Worm Weld Defects Explained: Wormhole Porosity, Worm Tracks, Causes, Fixes & Code Limits
Worm weld defects explained: what wormhole porosity and worm tracks are, why they form, how to find them,…
READ ARTICLE →The trailer hitch looked perfect — shiny bead, even ripples, a coat of paint over the top — until the first pothole snapped it clean off at the coupler, taking a $3,000 load of lumber into the ditch. The welder had charged the least, promised the most, and never once clamped, cleaned, or checked anything he laid down, which is exactly how a ‘bad welding job’ starts: with a weld that passes for looking good and fails the moment it matters.
The trailer hitch looked perfect. Shiny bead, even ripples, a fresh coat of black paint over the top. It held for three weeks. Then the first real pothole snapped it clean off at the coupler, sending a $3,000 load of lumber nose-down into the ditch. The welder had charged the least, promised the most, and never once clamped, cleaned, or checked a thing he laid down.
That is exactly how a bad welding job starts: with a weld that passes for looking good and fails the moment it actually matters.
If you landed here, you have met a version of that story. A gate post that cracked at the base the first windy day leaned on it. A “custom” exhaust that started hissing two months after you paid for it. Maybe you are a shop owner staring at a returned part, or a beginner who suspects your beads are not what the videos promised.
Whatever brought you here, the questions underneath are the same:
Welding is unusual as a trade and frightening when it goes wrong. In most crafts, bad work announces itself. A bad paint job drips. A bad tile job’s lippage catches your toe. A bad electrical job trips a breaker.
A bad weld can sit there looking like jewelry and be fused to absolutely nothing. The bead can be beautiful, evenly rippled, correctly convex, and still be riding on the base metal with no real bond underneath. You cannot see it. Paint covers it. It passes a casual glance. Then it lets go, without warning, under load, at speed.
That gap between appearance and reality is why “bad welding job” is such an anxious search. You are not asking whether it looks ugly. You are asking whether something you cannot see is going to hurt someone.
This is not a niche concern. The U.S. Bureau of Labor Statistics Occupational Outlook Handbook tracks welders, cutters, solderers and brazers at roughly 430,000 jobs with median pay near $49,000, one of the largest skilled-trades occupations in the country. That means millions of welds laid down every day across a very mixed pool of training and supervision.
Regulators take the risk seriously. OSHA’s welding, cutting and brazing standard, 29 CFR 1910.252, tells employers to insist that welders and their supervisors are trained in the safe operation of their equipment and the safe use of the process, because untrained work is a recognized hazard. The same standard requires a written permit and a fire watch of at least half an hour after welding or cutting ends, which shows how seriously regulators treat routine hot work. OSHA points to the American Welding Society and its ANSI Z49.1, Safety in Welding, Cutting and Allied Processes, as the reference for how this work should be done. On the health side, NIOSH has documented decades of long-term fume exposure risks, from manganese and manganism to neurobehavioral effects and rare conditions like welder’s anthrax.
Quick Note: You do not need to be a certified inspector to protect yourself. You need to know what to look for, the right questions to ask before work starts, and when a professional inspection is non-negotiable. If you are starting out, our welding PPE guide covers the gear that keeps you whole.
We will work through this in order, from symptoms to systems:
By the end, you should be able to look at any weld, ask four or five questions, and know whether it is safe or dangerous. Our breakdown of pipe welding procedures shows how formal process discipline gets once failure is not an option.
Let’s start with what a bad weld actually looks like.
A bad welding job, in plain language, is any weld that fails to do what it was supposed to do. That failure can show up four different ways: it looks wrong (visual), it breaks the governing standard (code), it fails in service, or it creates a hazard (safety). A single weld can flunk all four — or it can quietly pass three and fail the last one, which is exactly how dangerous work gets paid for and painted over.
Here’s what trips up most buyers: welding is one of the few trades where “looks fine” and “is fine” are genuinely different questions. A bead can be gorgeous and fused to nothing. It can look rough and hold a semi-truck. Judge by appearance alone, and eventually you get burned.
Inspectors split weld problems into two buckets.
Cosmetic (appearance) issues are surface problems: uneven ripple spacing, spatter balls welded to the plate, arc strikes (those scorch dots where the electrode touched off the steel), or a rough, slightly lumpy cap. They look bad and hint at a sloppy hand, but many are acceptable under code. Spatter, arc strikes, and bead profile are largely appearance concerns, which is why a good inspector separates “appearance” from “soundness” before condemning anything.
