Why travel speed is the number nobody hands you
Every setting chart gives voltage and wire feed speed for a material thickness. Almost none of them give travel speed, and the reason is that travel speed is not an independent setting. It falls out of two things: how much metal per minute the wire feed is delivering, and how much metal per inch the bead you want requires. Choose those two and travel speed is arithmetic.
The first half is a volume calculation. A wire of diameter d has a cross-section of pi times d squared over four. At 0.035 inch that is 0.000962 square inches. Feed it at 300 inches per minute and 0.2886 cubic inches of steel per minute leaves the contact tip, which at 0.2836 pounds per cubic inch is 4.91 pounds an hour. Not all of it lands: spatter, the vapour that leaves as fume, and on stick the stub you throw away, all come off the top. Call the survival rate the deposition efficiency, put it at 92 percent for solid wire under gas, and 4.52 pounds an hour actually becomes weld.
The second half is the bead. A fillet with equal legs of length W has a nominal triangular cross-section of W squared over two, so a quarter inch fillet is 0.03125 square inches. Real beads are convex rather than flat-faced, which is why there is a convexity allowance on this page, and 10 percent is a fair figure for a bead that looks normal.
Now divide. 4.52 pounds an hour at 0.2836 pounds per cubic inch is 15.9 cubic inches an hour, which is 0.2655 cubic inches a minute. Dividing by 0.03125 square inches gives 8.5 inches a minute of travel. That is the whole calculation, and it is why turning the wire up without moving faster gives you a bigger bead rather than a better one.
Measuring your actual wire feed speed
The dial on the machine is a reference, not a measurement, and on older equipment it can be a long way out. The check takes two minutes. Cut the wire flush at the contact tip, turn the gas off at the cylinder or set the machine to a purge mode so the arc cannot strike, pull the trigger for exactly six seconds, then measure the wire that came out. Multiply by ten and you have inches per minute. Do it at the settings you actually weld at, because feed rollers slip differently under load and a worn liner drags.
The same check catches a whole family of problems. Wire that comes out at a rate that wanders is a drive roll tension, liner or spool brake issue, and it is the underlying cause of a surprising amount of what gets blamed on technique.
When the answer comes out wrong-looking
| Result | What it usually means |
|---|---|
| Travel speed over about 30 in/min | The bead is small relative to the wire feed. Step the wire size down, turn the feed down, or accept that this is a mechanised speed. |
| Travel speed under about 4 in/min | The bead is large relative to the feed. Split it into more passes, which is what a real welder would do anyway. |
| Single pass over 0.06 in² | A large bead for the flat position and impossible out of position. Multi-pass. |
| Number looks right but the weld does not | Travel speed governs how much metal lands per inch. It has nothing to say about fusion, penetration or whether the arc length is right. |
That last row deserves emphasis. This page is a metal-delivery calculation. Two welds with identical travel speed and identical bead size can be sound and worthless respectively, because the difference lives in voltage, arc length, gun angle, joint fit and cleanliness, and none of those appear anywhere in the arithmetic.
Where the wire and gas actually go
Per foot of finished joint you get three consumption figures, and they answer different questions. Deposited metal per foot is what the joint weighs and is fixed by geometry. Wire per foot is deposited metal divided by efficiency, which is what you buy. Gas per foot is flow rate multiplied by arc time, and it is the one that surprises people, because gas is billed by the cylinder and consumed by the minute. At 35 CFH and 1.4 arc minutes per foot, a foot of quarter inch fillet burns about 0.8 cubic feet of gas, so an 80 cubic foot cylinder covers roughly a hundred feet of that joint before you count purging and the gas that runs during pre-flow and post-flow.
If you want the money rather than the quantities, the weld cost calculator takes the same geometry and adds filler price, gas price, operating factor and labour. This page and that one are deliberately different halves of the problem: this one is about how fast you move the gun, that one is about what the joint costs.
Fitting it into a day of work
Arc time per foot is the input the rest of the shop planning runs on. Feed it into the duty cycle calculator to find out whether the machine can sustain that arc time at your amperage, and how many parts an hour that permits before the thermal protection interrupts you. A comfortable travel speed on paper is still limited by the machine, and by the twenty other things that happen between parts.
