It started, as these things always do, with a sound. Dave’s 2004 F-150 had been quietly rusting for a decade in a New England driveway, and for most of that time the exhaust note was merely embarrassing — a soft metallic ticking at idle that his neighbors politely ignored. Then one Tuesday morning the ticking became a bark, the bark became a shout, and by the time he pulled into his own driveway at six in the evening the truck sounded like a piece of farm equipment trying to pick a fight. He slid underneath on a scrap of cardboard with a trouble light and found the problem in about four seconds: the two-bolt flange where the downpipe meets the y-pipe had rusted down to lace. One bolt head had snapped off clean; the other had reduced to a nub of orange powder. The gasket between the two faces had disintegrated years ago, and the resulting gap was puffing hot exhaust gas straight at his face every time he blipped the throttle.
The repair quotes came in fast and unhappily. One shop wanted $380 to cut out the rotted flange and weld in a new section, politely explaining that the surrounding pipe was thin enough they’d have to replace a foot of it. Another shop wanted $520 for a “full cat-back assessment” that Dave could read himself with a mirror and a flashlight. Meanwhile, sitting in the corner of his garage was a $199 140-amp flux core wire welder he had bought three years earlier for a lawn mower cart and a trailer hitch, a spool of .030″ wire, an auto-darkening helmet with a scratched lens, and exactly enough confidence to be dangerous. The question that consumed his entire Saturday was simple: can this gasless wire feeder actually weld an exhaust pipe back together, or is he about to turn a $380 repair into a $900 repair with a hole burned through a $120 catalytic converter?
That question comes up constantly, and the honest answer is yes — flux core can weld an exhaust system, with real conditions attached. Exhaust tubing is thin, coated, curved, usually rusty, and welded from underneath while you lie on your back; that is four of the least friendly variables you can combine on one job. It is not always the right process for the job, but with sensible settings, serious prep, and a disciplined approach to heat input, a gasless flux core machine will fuse 16- and 18-gauge aluminized tubing well enough to last years. What separates the welds that last from the welds that crack and leak in six months is not talent or an expensive machine — it is knowing what you are up against before you pull the trigger, which is exactly where this guide starts.
If you have ever laid a beautiful bead on 1/4″ plate in the shop and then tried to weld a piece of exhaust tubing, you know the whiplash. Exhaust work sits at the difficult end of almost every variable that makes arc welding challenging, and it stacks those variables on top of each other instead of isolating them one at a time. The tubing is thin. The surface is contaminated. The position is overhead. The geometry is a curve. The material is often rusted through. And the part has to actually seal when you are done, because a “strong enough” weld that leaks is a failed repair. Understanding why each of these factors fights you is the difference between guessing and making deliberate decisions at the machine.
Most factory exhaust tubing falls between 16 and 18 gauge — roughly 0.048″ for 18 gauge and 0.062″ for 16 gauge — and some cheaper replacement pipe is stamped as thin as 20 gauge in spots. Structural steel you might practice on is typically 3/16″ or 1/4″, four to five times thicker. The thermal math is unforgiving: there is very little metal to absorb and spread the arc’s heat, so the puddle goes from “just starting” to “hole” in a fraction of a second, especially if you dwell while getting oriented. That is why blow-through is the number one complaint on this job, and why settings chosen for plate will destroy tubing instantly. The same thinness creates distortion: as you weld one side of a joint, the heat shrinks on cooling and pulls the pipe out of alignment, opening gaps on the opposite side. We cover heat control on thin material in our guide to welding thin metal with an arc welder.
Two coating problems show up on nearly every exhaust job, and both will make your welds ugly or porous if you ignore them. Aluminized steel — mild steel sprayed with an aluminum-silicon alloy — is the default material for aftermarket exhaust pipe, chosen for the way it reflects heat and resists corrosion. The problem is aluminum oxide, which melts around 3,700°F, far above the ~2,600°F melting point of the steel underneath; drag the arc across a dirty, oxidized coating and you get a wandering arc, a puddle that will not wet out, and beads contaminated with floating oxide. The fix is mechanical: grind or wire-wheel the coating back to bright metal for at least half an inch on both sides of the joint before you strike an arc. The same principle applies to galvanized hangers and brackets — our guide to flux core welding galvanized steel covers the specifics, including the white zinc oxide smoke that will make you nauseous.
The other common material is stainless — usually 409 stainless (about 11% chromium) on OEM systems, sometimes 304 on aftermarket performance exhaust. Stainless is genuinely easier to burn through than mild steel because it conducts heat poorly: the heat stays pinned at the arc instead of running off into the pipe, so the puddle grows faster than you expect. It also discolors and, more importantly, welding stainless releases hexavalent chromium fumes, which is a genuine health concern rather than a theoretical one. Flux core can weld stainless if you use stainless flux core wire and accept some trade-offs in appearance and fume control, and we walk through that in our article on welding stainless steel with flux core. If your system is 409 and you own a gas bottle, TIG or gas MIG with 308L filler is the cleaner path; if it’s a rusty 16-gauge aluminized patch in a driveway, flux core remains perfectly serviceable.