Structural (soundness) issues are the real threats: cracks, lack of fusion, lack of penetration, undercut, and slag inclusions. These weaken the joint itself, and they do not care how smooth the cap pass looks.
Quick Note: “Ugly but sound” is a rough-looking weld that is fully fused and passes inspection. “Pretty but fused to nothing” is a smooth, evenly rippled bead sitting on cold, unclean steel with no penetration behind it. The second one is the dangerous one, and it is the one that gets called a “nice weld” by anyone who only looks at the surface.
You don’t need an X-ray machine to catch the obvious failures. Clean the weld first — wire brush off the slag and spatter — then look hard at each of these:
Run down that list on any critical joint and you’ll catch most bad work before it gets painted — and before you pay for it.
This is where amateur judgment ends and standards begin. The governing reference for structural steel in the U.S. is the American Welding Society’s Structural Welding Code – Steel, AWS D1.1. It sets hard, numeric limits for visual acceptance rather than opinions. Under D1.1, cracks are always rejectable. Porosity, undercut, and reinforcement have measurable tolerances.
Take undercut. It is a notch cut into the base metal at the weld toe — a classic stress raiser that codes limit by depth and length, not by looks. The same visible undercut might pass on a decorative gate and fail on a load-bearing column. That’s why “it looks okay to me” is not an inspection.
Also note that “rejectable” and “repairable” are different words. A weld that fails inspection isn’t automatically scrap; it may simply need to be ground out and re-welded to a qualified procedure. What matters is that someone actually made that call instead of letting it slide.
One quiet driver of bad work is fatigue. Welders running under-shaded lenses all day get eye strain and headaches, and nobody inspects a joint carefully when they can barely see it. OSHA’s shade-selection table in 29 CFR 1910.252(b)(2)(ii)(H) lists shade 10 for 1/8-inch SMAW electrodes and shades 11–12 for gas-shielded processes. The right lens matters for more than comfort — see the welding PPE guide — it’s a quality issue too.
Paint, powder coat, chrome, and a fast hit of rattle-can black are the great equalizers. One smooth coat covers cracks, porosity, and cold lap in seconds. That’s why serious inspection happens before finish — and why “it’s already painted” should be a question, never a reassurance.
Almost every failed weld you are asked to look at comes down to a short list of recognizable problems. Learn their names and you stop arguing about whether a bead is “ugly” and start knowing whether it will hold. Each entry below gives the look, the cause, and the fix.
Note: NIOSH notes that confined-space welding can significantly increase exposure to manganese because fumes build up. The same trapped atmosphere that fogs your lens also traps porosity-causing contamination, so good welding PPE and ventilation protect the welder and the weld at once.
| Defect | Main cause |
|---|---|
| Porosity / worm tracks | Dirty metal or lost shielding gas |
| Undercut | Too much current, travel too fast |
| Overlap | Too little current, wrong angle |
| Lack of fusion | Arc never reaches the side wall |
| Lack of penetration | Too little heat at the root |
| Hot crack | Restraint plus wrong filler |
| Cold crack | Hydrogen and no preheat |
| Slag inclusion | Chipping skipped between passes |
| Reinforcement / underfill | Settings off, travel inconsistent |
| Misalignment / distortion | Bad fit-up and tack sequence |
| Burn-through / arc strikes | Excess heat, careless electrode handling |
Short Note: Three defects cause the most real-world failures: lack of fusion (invisible, passes a casual look), hot cracks (they start as micro-tears under load), and undercut (a permanent stress riser at the toe). If you only inspect for these three, you catch most disasters early.
The causes are never mysterious: dirty metal, wrong current, wrong travel speed, wind blowing shielding gas, wet electrodes, the wrong electrode for the alloy, and no preheat. These are setup and process failures, which is why OSHA’s welding standard requires completed welds to be visually inspected and free of visible cracks, lack of fusion, and lack of penetration, and why the American Welding Society publishes acceptance criteria rather than leaving it to taste.