The hazards, named once
None of the arithmetic on this site makes welding safe, and none of it is a substitute for being taught. Four hazards kill or maim people who were only doing a small job.
Fume is a respiratory hazard. Welding fume is condensed metal vapour fine enough to reach the deepest part of the lung, and it is harmful from every process and every base metal. Galvanized coating is worse again: the zinc burns off and produces fume that causes an acute illness, and coatings, platings, paints and residues each produce their own decomposition products. Grinding a coating off a joint before welding is not the same as ventilating, and neither is an open roller door. This is an area where the right answer comes from training and a competent assessment of the actual space, not from a web page.
Arc radiation burns eyes and skin, including bystanders. The ultraviolet output of an arc injures the cornea of anyone who looks at it without protection, from across a room, and the person holding the torch is the one wearing a helmet. Reflected light off a light-coloured wall does it too. Anyone in the space needs to be screened or protected.
Sparks start fires that smoulder for hours. Spatter and grinding sparks travel much further than people expect, fall through gaps, and settle into insulation, sawdust, rags and roof cavities where they can burn quietly for a long time after the work stops and everyone goes home. Fire watch after hot work is a real practice with a real reason behind it.
Welding on a sealed or previously filled container can explode. Drums, tanks, pipes and hollow sections that held anything flammable, or that are merely closed, are not repair candidates for anyone without the specific training for that work. Residue and vapour do not need to be visible, and a sealed void heated by an arc has nowhere for the pressure to go.
Learn this from an instructor, a course or a shop that will teach you, and from the manufacturer documentation that came with your equipment and consumables. This site gives you numbers, not procedures.
Questions people ask
Is a higher travel speed better or worse?
Neither on its own. Travel speed is a dependent variable: it is whatever it has to be to lay the bead you want with the metal you are delivering. If you increase travel speed without changing anything else, you get a smaller bead, not a faster weld of the same size. What genuinely varies with travel speed is heat input per inch, which falls as you speed up, and that changes the cooling rate, the width of the heat affected zone, and how much the part distorts. Faster is not free, it is a different set of trade-offs.
What deposition efficiency should I use?
It depends on the process and it varies with how you are running it. Solid wire under shielding gas is the most efficient of the common processes because nothing is lost to slag or a stub, and the losses are spatter and vapour. Gas shielded flux cored loses slag as well. Self shielded flux cored loses more again. Stick loses the slag coating plus the two inches of stub you cannot burn, which is a large fraction of a short electrode. Rather than trusting any single number, weigh a spool before and after a known length of weld once, work out what you actually got, and use that figure. It will be specific to your machine, your settings and your habits, which is the point.
Why does my quarter inch fillet need more than one pass?
Because a single pass carrying 0.031 square inches is at the upper end of what a manual welder can control in the flat position, and out of position gravity takes the puddle before it freezes. Multi-pass is not a compromise, it is normal fabrication practice, and it has a second advantage: each pass tempers the one below it and the smaller beads distort the part less than one large one. The passes field on this page divides the cross-section evenly, which is a planning simplification. In practice a root pass is smaller than the fill passes and the cap is sized for appearance.
Does this work for stick and TIG?
Partly. The deposition side of the calculation assumes a continuously fed wire of known diameter, so it maps directly onto wire processes and not onto stick, where you feed by hand at a rate you do not measure. For TIG the wire feed speed is also manual, so the deposition side does not apply. What does carry across is the second half: the bead cross-section per inch is geometry and is the same regardless of process, so if you know your deposition rate in pounds per hour from any source, the travel speed arithmetic still holds. The bead area mode on this page lets you drive it that way.
How do I check the travel speed I am actually using?
Time a known length. Mark the plate at the start and at twelve inches, weld the run at a pace that feels normal, and time it with a stopwatch. Twelve inches in ninety seconds is eight inches a minute. Most people are surprised, usually because they are slower than they thought. Comparing your measured speed against the number this page gives for your settings tells you whether the bead you are getting matches the bead you intended, and if it does not, one of the two inputs is wrong.