Rust is the variable that turns a straightforward weld into surgery. Exhaust systems live in the worst environment on the vehicle: they cycle from ambient to 1,200°F and back, they get doused with road salt and water, and they sit inches from the road where stones chip their coating. By the time something fails, the surrounding metal is often thin, scaled, and full of pinholes — welding directly onto it is an exercise in frustration, as the arc blows through the thin spots while refusing to fuse the scaled ones. The professional approach is to cut back to metal that is still full thickness, usually an inch or two more pipe than feels necessary, and join new tubing to what remains. And all of this happens on your back under a car, with gravity pulling the puddle out of the joint, a limited view through a helmet, and a jack stand in your shoulder blades. Position work has its own settings and travel-speed rules — our horizontal flux core settings and vertical-up settings explain how to adjust when the pipe forces you out of flat. General habits for new welders are in our beginner welding tips collection.
Here is the part where an honest guide stops cheerleading. Flux core is a legitimate process for exhaust repair, but it is not the process a professional shop would choose for thin-wall tubing, and pretending otherwise would waste your time. Gas-shielded MIG — short-circuit transfer with 75/25 argon-CO2 — is the industry default for exhaust work because it runs cooler, cleaner, with less spatter, no slag to chip, and far better visibility of a puddle that is only a couple of beads wide. TIG produces the prettiest and most controllable exhaust welds on earth, and it is what high-end stainless systems get built with, but it demands two-hand coordination, expensive filler rod, and a level of patience that most driveway repairs cannot afford. Stick welding is genuinely difficult on 16-gauge tubing — the electrodes run hot enough that 6011 at minimum amperage will still punch holes — though a skilled hand with a 6011 or 7018 can absolutely join heavier flange material. We have a full comparison in our stick welding exhaust guide if that is the machine you own.
The strongest argument for flux core on exhaust is practical rather than metallurgical: it works without a shielding gas bottle, which matters when the alternative to welding it yourself is paying a shop. Self-shielded flux core also tolerates dirty, rusty, slightly oily surfaces far better than gas MIG, because the flux generates its own protective envelope right at the arc instead of relying on a gas cloud a 15 mph breeze can blow away — and that matters enormously when you are working outdoors next to a pipe you could not get perfectly clean. Flux core deposits metal faster and penetrates deeper per amp, which means you can often fuse a gap or a slightly misaligned joint that would stall a short-circuit MIG arc. If you are weighing a machine purchase, our beginner flux core welder guide covers what actually matters for jobs like this one. For a primer on the process itself, start with what is flux core welding, and for the head-to-head against gas wire see flux core arc welding vs MIG.
Now the costs. Flux core makes more spatter than gas MIG — noticeably more on thin tubing running low settings and short bursts — and the spatter sticks, so you will spend time after the weld with a wire brush knocking off berries. It generates more fume, which under a car with limited airflow is a real consideration we will return to in the safety section. It leaves a layer of slag over every bead that must be removed before you can even see whether your weld is good, and it hides defects in the meantime; slag inclusion and worm-tracking are the signature flux core problems, covered in our porosity in welding guide and our flux core technique article. Appearance suffers too: flux core beads on thin tubing tend to look lumpy compared with a neat MIG stack, and exhaust sections are sometimes visible under a lifted truck. Finally, flux core’s minimum practical heat on a 140-amp machine is higher than gas MIG’s, so the window for welding 18-gauge material without burning through is narrower — a problem you solve with stitch timing, not with the machine. Our flux core vs MIG comparison lays out the full picture.
The practical verdict: if you own a gas-capable MIG machine, use gas MIG for the visible, structural parts of the exhaust and save flux core for the dirty, out-of-position patches. If you own only a flux core machine — which describes most people attempting this repair — it will do the job well if you respect the heat and the prep. And if the joint is a flange-to-flange connection with thick material, remember that flux core welds are plenty strong at those thicknesses; the thin tubing is the challenge, not the strength. One more process note before moving on: verify your polarity. Self-shielded flux core must run electrode positive, and hooking up a machine backwards is one of the most common setup errors in the hobby — our polarity guide shows the correct connection in thirty seconds.
Exhaust welding rewards preparation more than almost any other repair, because once the arc is lit you will have roughly a second per decision. Everything that can be done ahead of time — cleaning, cutting, deburring, clamping, aligning — should be done ahead of time, on the bench if at all possible, because working upside down under a car is where good prep plans go to die. The full toolkit is not exotic: an angle grinder with a cut-off wheel and a flap disc, a wire wheel for the drill, a reciprocating saw or tubing cutter for removing rotten sections, exhaust clamps to hold things in alignment while you tack, a file or deburring tool, a magnet or two, and something to protect the surroundings from spatter. Add a fire extinguisher, a welding blanket or wet cardboard to shield fuel lines and plastic, and jack stands rated for the vehicle — never weld under a car supported only by a jack.