Here’s the uncomfortable truth that most welder-hiring guides skip: a bad welding job is almost never a one-off accident. It is the predictable output of a process that was quietly skipped. Talk to any experienced inspector and you’ll hear the same thing from every direction — the cracked gate post, the leaking custom exhaust and the snapped trailer hitch almost always trace back to a handful of decisions made before the arc was ever struck. Welding punishes shortcuts harder than most trades, because the evidence hides inside the joint until the day the joint lets go.
Short Note: Ventilation discipline is a fingerprint of overall shop discipline. OSHA 1910.252 requires filler-metal containers to carry the notice: “Do not use in areas without adequate ventilation. See ANSI Z49.1… published by the American Welding Society.” OSHA 1926.353 requires mechanical or local exhaust ventilation whenever welding happens in a confined space, and airline respirators plus an outside attendant when ventilation can’t be achieved. Stick welding on stainless must meet toxic-metal rules because of dangerous nitrogen dioxide levels (1926.353(d)(1)(iv)). And manganese limits diverge sharply: OSHA’s PEL is a 5 mg/m³ ceiling, while NIOSH sets 1 mg/m³ TWA and 3 mg/m³ STEL and ACGIH lists a 0.02 mg/m³ respirable TLV. A shop that ignores fume control is almost always ignoring procedure discipline everywhere else — MIG welding safety and proper welding PPE are non-negotiable.
| Mistake | What it usually produces |
|---|---|
| Grinding the base metal with a dirty or oily disc | Embedded contamination, porosity, weld rejection |
| Welding over paint, rust, oil or mill scale | Gas holes, slag inclusions, toxic fumes |
| Using an undersized machine run at max duty cycle | Cold lap, no penetration, burn-through on thin metal |
| No purge on stainless root joints | “Sugared” root, corrosion failure later |
| Ignoring preheat on thick or high-carbon steel | Hydrogen (cold) cracking days after welding |
| Wrong polarity, contact tip size or drive-roll tension | Erratic arc, spatter, inconsistent feed |
| Running the wrong rod size for the amperage | Undercut, lack of fusion at the toes |
| Welding in a breeze without a screen | Shielding gas loss and porosity |
| Dragging damp rods out of a truck bed | Hydrogen pickup, worm tracks, cracking |
| Welding overhead unpracticed | Sagging beads, lack of fusion in 4F/4G work |
The expert insight worth remembering is simple: bad jobs are systems failures, not bad luck. The same shop that skips cleaning and ignores wind also skips preheat, runs unqualified welders and paints over welds before anyone can inspect them. Fix the process — a real WPS, correct settings such as proper E7018 amperage control, clean fit-up, controlled consumables, trained hands in the positions they’ve practiced — and the defect rate collapses on its own. The defects in the next section are simply what happens when these ten roots are left to grow.
People usually price a bad welding job by what they paid the welder. That number is almost never the real bill. The real bill arrives later: a cracked handrail, a scrapped assembly, a delayed shipment, an insurance letter, or a doctor’s appointment that better fume extraction would have avoided.
Welding is one of the few trades where a single mistake can cost fifty times the price of the original bead. Understanding why is the fastest way to stop treating weld quality as a cosmetic preference.
Removing a defective weld is expensive because it runs backwards. A good weld takes one pass forward; a bad one has to be cut, gouged or ground out, the base metal checked for hidden damage, the joint re-prepped and welded again. In practice, grinding out and re-welding a rejected joint routinely costs two to five times what the original weld cost, and it climbs fast if the part needs re-fixturing.
Then there is downtime: the assembly sits, the fitter waits, the paint booth idles. One rejectable weld can push an entire shipment, and one delayed shipment can sour a client relationship built over years. Scrapping is the extreme version: burn-through on thin sheet or a crack in a machined component often means the piece and the machine time go into the bin.
Quick Note: Warranty claims and reputation damage are the costs nobody quotes upfront: a shop that redoes work for free earns nothing either time. One publicly failed weld can cost more business than a decade of good work earned.
Money is recoverable. Injury is not. Structural failures trace back to a small set of defects: lack of fusion at the toe of a joint, an undercut that acts like a pre-cut crack, or a hot crack that grows under cyclic load. That is how a weld with visible undercut passes a glance and then fails on a coupler, a handrail, a frame rail or a pressure component. Handrails are the classic case: welded once, loaded suddenly by a falling person, rarely inspected again.
