Brothers explains. Chip the slag, brush the toes and tie-ins, then run a hot pass to burn out slag pockets in the root toes. Trapped flux core slag is the most common between-pass defect, per GoWelding.org, so never pass over uncleaned slag.
6. Cap, cool and inspect. Run the cap slightly faster with a controlled weave to hold reinforcement low. D1.1 requires the weldment to cool to ambient temperature without external quench media, since rapid temperature changes cause problems. Finish with visual inspection to the acceptance criteria, then remove slag and spatter.

Troubleshooting Recap

Symptom	Likely Cause	Corrective Action
Sagging bead	Too much heat, fast feed, slow travel	Cut feed and voltage, speed travel, smaller puddle
Undercut	Long arc, excess speed, wire off sidewall	Shorten stickout, fix angle, pause at toes
Lack of fusion	Arc too short, feed too slow, puddle walked	Add heat, re-establish the puddle
Porosity	Damp metal, wind, trapped slag	Clean to bright metal, shield joint, check gas
Spatter	High voltage, wrong polarity, long arc	Trim stickout, check polarity, shorten arc
Slag inclusion	Wide weave, slag ahead of puddle, poor cleaning	Narrow weave, wire in leading third, clean between passes
Nozzle and liner wear	Abrasive self-shielded slag	Extend nozzle, swap liner and tip

If a defect will not respond to setting changes, look at fit-up, restraint and the machine instead. Most stubborn vertical-up problems live in the joint, not the dial.

Field Tips Worth Memorizing

• On gas-shielded wire, the Suggested Specification for FCAW of Carbon Steel wants the contact tube recessed in the cup nozzle by at least 6 mm (1/4 in).
• Set the spool and machine on the correct side so the wire feeds straight in, and back off feeder tension if the wire bird-nests.
• Start every job with a fresh liner and contact tip. A kink in the liner looks exactly like a wire feed problem.
• Slag off overhead passes burns clothing and skin, so assume every chip you knock loose is live.
• Keep the nozzle extended so the gas cup is shielded from spatter, and keep a chipping hammer and wire brush within arm’s reach.
• Drink water. Vertical-up in full gear is punishing, and fatigue shows up in the last hour of the shift.

For shop-floor context on the processes you might be weighing instead, see why stick welding still wins in certain situations and how to pick all-position welding rod.

How to Practice

Set the machine once on scrap of the real thickness, run three or four passes, and film them with your phone. The footage is uncomfortable and extremely useful: you will spot the drift off vertical, the stickout creeping longer as you tire, and the exact moment the puddle started running. Grind one pass and study the profile. A good vertical-up bead has a slightly convex face, even width, a toe blending into the base metal and no root reinforcement to trim. High reinforcement means travel speed was too slow; a flat, wide bead with rounded edges means too hot or too fast.

Conclusion

Vertical-up flux core is not a mystery process. It differs from flat work because gravity acts on the molten pool and the slag, so you manage heat input and travel speed instead of just pushing wire. Respect the data sheet, the stickout, the small puddle and the cleaning between passes, because trapped slag causes more downstream rework than anything else you will get wrong.

The biggest gain most fabricators will ever get comes from two habits: reading the manufacturer’s data sheet instead of guessing, and welding the first inch slowly and deliberately. That first inch tells you what the puddle wants to do.

Key Takeaways

• Vertical-up flux core is FCAW with upward travel, and the rutile core makes it forgiving.
• Know FCAW-S vs FCAW-G: it decides polarity, stickout, shielding and wind tolerance.
• Dial down 10 to 20 percent from flat settings, then tune from the puddle.
• Travel speed is puddle control: slower means more heat per inch.
• Clean every pass. Trapped slag causes most inclusions and porosity.
• On self-shielded wire, the slag shows how fast you can travel.

If you take one action from this article, make it the next one: grab a scrap coupon of the exact plate you are about to weld, pull the vertical-up numbers for your wire off the manufacturer’s sheet, and run a slow, deliberate first inch. All-position wire and rod is where to start if you are still choosing a consumable. Once the coupon looks right, the joint on the clock will look right too.