For 16- and 18-gauge aluminized or mild steel tubing, .030″ self-shielded flux core wire (E71T-11 or E71T-14 classification) is the right starting point and honestly the right endpoint for most driveway exhaust work. .035″ wire ships with many machines and will weld 16 gauge, but it feeds more metal than you want on 18 gauge and demands more amps to run cleanly — heat you are trying to avoid. .023″ wire is overkill: it is designed for sheet metal thinner than anything on your car’s exhaust. The debate in full lives in our .030 vs .035 comparison, and our best flux core wire guide ranks the brands that feed smoothly through cheap machines. Two setup details matter as much as the wire: make sure the drive rolls are grooved for the diameter you are running (a .035 groove running .030 wire slips and feeds erratically — see our drive rolls guide), and confirm the machine is in flux core mode — electrode positive with the gas solenoid off. If you have ever wondered whether solid MIG wire can substitute, our article on using MIG wire in a flux core welder explains why that usually ends badly.
Before any welding happens, deal with the rust honestly. Probe the area around the failure with a screwdriver; if you can push the tip through the pipe wall, that metal is gone and no bead will save it. Cut outward from the failure until you reach tubing that is solid enough to resist a firm screwdriver poke — often an inch or two beyond where the visible hole ends, sometimes the entire length of a section. A reciprocating saw with a fresh metal blade is the fastest way under a car; a tubing cutter gives a cleaner, squareer edge if you have room to swing it, though it cannot reach in tight spaces. Deburr both cut ends inside and out with a file, because the raised lip left by a cut-off wheel creates a gap that flux core struggles to bridge cleanly. If the failure is at a flange face — as in our driveway story — and the flange itself is eroded or the bolt holes are oval, replace the flange rather than trying to rebuild it: a warped flange will leak no matter how good your weld is.
Grind or wire-wheel at least half an inch of bare, bright metal back from each side of the joint — both pieces. On aluminized tubing this removes the coating that causes arc wander and oxide inclusions; on rusty tubing it removes the scale that causes lack of fusion; on either, it removes the road film and undercoating that cause porosity. Degrease afterwards with brake cleaner or acetone if the pipe has oil on it, and let it dry completely — solvent trapped in a seam will boil into your puddle and pop porosity holes as you weld. Ground clamp placement matters: attach it directly to the pipe as close to the weld as possible, not to the jack, the frame, or a hanger three feet away, because a poor ground on thin tubing destabilizes the puddle. Fit-up is the last prep step and the one most often rushed: a butt joint gap should be no wider than about 0.030″, a lap joint should overlap 1/4″ to 3/8″, and the pieces should be aligned with exhaust clamps, padded vise grips, or a piece of pipe slid inside both ends as an alignment splint. If the gap is wider than the material thickness, you are no longer welding — you are filling.
Settings are where thin-tubing welding is won or lost, and the most useful habit you can build is testing on a scrap piece of the exact same tubing before touching the car — the difference between “about right” and “hole burned through” on 18 gauge is often a single volt. The numbers below are typical starting ranges for self-shielded flux core with a 120–140 amp class machine; treat them as the middle of the road and adjust from what your arc tells you. A chart on the inside of your machine’s door makes a useful second opinion — our wire speed and voltage chart covers the full thickness range, and the flux core rod and technique article explains what each adjustment does to the puddle.
| Material / Joint | Wire | Volts (typical) | Wire Feed (ipm) | Approx. Amps | Notes |
|---|---|---|---|---|---|
| 18 ga (0.048″) aluminized, butt joint | .030″ E71T-11 | 17.0–18.0 | 90–110 | 65–80 | Cool end of the range; short stitches, frequent pauses |
| 18 ga, lap or overlap joint | .030″ E71T-11 | 17.5–18.5 | 95–115 | 70–85 | Overlap gives a heat sink; slightly hotter is fine |
| 16 ga (0.062″) aluminized, butt joint | .030″ E71T-11 | 18.0–19.0 | 100–125 | 80–95 | The sweet spot for most exhaust repairs |
| 16 ga, flange or thick bracket | .030″ or .035″ | 18.5–20.0 | 115–150 | 90–110 | Thicker material tolerates real heat; watch the thin pipe side |
| 14 ga (0.075″) downpipe | .035″ | 19.0–20.5 | 130–160 | 100–120 | Most 140-amp machines handle this comfortably |
| 409/304 stainless, 16–18 ga | .030″ stainless flux core | 17.0–18.5 | 90–120 | 65–90 | Stainless holds heat; shorter stitches and lower amps |
| Rusted / scaled pipe, back to sound metal | .030″ | 18.0–19.0 | 100–125 | 80–95 | Slightly hotter to burn through scale — only on sound metal |
Those amperage figures are approximations derived from published wire-feed-to-amperage relationships; wire feed speed, not a dial reading, is what actually controls heat on a constant-voltage machine. If your welder has an amperage display, great — use it. If it only has numbers from 1 to 10, ignore the numbers entirely and tune by ear and eye: a smooth, quiet hiss with a fluid puddle is right; a violent crackle with big spatter balls means too much voltage or wire feeding too fast for the setting; a stubbing, popping arc that keeps sticking to the work means not enough voltage or too little wire for the arc to sustain. Our amperage chart shows how to translate between systems, and machine-specific dial logic is discussed in our flux core settings guide — worth comparing against the manufacturer starting points published in Miller’s resource library and Lincoln Electric’s technical library, since every brand scales its dials differently. A practical tip from the field: write your working combination on a piece of tape inside the machine’s door the moment you find it. Exhaust work involves repeated sessions — tack today, weld tomorrow — and remembering that 18.0 volts at 105 ipm is what worked saves you twenty minutes of re-discovery next time.