Pressure and pipe work is less forgiving still, governed by written procedures and third-party inspection rather than opinion. Get that discipline wrong and you can end up with a joint that looks right and metallurgically is not. Our walkthrough of pipe welding procedures, positions, passes and inspection shows how much process sits behind a single acceptable tie-in.
Legal liability follows the failure. If a welded assembly causes property damage or bodily harm, the party who signed off on it, or the employer who allowed unqualified work, can be held responsible. Regulators treat cutting, welding and brazing as fire-risk operations for a reason: OSHA’s 1910.252 standard exists because welding starts thousands of fires a year.
Three provisions of that standard are worth memorizing:
Skip these and the worst case is not rework but a structure fire or a vessel explosion. Sound MIG welding safety practice and correct welding PPE are the everyday version, and the OSHA welding, cutting and brazing page collects them in one place.
Welding fume is not a vague irritant. NIOSH reports that welders exposed to low manganese levels below 0.2 mg/m³ already performed more poorly on tests of brain function and motor skills. Prolonged exposure above 1 mg/m³ may lead to a Parkinsonian syndrome known as manganism. NIOSH researchers have also identified welder’s anthrax, a rare but serious occupational disease among metalworkers.
Those are career-ending diagnoses, not bad days. They come from years of uncontrolled exposure: no ventilation, no respiratory protection, no idea what the consumable puts into the air. It is exactly what a cheap, unqualified welding job never prices in.
Note: One $150 shortcut weld that cracks or triggers rework can cost more than a qualified professional’s full quote. Add downtime, scrap, liability and the fume exposure, and the “cheap” bid is almost always the expensive one.
The trade’s wage and injury picture is tracked by the U.S. Bureau of Labor Statistics, and acceptance standards come from the American Welding Society. Welding is a skilled, regulated, consequential trade; treating it as optional only moves the bill somewhere less visible.
You do not need a certification to spot a terrible weld. You need ten minutes, a bright light, and a willingness to look where the paint is. Most bad welding jobs fail visually long before they fail physically, which means the average buyer has a real chance of catching the problem before money changes hands. What you cannot do at home is prove that a weld is sound. That gap between “looks bad” and “proven good” is exactly where this section lives.
Start by stripping the evidence. Remove fresh paint, spatter, and grinding dust with a rag and a wire brush. Paint is the single best friend a bad welder has, because it levels the visual field and hides undercut and pinholes in one pass. Then work a flashlight along the joint at a low angle, the way raking light reveals scratches in a floor. Do not inspect a hot or freshly ground surface without protection: OSHA is explicit that 1926.353(e)(2) requires employees performing any type of welding, cutting or heating to be protected by suitable eye protective equipment, and that includes anyone standing nearby with a flashlight. Our Welding PPE: The Complete Guide to Protective Gear Every Welder Needs (2026) covers the basics.
Run this checklist in order:
Two cheap tools make the job far more reliable: a weld gauge (under $30, measures undercut depth, reinforcement height, and bead width against code limits) and a small inspection mirror for the back side of joints you cannot reach. Write down what you find with dates and photos before anyone applies a coat of paint.
Quick Note: Ask five questions before you pay: What WPS did you follow? What process and filler metal? What amperage and travel speed? Was preheat required and performed? Will you send photos of the raw, unpainted weld? A welder who answers instantly is a welder with a system. A welder who gets defensive is telling you something.
Visual inspection catches surface problems and nothing else. Everything internal requires instrumentation. The American Welding Society administers the Certified Welding Inspector (CWI) program, the most widely recognized welding inspection credential in the U.S., and that is who you hire when the joint matters.
| Method | Best at detecting |
|---|---|
| Visual (VT) | Profile, undercut, porosity, misalignment, workmanship |
| Magnetic particle (MT) | Surface and near-surface cracks in ferrous parts |
| Dye penetrant (PT) | Surface-breaking cracks and lack of fusion, any metal |
| Ultrasonic (UT) | Internal voids, lack of penetration, deep cracks |
| Radiographic (RT) | Internal voids, slag, porosity, incomplete penetration |
| Hydrostatic test | Leaks and weakness in pressure systems under load |
MT and PT find surface-breaking defects such as cracks and lack of fusion. UT and RT find internal defects such as voids and lack of penetration. Neither pair substitutes for the other. For pressure and pipe work, the governing procedure and inspection plan usually follow AWS D1.1 or ASME boiler code, which is the world our Pipe Welding Procedures: The Complete Guide to WPS, Positions, Passes & Inspection covers in detail.
