Shielding Gas Flow Rates & How to Fix Weld Porosity

Set the Right Flow Rate First: CFH by Process, Cup, and Environment

Flow rate is measured in CFH (cubic feet per hour), and the correct number is a range, not a single magic value. It depends on your process, the nozzle or cup size, the shielding gas, and the environment you are welding in. The most common mistake welders make is treating the flowmeter like a volume knob and turning it up whenever a weld looks bad. That instinct causes as many problems as it solves.

The governing principle is coverage without turbulence. You need enough gas to blanket the molten pool and the hot solidifying metal behind the arc, but not so much that the stream tumbles and pulls in room air. Larger cups and nozzles need more CFH because they have more cross-sectional area to fill; a #8 TIG cup simply cannot be shielded at the same flow as a #5. For GTAW, a good starting rule is roughly two to three times the cup number in CFH for argon (a #8 cup lands around 16 to 20 CFH), and a gas lens lets you run at the low end of that range while extending stickout.

Use the table below as a starting point, then verify with the weld itself. A clean, bright, well-shielded bead with no pinholes tells you the flow is right far more reliably than any number on the gauge. If you are welding reactive or thick material, or running helium-rich mixes for aluminum, expect to sit at the higher end because helium is lighter than air and disperses faster.

• Match the cup to the job: A wider cup or nozzle shields a bigger pool but demands proportionally more CFH; do not run a large cup at small-cup flow.
• Verify at the puddle, not the gauge: The flowmeter reads at the regulator; what matters is coverage at the arc. Trust a clean bead over a number.
• Standards reference: AWS D1.1 Structural Welding Code holds you responsible for sound, pore-free welds; there is no shortcut flow number that guarantees that if the setup is wrong.

Why Too-High Flow Causes Porosity: Turbulence and Aspiration

This is the counterintuitive truth that surprises most self-taught welders: turning the gas up past a certain point makes porosity worse, not better. Shielding gas needs to exit the nozzle in smooth, laminar flow so it lays down a stable protective blanket over the weld. Above a critical velocity, the stream becomes turbulent, tumbling and rolling as it leaves the cup. Turbulent gas does not shield; it churns the atmosphere right into your weld pool.

There is a second, sneakier mechanism called aspiration, or the Venturi effect. A high-velocity jet of gas creates a low-pressure zone around itself, and that low pressure literally sucks the surrounding room air into the shielding envelope. So the very gas you added to protect the weld ends up drawing in the oxygen, nitrogen, and moisture that cause porosity. On TIG especially, blasting 40 CFH through a small cup guarantees this, which is why experienced hands run far less gas than beginners expect.

The practical takeaway: if you already have adequate flow and you are still getting porosity, adding more gas is almost always the wrong direction. Back the flow down to the recommended range for your cup size and fix the actual cause, whether that is a draft, a leak, or dirty metal. When in doubt, err toward the lower end of the range and only increase if the puddle is visibly turning gray, sooty, or porous from too little coverage.

• Turbulence threshold: Most standard MIG nozzles turn turbulent somewhere north of 45-50 CFH; small TIG cups turn much sooner. Staying in-range keeps the flow laminar.
• Aspiration: A fast gas jet pulls room air in by the Venturi effect, so excess flow imports the exact contaminants you are trying to exclude.
• Symptom overlap: Too-low and too-high flow produce nearly identical scattered pinholes, which is why welders chase this problem for hours turning the knob the wrong way.

Drafts, Wind, and Windscreens

A shielding gas envelope is fragile. It does not take a storm to blow it away; a gentle cross-breeze from a shop fan, an open bay door, or an HVAC vent is enough to strip coverage off the pool and let air in. The widely used field limit is roughly 5 mph of air movement for GMAW, FCAW-G, and GTAW. Above that, you cannot reliably shield the weld no matter how much gas you push, and pushing more only makes the turbulence problem worse.

The correct fix is to block the wind, not to fight it with flow. Erect a welding screen, close the door, shut off the fan near the arc, or build a simple cardboard or sheet-metal windbreak around the joint. On field and structural jobsites across the US, fabricators routinely tent off weld areas for exactly this reason. If you genuinely cannot control the wind, that is the situation self-shielded flux-cored wire (FCAW-S, AWS A5.20 electrodes like E71T-11) was built for; it carries its own shielding in the flux and tolerates drafts that would destroy a gas-shielded weld.