One adjustment most beginners underestimate is electrical stickout — the distance from the contact tip to the work. Flux core runs at a longer stickout than MIG, typically 1/2″ to 3/4″ versus 1/4″, and the longer stickout reduces amperage for a given wire speed because the wire has more resistance before it reaches the arc, which makes it a legitimate way to shave heat on thin tubing without touching your dials; going past an inch, though, causes erratic arcs and cold lap. Travel speed works the same way: slow down at all and the puddle will drop through, so on butt joints each 1/2″ of bead should take roughly one to one-and-a-half seconds. Finally, check your wire speed accuracy once — most budget machines deliver less wire than they claim, so the “90 ipm” on the knob might really be 75. Hold the trigger over a tape measure for six seconds to find the truth, then choose the setting that gives you the real number you need.
With prep done and settings dialed in on scrap, the actual welding follows a sequence that never changes much: fit, tack, check, stitch, cool, rotate, repeat. The discipline lives in the middle of that list — every shortcut from it is a shortcut toward a warped, leaky joint, and the difference between a repair that lasts three years and one that lasts three weeks is almost always how strictly the stitch-and-cool rhythm was followed. Exhaust tubing carries no structural load, so you are never in a hurry; the metal only needs to seal, not hold up a bridge, and patience costs you nothing but a few minutes.
Once the pieces are clamped in final position with the gap within spec, tack first and weld second — no exceptions. Place the first tack at twelve o’clock (the top of the pipe), holding the trigger for roughly one second so the tack is a small button of metal about 1/8″ across rather than a crater. Move directly to the opposite side — six o’clock — and place the second tack there. That opposition matters: a tack on one side pulls the joint toward it as it cools, and its twin on the opposite side cancels that pull instead of letting the joint walk out of alignment. Then add tacks at three and nine o’clock the same way, checking between each pair that the gap has not opened. If the joint has shifted, fix it now: re-grinding a tack and re-tacking takes thirty seconds, while the same correction after six stitches takes twenty minutes. Four tacks at the cardinal points is the baseline for any butt joint; on a larger diameter pipe, use six or eight for the same reason. For general tack-welding practice, our flux core technique guide covers fundamentals that apply far beyond exhaust work.
Stitch welding is the single most important technique on thin tubing, and it works by treating the weld as a series of short, discrete events rather than one continuous line. Lay a bead segment roughly 1/2″ to 3/4″ long — about one to one-and-a-half seconds of travel — then release the trigger and let the arc die. Move to a position directly opposite the segment you just laid and run another segment of the same length. Keep hopping around the circumference in that same star sequence, so heat never concentrates on one side long enough to warp it. Between each segment, wait until the pipe is merely warm to the touch — if you can’t hold a finger on it comfortably, it is still too hot to add more heat to that area. That usually means ten to twenty seconds of rest for 16-gauge tubing, and longer near a flange or bracket acting as a heat sink.
The overlap discipline is what makes stitches join into a sealed seam instead of a string of isolated islands. Start each new segment on the trailing edge of the previous one, about 1/4″ onto already-deposited metal, so the new bead ties in rather than starting cold on bare tubing; starting on cold metal is how you get a cold-lap lump that leaks even though it looks welded. Finish each segment by slowing slightly to fill the crater, because a crater left on thin tubing is a crack starter. Between passes, chip the slag completely and wire-brush the seam; flux core slag left on the surface contaminates the next bead’s tie-in and shows up later as worm tracks or porosity buried under a bead you thought was fine. If you find yourself tempted to run one long continuous bead “because it’s going well,” remember what happens on 18 gauge when heat has nowhere to go: you will get two inches of beautiful weld and a molten hole in the third. Our guide on push vs pull technique covers the speed and angle habits that make stitch work cleaner.
For self-shielded flux core you drag the gun — point the nozzle back toward the direction you are traveling, with roughly 10 to 20 degrees of angle off vertical. Dragging buries the wire in the puddle and lets the flux shield right at the arc; pushing blows the flux ahead and leaves you welding blind through a shower of it. Hold your electrical stickout around 1/2″ and keep it consistent, because wandering the stickout changes your amperage mid-bead without you touching a dial. Travel speed should be brisk and steady: on a butt joint in 16 gauge, think “half inch per second,” moving fast enough that the puddle stays shallow and never turns mirror-bright and saggy. Watch the leading edge of the puddle rather than the wire — its width tells you everything. A puddle narrower than the wire diameter means you are too cold; one that shimmers and drops means you are too hot or too slow.