Frequently Asked Questions

What is the best flux core wire for vertical up welding?

For carbon steel in the vertical-up position, an all-position E71T-1 class wire is the standard choice, and smaller diameters such as 0.035 or 0.045 inch are easier to control out of position than 1/16 inch. Choose self-shielded (FCAW-S) for field work where wind or gas supply is a problem, and gas-shielded (FCAW-G) for clean shop work where a slag-free surface and easier slag control matter. Always confirm the wire you choose is rated for vertical-up on the manufacturer’s data sheet.

What voltage and wire feed speed should I use for vertical up flux core?

Manufacturer-derived starting ranges for 0.045 inch E71T-1 in vertical-up are roughly 22–26 volts and 150–250 amps, with an optimum near 25 volts and 200 amps. Self-shielded wire of the same diameter typically runs a touch cooler, around 20–24 volts and 180–350 IPM of wire feed. Treat all of these as starting points only, because your machine, stickout, plate thickness and qualified WPS will shift them.

Should I weld vertical up or vertical down with flux core?

Vertical-up is the standard and usually the only prequalified progression for structural work under AWS D1.1, since upward travel gives better fusion and slag control. Downward vertical progression is generally not permitted on a prequalified WPS except in the limited cases the code lists, such as small weld sizes, undercut repair, or where the Engineer approves it in writing. If you need downhill capability, it has to be qualified and documented.

Do I need shielding gas for flux core vertical welding?

Only if you are running a gas-shielded wire. Self-shielded flux core produces its own shielding atmosphere from the flux core and needs no gas cylinder, which makes it the better choice on windy outdoor sites. Gas-shielded flux core needs 100% CO2 or a 75/25 argon-CO2 blend flowing at roughly 35 to 45 CFH, and it must be sheltered from wind to protect the puddle.

What polarity should I use for flux core?

It depends on the wire, not on the position. Gas-shielded rutile wires such as E71T-1 are generally run DCEP (electrode positive), while most self-shielded wires such as Hobart Fabshield are run DCEN. Because the polarity differs by wire family, switch the leads before you start rather than discovering it when the bead looks wrong.

How do I stop a vertical up flux core weld from sagging?

Lower the wire feed speed and voltage, speed up your travel slightly, and keep the puddle small by holding the wire in the leading third of the pool rather than dragging it through. Also check that you are not over-feeding wire for the diameter, and confirm the stickout is not too short, since a short stickout raises current and lengthens the arc.

Why is my vertical up flux core weld undercutting at the toes?

Undercut at the toes almost always means the arc is too long or the puddle is drifting off the sidewall. Shorten your stickout, tighten the torch angle so the arc is directed into the corner, and slow your travel slightly at the toes before moving quickly through the center. Dirty or oxidized toes that are not ground clean can also undercut, so clean to bright metal before tacking.

Is flux core faster than stick for vertical up welding?

Generally yes. All-position flux cored electrodes are reported to achieve vertical-up deposition rates of about 10 to 14 lb/h, compared with roughly 2 to 4 lb/h for GMAW or SMAW. In practice that means fewer passes, less grinding and less time standing on a step ladder, though the wire and machine cost more than a stick electrode.

What stickout should I use for vertical up flux core?

Plan for more stickout than you use in flat position because the gun is angled upward. Self-shielded wires typically run about 0.75 to 1.00 inch of stickout, while gas-shielded wires usually need closer to 0.50 to 0.75 inch so the gas cup can still protect the puddle. Measure it rather than guessing, because stickout drives both current and arc length.

Can I get certified for 3G vertical up in flux core?

Yes. The 3G plate test in AWS D1.1 is a vertical-up groove weld test, and FCAW is one of the processes commonly used for it. The coupon is typically 3/8 inch plate with a single-V groove, and the welder must complete all passes in the upward progression. Passing a bend test on that coupon generally qualifies you for vertical-up groove welds within the thickness range the code allows for that test plate.