One caution when you finally get out of the wind: do not forget to bring the flow back down. Welders who bumped the CFH up outdoors often leave it there when they move back inside, and then the excess flow starts causing porosity through turbulence and aspiration. Reset to your in-range value for the conditions you are actually welding in.

• 5 mph rule: Air movement above about 5 mph overwhelms a gas shield. Measure with a cheap anemometer or the back of a wet hand if you are unsure.
• Block, do not blast: Windscreens, closed doors, and killed fans fix drafts. More CFH does not, and adds turbulence.
• Switch processes when needed: For genuinely windy field work, self-shielded FCAW-S removes the shielding gas problem entirely.

Leak Checks: Hoses, O-Rings, and the Silent Porosity Source

A slow leak anywhere between the cylinder and the nozzle is one of the most maddening causes of porosity because the gauge still reads a normal flow and everything looks fine. There are two ways a leak hurts you. The obvious one is gas escaping so less reaches the weld. The subtler one is aspiration in reverse: because the gas stream downstream of a pinhole is moving fast and at low pressure, a leaky hose or fitting can actually draw room air INTO the line, so contaminated gas arrives at the puddle even though plenty of gas is flowing.

Cheap clear vinyl gas hose is a frequent culprit. It is permeable to moisture and air over time, cracks at the ferrules, and goes brittle. Replace it with braided or barrier-grade shielding gas hose. Check every connection: the CGA-580 cylinder fitting for inert gases, the regulator-to-flowmeter joint, the hose barbs, the quick-connect, and the torch or gun connections. The O-rings on TIG torch backcaps, gas lens collet bodies, and MIG gun necks harden and shrink with heat and age; a $2 O-ring kit fixes porosity that has stumped welders for weeks.

To leak-check properly, brush a leak-detection solution (or plain soapy water, never oil near oxygen equipment) on every joint with the gas flowing and watch for bubbles. For a whole-system check, pressurize the line, close the cylinder valve, and watch the high-pressure gauge: if it bleeds down with the torch trigger released, you have a leak to hunt. Also confirm the gun or torch gas solenoid is snapping fully closed and open, because a sticking solenoid gives intermittent, hard-to-diagnose porosity.

• Ditch vinyl hose: Clear vinyl line is permeable and cracks; braided or barrier hose is a cheap, permanent upgrade that ends a whole class of porosity.
• O-rings first: Backcap, collet-body, and gun-neck O-rings are the most overlooked leak point. A $2-$8 kit is the best-value fix in the shop.
• Soapy water, not oil: Use soap solution or a rated leak detector to find bubbles; never put oil or grease near oxygen or high-pressure fittings (OSHA 1910.253).
• Leak-down test: Pressurize, close the cylinder, and watch the gauge. A dropping needle with the trigger off means a leak downstream.

Regulator vs. Flowmeter: Reading Your Gas Correctly

There are two common ways to set gas, and confusing them leads to wildly wrong flow. A flow gauge (also called a flow regulator or pressure-compensated gauge) reads CFH on a dial by holding a fixed downstream pressure across a fixed orifice. A flowmeter uses a tapered glass or plastic tube with a floating ball or bearing; the height of the ball indicates flow. Both are calibrated for a specific gas at a specific outlet pressure, and that calibration is the trap.

A flowmeter or flow gauge marked for argon will read incorrectly on CO2 or on an argon/CO2 blend because those gases have different densities. The good news is that C25 and other argon-rich blends read close enough to a pure-argon scale for practical shop work, but pure CO2 reads noticeably off. Read a ball float at its center (or per the tube's instructions, some read the top edge), keep the flowmeter mounted vertical or it lies to you, and remember the meter shows flow at the regulator, not at the nozzle several feet of hose and a solenoid away.

One more setup gotcha: the pre-flow and post-flow surge. When you pull the trigger, the pressure that built up in the hose dumps out as a brief high-flow blast, then settles to your set value. That start surge can blow turbulence into the very beginning of the weld, which is why so many porosity spots and starting pinholes appear right at the tie-in. Setting a short pre-flow (on machines that offer it) and not pulling the trigger a foot away from the joint both help. Never set flow by the burst you hear; set it by the steady-state reading.