Position changes the recipe slightly when the pipe forces you out of flat. Overhead segments — the bottom of a pipe still on the car — generally want about 5 to 10 percent less heat and slightly faster travel, because gravity is trying to pull the molten metal out of the joint and your job is to freeze it in place before it can sag. Vertical sections are common on downpipes; vertical-up gives the best control of the puddle, though on thin tubing many welders simply reposition the pipe and avoid the argument entirely. Machine-side adjustments for position are covered in our vertical-up settings and horizontal settings articles. When the exhaust section can be removed from the car — which is almost always true for a muffler, a tailpipe, or a mid-pipe — take it out and weld it on the bench in flat position. That one decision upgrades every aspect of the job: better visibility, better ground, better position, and a much easier cleanup.
Real repairs rarely present you with a perfect fit, and the two scenarios that ruin most driveway exhaust jobs are gaps wider than the material and metal too thin to weld. The rule of thumb: a butt joint gap should be no larger than the wall thickness — for 16 gauge, under about 0.060″. A gap inside that limit can be bridged tack-and-stitch: lock the geometry with tacks on both edges, then run short stitches across the gap at a slightly faster travel speed, letting each dab freeze before the next, almost like building a ladder of small welds rather than running a bead along one edge. Speed is your friend in a gap because the puddle has less metal underneath it to hold heat, so slow down for one second and you will watch the edges fall away from each other.
If the gap exceeds roughly an eighth of an inch — or if the two pieces are mismatched in diameter, which happens constantly when someone combines aftermarket and factory tubing — stop trying to weld it and put metal back. Cut a short sleeve from a larger-diameter piece of exhaust tubing (or buy a pre-made repair sleeve), slide it over the joint so it overlaps each side by about half an inch, and weld the two sleeve edges rather than trying to fill the original void. A sleeve also solves the rust problem: when the original pipe is too pitted or thin near the failure, cut back past the damage and sleeve the joint, giving you two sound, full-thickness edges instead of one paper-thin edge that blows through on contact. Welding over rusted, scaled metal directly is the classic beginner error — the arc blows holes in the thin sections while refusing to fuse the scaled ones, leaving a bead bridging a field of pinholes. And welding a hole shut is only ever a cosmetic fix: if the surrounding wall has lost thickness, the pipe will perforate again next winter, which is why cutting back to sound metal remains the only permanent repair.
Once the circumference is stitched closed and everything has cooled, resist the temptation to reinstall immediately and call it done. Chip every bit of slag off the seam with a chipping hammer or a flat screwdriver, wire-brush until the bead is visible, and look at what you actually made. On 16-gauge tubing a good flux core weld shows a consistent bead width and — most importantly — no visible pinholes, cracks at the crater ends, or areas where the bead sits proud on one edge without having fused into it. Undercut along the toe is both a leak path and a crack starter; a quick pass with slightly higher wire speed usually fills it. Any grinding you do should be light and aimed at knocking down high spots only; aggressive grinding into a thin-walled weld is how you convert a good repair into a new hole.
With the system fully assembled and the engine cold, spray the entire seam with soapy water from a spray bottle, then start the engine and let it idle — in an open, ventilated area, never in an enclosed garage, because you are introducing both exhaust gas and carbon monoxide into the test. Where the weld leaks, the soapy film will bloom into a steady stream of bubbles; a good weld stays flat and unremarkable. An old trick: loosely block the tailpipe with a rag to build backpressure so even a hairline leak announces itself — never seal it tight, or other gaskets will give you false positives. If bubbles appear at the weld, let it cool, wire-brush the area, and place one or two short stitches directly over the pinhole at reduced settings; the alignment work is already done, so the touch-up takes two minutes. For why a sealed exhaust matters beyond noise — unburned hydrocarbons escaping upstream of the catalytic converter raise the vehicle’s emissions — see the EPA’s vehicle emissions resources, and our full penetration vs fillet weld explainer covers why a surface bead sealing a butt joint is not a structural fusion weld. Once the test passes, run a thin bead of sealant along the seam and put anti-seize on disturbed flange bolts — the job is done.
Welding is not always the right answer for a driveway exhaust repair, and plenty of professional repairs on older vehicles don’t involve an arc at all. The first and most underused option is the clamp repair: cut out the rotten section, slide a butt-coupling exhaust clamp or a lap-joint sleeve over both ends, tighten the band, and seal the seams with exhaust compound. It takes twenty minutes with basic hand tools — no heat, no grinding, no skill beyond cutting square — and on a vehicle nearing the end of its life, spending four hours welding a perfect joint onto a dying car is a poor investment. The trade-off is longevity: clamps on thin tubing eventually develop their own small leaks as the pipe continues to corrode, and they must be checked periodically.