Quick Note: A visual check is not enough when the weld is pressure-bound, structural, overhead, or hidden from both sides. Also remember that inspection happens at height: OSHA 1910.252(b)(1)(i) requires welders working on platforms, scaffolds or runways to be protected against falling. Same exposure applies to the inspector leaning over the joint with a mirror.
Not every bad welding job carries the same weight. A small spot weld on a body panel is a cosmetic headache; that same defect on a trailer coupler or a pressure vessel can be catastrophic. As the consequences of failure rise, so does the quality control required to prevent them.
The automotive world sees some of the most deceptive bad welding jobs, thanks to rushed repairs and cover-ups. From spot welds on unibody frames to custom exhausts, precision is a crash-safety requirement.
Quick Note: Tight vehicle confines amplify fume exposure, so welding PPE and ventilation matter as much as the weld itself.
Handrails, beams, and moment connections are held to a high standard. A single undercut or cold lap creates a stress riser that can jeopardize an entire frame.
E7018 electrodes are a structural staple, but without correct parameters and fit-up, even good filler metal produces a bad weld.
Transmission girth welds endure immense pressure, corrosion, and temperature swings for decades, so this industry runs some of the strictest QA/QC regimes on Earth.
Short Note: Cross-country lines do use specialized joining methods, but arc welding dominates girth welds, where purging and preheat prevent cracking.
Pressure vessels hold volatile substances under force. Fatigue, creep, and improper repairs can cause a violent rupture with little warning.
For toxicity concerns flagged by NIOSH, shops handling exotic alloys need local exhaust ventilation and, when required, airline respirators, particularly with beryllium-bearing materials under 1926.353(c)(2)(iv).
Shipyards are brutal welding environments: wind, moisture, confined spaces, and enormous scale combine into a perfect storm for bad welds.
This is where everyday people most often meet a bad welding job. These items are load-bearing, yet frequently fabricated with little or no formal QA.
Note: Cleanliness is king everywhere. OSHA 1926.352 requires chlorinated solvents kept at least 200 feet from the arc unless shielded, since UV radiation decomposes them into toxic byproducts.
| Industry | Common Defect | Consequence | Who Checks It |
|---|---|---|---|
| Automotive | Lack of fusion / Burn-through | Crash structure failure | OEM Certified Technicians |
| Structural Steel | Undercut / Cold lap | Building load failure | AWS CWI |
| Pipeline & Oil & Gas | Lack of penetration / Cracking | Leaks or explosions | API Inspectors & NDT Specialists |
| Pressure Vessels | Hot cracks / Underfill | Catastrophic rupture | ASME Authorized Inspectors |
| Shipbuilding/Repair | Porosity / Distortion | Hull failure at sea | NAVSEA/QC Inspectors |
| Consumer Fabrication | Cold lap / Lack of fusion | Sudden load failure | End-user Visual (or 3rd party) |
The takeaway: the higher the consequence of failure, the more formal the QA/QC must be. From a tiny MIG tack weld to a pressure vessel seam, cutting corners always costs more than doing it right the first time.
Here is the uncomfortable truth about hiring a welder: you usually cannot tell a great weld from a terrible one after the paint dries. Both look smooth. Both come with a confident promise. The difference shows up six months later, in the rain, under load. So the only real defense you have is the hiring conversation itself, and most people walk into it unprepared.
When you know what to compare, bad-job shops and good-job shops stop sounding alike almost immediately.
| Dimension | Bad welding job shop | Good welding job shop |
|---|---|---|
| Certification | “My uncle taught me” or a card nobody can verify | AWS performance-based qualification records for welders and procedures |
| Written procedure | Settings chosen by feel, different every job | A Written Procedure Specification (WPS) with qualified ranges |
| Fit-up and cleaning | Welded over mill scale, paint, and oil | Ground, degreased, fit-tight, and tack-welded before the first pass |
| Machine capability | One undersized buzz box for everything | Correct process and output for the material, including adequate duty cycle |
| Filler traceability | Whatever rod was in the drawer | Matching filler, heat numbers, and batch records kept with the job |
| QA documentation | None, or a verbal “yeah, it’s good” | Photos, inspection records, and a signed acceptance |
| Photos before paint | Refuses or never takes them | Sends them unprompted |
| Warranty | Verbal, or none at all | Written terms with clear exclusions |
| Price | Lowest bid on the board | Higher, itemized, and explained |
Notice that certification sits at the top of that list for a reason. AWS qualification documents are the standard way fabricators demonstrate welder and procedure capability to clients and inspectors, and they are the closest thing to a proof of competence you can ask for.