• Gas-specific scales: An argon-calibrated meter misreads on pure CO2. Argon-rich blends like C25 are close enough for most shop work.
• Mount it vertical: A tube flowmeter tilted off vertical gives a false reading. Keep it plumb.
• Mind the start surge: The hose dumps a high-flow burst on trigger pull, causing start porosity. Set flow by the steady reading, not the burst.

Contamination: Mill Scale, Moisture, Oil, and Hydrogen

Even with perfect gas coverage, dirty base metal and dirty consumables will pump gas into your weld. Hot-rolled steel comes coated in mill scale, a bluish-black iron-oxide layer that traps moisture and outgasses as it melts. Rust is hydrated iron oxide and is worse. Grind or wire-brush down to bright metal at least an inch on each side of the joint. Oil, grease, cutting fluid, layout marker, and paint all crack under the arc into gases that get trapped as porosity, so degrease with acetone or a dedicated solvent before you strike an arc.

Moisture is the hidden enemy and the source of hydrogen. Water on the plate, humidity condensing on cold steel brought in from an unheated yard, damp flux-cored wire, and a wet or contaminated gas stream all introduce hydrogen. Hydrogen dissolves in the molten pool and comes out of solution as it freezes, leaving pores, and in higher-strength steels it also drives hydrogen-induced (cold) cracking. Store low-hydrogen stick electrodes and flux-cored spools per manufacturer and AWS A5.1/A5.20 guidance, keep opened FCAW spools in a dry cabinet, and let cold steel warm to shop temperature before welding.

Galvanized coating deserves its own warning. Zinc boils off at a temperature far below steel's melting point and produces violent porosity, wormtracks, and toxic zinc-oxide fume. Grind the galvanizing back from the joint, ventilate aggressively, and follow OSHA fume limits. Finally, do not overlook the wire itself: rusty or fingerprint-contaminated MIG wire and old spools that have sat in humid air are a real porosity source. Where some surface contamination is unavoidable, a filler with extra deoxidizers such as ER70S-6 (AWS A5.18), rich in silicon and manganese, tolerates it far better than a leaner ER70S-3.

• Grind to bright metal: Remove mill scale, rust, and paint at least 1 inch back from the joint on both sides.
• Kill the moisture: Warm cold steel, store flux-cored and low-hydrogen consumables dry, and never weld over condensation or damp surfaces.
• Galvanizing: Grind zinc coating back, ventilate for the toxic fume, and expect porosity if you weld over it.
• Use a forgiving filler: ER70S-6 carries extra silicon and manganese deoxidizers that scavenge oxygen and tolerate light surface contamination better than ER70S-3.

Step-by-Step Porosity Diagnosis Checklist

When porosity shows up, resist the urge to grab the flowmeter first. Work this checklist in order, changing one variable at a time and laying a test bead after each change so you know what actually fixed it. Rushing and changing three things at once is how welders end up not knowing the cause and getting bitten again next week.

This sequence moves from the fastest, most common causes to the deeper ones. In most shops across the country the problem is solved by step three or four, long before you need to question the gas supply itself. If you reach the end and still have porosity, that is when a call to your local welding and gas distributor pays off, because a bad cylinder fill or moisture in the gas does happen and they can swap it or test it.

• 1. Confirm flow is in range: Set CFH to the recommended value for your cup or nozzle and gas. If already in range, do NOT increase it; move on.
• 2. Clear the nozzle and check consumables: Remove spatter buildup from the MIG nozzle, replace a clogged or worn contact tip or gas lens, and confirm the diffuser holes are open.
• 3. Kill drafts: Shut off fans, close doors, and screen the joint. Confirm air movement at the arc is under about 5 mph.
• 4. Clean the base metal: Grind to bright metal and degrease with acetone. Remove mill scale, rust, oil, paint, and galvanizing back from the joint.
• 5. Leak-check the gas line: Soap-test every fitting, inspect and replace hardened O-rings, swap cheap vinyl hose for braided/barrier hose, and run a leak-down test.
• 6. Verify the meter and reading: Confirm the flowmeter hangs vertical, is calibrated for your gas, and that the solenoid opens and closes cleanly with no start surge blowing into the tie-in.
• 7. Question the gas and consumables: Check for a near-empty cylinder delivering erratic pressure, damp flux-cored wire, or rusty MIG wire. If all else is clean, ask your distributor to check or swap the cylinder for moisture or a bad fill.

Frequently Asked Questions