The second alternative is torch brazing with a silver alloy rod — typically 40% or 56% silver, flowing around 1,150°F to 1,400°F, well below the melting point of the steel underneath. Brazing, which Britannica’s welding overview distinguishes from fusion welding by the fact that the base metal never melts, is genuinely excellent on thin, rusty, hard-to-weld tubing: the filler wets and bonds to cleaned surfaces, and the lower heat input makes burn-through nearly impossible. The requirements are clean metal, brazing flux, and a torch steady enough to control the puddle — but for a small leak or a butt joint on paper-thin pipe, many experienced fabricators reach for the torch before the welder. Our welding versus soldering comparison explains where brazing sits in that spectrum. A third option, sold at any auto parts store, is tack-weld-plus-exhaust-paste: a couple of small tacks to mechanically hold a patch or sleeve, with the perimeter sealed in high-temperature exhaust paste. That combination is legitimate for small leaks — the tacks provide the mechanical retention paste alone lacks — but paste is not a weld, must be cured by running the engine, and will fail under thermal cycling. Finally, for flange failures, a pre-fabricated replacement section fitted with clamps is often cheaper than any repair, including your own labor — and when the section carries emissions-critical components like the catalytic converter and oxygen sensors, proper replacement also protects the vehicle’s emissions compliance, an area where state inspection programs leave no ambiguity.
Exhaust welding deserves its own safety section because the material you are burning is about as far from clean plate steel as it gets. Every layer on an exhaust pipe is a chemical you are about to aerosolize: the aluminum-silicon coating, the zinc on galvanized hangers, the chromium in 409 and 304 stainless, the undercoating and road tar outside, the oil and carbon deposits inside, and the rust itself, which carries decades of absorbed road salt. Flux core amplifies the problem because it produces noticeably more fume than gas-shielded MIG. The occupational guidance on this is unambiguous: OSHA’s welding, cutting, and brazing standards require ventilation sufficient to keep fumes below exposure limits, and NIOSH’s welding topic pages document the respiratory hazards, including metal fume fever from zinc, pulmonary effects from iron oxide, and the elevated lung cancer risk associated with stainless steel and hexavalent chromium. The profession’s consensus standards, published through AWS Standards, provide the safety practices and filler classifications shops build procedures around.
Practically, that means three things. First, ventilation: weld outdoors when you can, or position a fan so it blows fumes away from your breathing zone rather than across it, and keep your head out from under the car when you are not actively striking an arc — the area under a vehicle traps fume in a pocket that has nowhere to go. Second, respiratory protection: a NIOSH-approved respirator fitted to your face with P100 particulate filters is the baseline for this work, with combination particulate/organic vapor cartridges if you are burning off coatings or sealants; a mask that leaks around the cheeks because of a beard or a bad strap is decorative. Our best welding respirator guide explains the fit and filter options, and are welding fumes toxic covers what the fumes actually contain. Third, remove or shield everything flammable: fuel lines, plastic clips, rubber hangers, sound deadening, dry leaves and oil-soaked rags — and never weld within reach of a fuel tank that has not been purged and vented. Keep a charged fire extinguisher within arm’s reach of where you are lying, not on the bench across the garage.
The other hazard unique to this job is the engine itself. Every leak test requires running the engine, which means carbon monoxide — an odorless, lethal gas that has killed plenty of home mechanics who “just left it running for a minute” in a closed garage. Run leak tests outdoors or with the garage door fully open, never sit in the car with the engine running during a test, and remember that a helmet and a running engine under a car is exactly where you lose track of time. General shop habits for work like this are in our welding safety tips library, and a good auto-darkening helmet makes out-of-position work far less miserable than a fixed-shade lens.
Most failed exhaust welds trace back to a short list of repeatable errors, and knowing them in advance is cheaper than discovering them through trial and error on a $120 pipe. The big three are heat, heat, and heat: settings borrowed from thicker material, travel speed too slow for thin tubing, and dwell time in one spot while the welder gets oriented. Each of those manifests as a hole; together they are responsible for the majority of ruined exhaust sections. Beyond heat, the errors cluster into prep failures, setup failures, and sequencing failures. Prep failures — grinding only half the joint, welding over coating, rust, or undercoating — produce porosity and lack of fusion that no technique can rescue. Setup failures include wrong polarity for self-shielded wire, drive rolls sized for the wrong wire diameter, a ground clamp attached to the frame instead of the workpiece, and contact tips worn oversized from previous projects. Sequencing failures feel like time-saving shortcuts: running a continuous bead, skipping cool-down intervals, not chipping slag between stitches, and welding the whole circumference from the easiest spot instead of tacking and rotating in opposition. Our consolidated article on flux core mistakes worth avoiding is worth a read before you start: arc strikes outside the joint, crater cracks, and welding over paint are invisible until the job is finished.
There are quieter mistakes, too, that cost money rather than welds. Buying the wrong wire: the cheapest spool on the shelf may be .035″ hard-drawn wire with a flux formula designed for thicker structural work, and it will fight you the whole way on 18 gauge; spending a few dollars more on a quality .030″ E71T-11 spool that feeds smoothly through a budget machine changes the entire experience, and our best flux core wire guide points you toward the ones that do. Ignoring the machine’s own maintenance — a liner packed with flux dust, a drive roll tension cranked too hard (which deforms the wire and causes erratic feed), or a ground clamp with a corroded jaw — turns a settings problem into a mystery. Welding when tired or rushed matters more on exhaust than almost any other job, because the work is overhead with your arms extended and every decision is made in about a second; a weld you would reject on the bench becomes the weld you accept when your arms are shaking. And finally, not testing settings on scrap — our settings guide hammers the point that two minutes of scrap testing beats twenty minutes of hole-filling.
When something goes wrong mid-repair, diagnosing from the symptom is faster than adjusting dials at random, and exhaust work produces a predictable set of defects with predictable causes. Porosity presents as pinholes or a spongy, cratered bead once the slag is chipped away; on tubing it is almost always contamination — coating, oil, rust scale, or moisture — or wind disrupting the flux’s shielding, and the fix is mechanical cleaning and a wind break rather than a settings change. Blow-through is the sound you were dreading — a drop of molten metal falling on your shirt — and it means heat input exceeded what the wall thickness could hold: too many volts, travel too slow, too long a dwell, or a gap too wide to bridge. Less obvious is worm-tracking, thin slag lines running through an otherwise decent bead, caused by re-welding over slag or overlapping onto a contaminated surface; chipping and brushing between every stitch eliminates it. Undercut — a groove melted alongside the bead — signals too much voltage or too aggressive an angle, and on thin tubing it is a leak path and a crack starter. The full diagnostic catalog lives in our flux core troubleshooting guide and the porosity deep dive; the table below narrows it to exhaust tubing.
| Symptom | Most Likely Cause | Quick Fix |
|---|---|---|
| Pinholes or spongy bead after slag removal | Coating, oil, or scale not ground off; wind on the arc | Grind to bright metal, add a wind break, re-weld the affected area |
| Burn-through / molten drops falling | Volts too high, travel too slow, dwell in one spot, gap too wide | Drop 0.5V or 10 ipm, speed up, lengthen stickout slightly |
| Excessive spatter that won’t brush off | Voltage out of balance with wire speed; worn contact tip | Balance the dials per our spatter reduction guide |
| Arc stubbing and sticking constantly | Voltage too low for the feed speed; stickout too short; dirty wire | Raise voltage slightly, slow the feed, clean the liner and tip |
| Worm tracks in the bead | Slag not removed between stitches; welding over contaminated metal | Chip and wire-brush to bare metal before the next stitch |
| Cold lap — bead sitting on top without fusing | Amps too low, stickout too long, surface not cleaned | Increase feed speed 10–15 ipm, shorten stickout, re-grind |
| Arc won’t start or keeps extinguishing | Poor ground, wrong polarity, clogged liner or contact tip | Move the clamp to the pipe, verify electrode positive, clear the feed path |
| Crack at the end of a stitch | Crater left unfilled; too much heat locked into the joint | Slow at the end of each segment, add cool-down time, stitch in opposition |
Two broader diagnostic notes save time before you blame technique. First, if the arc is unstable from the moment you strike it — sputtering, wandering, refusing to hold a consistent sound — the problem is setup, not skill: check polarity, check that the drive rolls match the wire, check your ground, and check your stickout. Second, if your welds look fine on scrap but fail on the car, the difference is position and access; the fix is usually a slight reduction in heat and a deliberate shortening of each stitch to compensate for reduced visibility. Parameter-level fine tuning for these situations is mapped out in our flux core parameters article. Technical background on what the flux actually does inside that arc — shielding, slagging, and alloying in one pass — is well covered by TWI’s technical knowledge base, worth bookmarking whenever you want the engineering answer behind the practical one.
Yes, when the settings are matched to thin tubing and the heat discipline is respected. Flux core will weld 16- and 18-gauge exhaust material successfully using the ranges earlier in this guide — .030″ wire, roughly 17–19 volts, 90–120 ipm — combined with stitch technique and proper prep. The burn-through reputation comes from welders running plate settings on tubing or welding continuously instead of in short segments; test on scrap of the same pipe first and the holes disappear from the equation.
Use .030″ self-shielded flux core wire, E71T-11 or E71T-14 classification, for both 16- and 18-gauge aluminized or mild steel exhaust. It feeds cleanly through most 120–140 amp machines and puts down less metal than .035″, which matters on thin walls; for stainless exhaust, switch to a stainless flux core wire matched to the alloy. Our .030 vs .035 comparison covers the trade-offs in detail.
Gas MIG is better if you have the option — it runs cleaner, with less spatter, less fume, no slag, and a more controllable puddle on thin tubing, which is why shops use it. Flux core is better when you have no gas bottle, when you are welding outside where wind would disrupt shielding gas, or when the surface is dirty and hard to clean perfectly. Both produce sound welds on exhaust material when set up correctly; our MIG exhaust guide and the flux core vs MIG comparison walk through the decision.
You should remove at least half an inch of coating back from each side of the joint on both pieces, grinding to bright metal. The aluminum-silicon coating creates aluminum oxide, which melts far above steel’s temperature and causes arc wander, lack of fusion, and inclusions if left in the weld zone. It is the single most common cause of “impossible” exhaust welds, and the same cleaning logic applies to galvanized and zinc coatings.
Electrode positive, also called DCEP or DC reverse — the work clamp on the negative side and the gun on the positive. Self-shielded flux core requires it for proper arc characteristics and penetration; running it electrode negative produces a weak, sputtery arc with poor fusion and excessive spatter. If your machine has a polarity switch or a quick-change lead, verify it before your first weld rather than after twenty minutes of confusion; our polarity guide shows the correct setup with photos.
You don’t weld the rust — you remove it. Cut back until the tubing resists a firm screwdriver poke and reaches full original thickness, then join sound metal to sound metal using a sleeve, a patch, or a new section of pipe. Welding directly onto scaled, thinned metal produces a bead bridging pinholes that will leak and re-perforate within months. If the damage extends over a long stretch, replace the section; it is faster, cheaper, and permanent.
Yes, with stainless flux core wire matched to the alloy — 308L wire for 304 stainless, and a suitable wire for 409. Stainless holds heat at the arc and burns through faster than mild steel, so run at the lower end of the settings range with shorter stitches. Welding stainless produces hexavalent chromium fumes, so ventilation and a fitted respirator are especially important on stainless work — see our welding stainless with flux core article for the full procedure.
A properly fused weld seals itself; exhaust sealant is insurance, not a substitute. After welding, chip the slag completely, inspect for pinholes, and pressure-test with soapy water while the engine idles — if no bubbles form, the weld is sound. A thin bead of high-temperature sealant along the outside seam adds peace of mind, but paste over a poorly fused bead will eventually blow out — if the test shows bubbles, re-weld the leak rather than sealing over it.
You can, and often must — but remove the section whenever access allows, because bench welding in flat position is easier, safer, and produces better results. If the pipe must stay on the vehicle, support it properly, shield fuel lines and plastic with a welding blanket or wet cardboard, keep a fire extinguisher nearby, and plan your stitch sequence to control heat. Expect position to force slightly lower settings and shorter stitches than the flat-position starting points in this guide.
They can be, because exhaust material is coated, oiled, rusty, and often stainless — each adds a hazard to the base metal fume flux core already produces. OSHA and NIOSH both address welding fume exposure directly, and the practical controls are ventilation plus a properly fitted P100 respirator: zinc causes metal fume fever, and stainless produces hexavalent chromium.
You still have good options. Exhaust repair clamps and sleeves handle most section failures with hand tools, silver brazing with a torch bonds to cleaned steel at far lower heat than welding, and tack-weld-plus-exhaust-paste works for small leaks if you can borrow a machine for five minutes. Pre-fabricated replacement sections from an auto parts store are often the cheapest path on a common vehicle. A flux core machine is still worth owning for the many other jobs around the house — our beginner welder guide covers where to start.
Dave’s Saturday ended at 4 p.m. — later than he planned, earlier than the shop could have started. The flange was cut out, a new section of 16-gauge aluminized tubing was sleeved in and stitched with about forty short beads from a $199 gasless wire feeder, and the leak test with a spray bottle showed nothing but flat soapy water along the entire seam. The weld isn’t pretty — flux core on thin tubing rarely is — but it is fused, sealed, and it has survived two New England winters of salt and thermal cycling without a whisper. That result is repeatable, and it does not require talent; it requires respecting four facts. Exhaust tubing is thin, so your settings and stitch discipline must be built for it. The surface is contaminated, so prep is half the job. The position is hostile, so removing the pipe and welding on the bench is always worth the extra effort. And the fumes are real, so ventilation and a respirator are part of the toolkit, not an afterthought.
If you take nothing else from this guide, take the workflow: cut out the rot, clean to bright metal on both sides, fit tight, verify polarity, dial in your settings on scrap, tack in a cross, stitch in opposition, cool between segments, chip your slag, and test with soapy water before you lower the car. Every step is small, and together they are the difference between an exhaust repair that lasts and one you are doing again next spring. Flux core is a legitimate, honest tool for this job — not the first choice a professional shop would make, but absolutely the right choice for a driveway, a gasless machine, and a pipe that needs to hold. When the arc starts up and the puddle goes quiet and fluid instead of violent, you will know you have it — from there it is just a matter of keeping your head, your stickout, and your travel speed steady. If a particular symptom still refuses to behave, our troubleshooting guide and our core flux core technique article are the two pages worth having open in the garage.