Run the math once and it never looks the same again. A hack quotes $150 for a structural bracket. A qualified fabricator quotes $400, including a WPS, fit-up, and inspection. Six months later the cheap bracket fails, and you pay $900 to have it cut out, the base metal restored, and a proper weld put in, plus downtime and whatever the failure destroyed. Total: $1,050 for work that should have cost $400 the first time.
Low bids are rarely efficiency. BLS occupational wage data shows welding pay spans a wide range, and a bid far below market almost always reflects unqualified labor rather than a smarter process.
Quick Note: Ask what the welder is being paid, not what you are being charged. If the labor behind your pressure-bearing weld is priced like fast food, the workmanship will be too.
A good shop answers all ten without hesitation. A bad shop deflects at question three. Professionals also treat authorization as paperwork, not improvisation: OSHA 1910.252 places responsibility on management to designate an individual who authorizes cutting and welding operations, preferably through a written permit, and supervisors must determine the combustible materials and hazards present before work begins. That habit of documenting before striking an arc separates planned work from improvisation.
Also watch how they talk about setup and safety. A fabricator who takes equipment maintenance seriously, from cable connectors to gas lines, usually takes joint quality seriously too.
Ask to see the process in action as well. Shops that explain their MIG machine setup or their pipe welding procedures without getting defensive are telling you they have nothing to hide. That transparency, not the price tag, is the real predictor of a weld that will still be holding when it matters.
Finding a bad weld is the easy part. What happens next decides whether you end up with a stronger joint or a hidden failure waiting under a fresh coat of paint. Most people’s first instinct is to run another bead over the top. That is the single worst thing you can do, and it is worth understanding why before we get into the correct sequence.
A new pass laid over porosity, slag inclusion, or an unfused toe does not fix the original defect. It traps it. The trapped void stays exactly where it was, now sealed under a bead that looks acceptable on the surface, and it becomes a stress concentrator inspectors cannot see without removing metal. The same logic applies to undercut at the toe of a previous pass: filling it cosmetically without grinding a smooth transition leaves a notch that keeps propagating. As a rule, if the first weld was rejected, the material that made it rejected has to come out. There is no shortcut around that, which is why undercut acceptance limits matter before you ever pick up a grinder.
Short Note: A pretty bead over unfused metal is not a repair, it is a disguise. If you cannot explain what is underneath the new weld, you do not have a repair — you have a cosmetically improved failure.
Two methods are accepted in fabrication practice: grinding and carbon-arc gouging. Grinding is the go-to for thin material, visible surface defects, and anything where heat input has to stay controlled. Carbon-arc gouging is faster and deeper, and it is the practical choice for heavy sections and thick pipe where you need to reach the root of a lack-of-penetration defect. Where gouging burns into the base metal, that damaged layer is then ground away to clean, bright steel. If burn-through has thinned the plate beyond the design wall thickness, you are no longer in repair territory — you are in replacement territory.
Before any preheating, gouging, or cutting on a closed section, OSHA is explicit: 1910.252(a)(3)(ii) requires all hollow spaces, cavities, or containers to be vented before preheating, cutting, or welding, with inert-gas purging recommended as well. Skipped on a trailer tongue, a sealed tube, or a former fuel container, that step causes both bad welds and explosions. Hot work around combustibles carries its own rules under OSHA 1926.352, including clearance of combustibles and a fire watch.
Once the bad metal is out, the joint goes back to fit-up: correct gap, correct land, correct bevel, members aligned within tolerance. Grind the edges bright, remove spatter and cutting residue, degrease, and dry any consumables. Then re-establish your ground clamp placement and confirm machine settings against the procedure — this is where people go back to guessing. Pull the actual parameters from your written procedure rather than dialing by feel; if you are working with low-hydrogen stick, our 3/32 E7018 amperage and polarity guide covers the settings that keep hydrogen cracking from recurring.
Re-welding should follow the same qualified procedure specification used on the original joint — preheat temperature, interpass temperature, filler classification, heat input, and travel sequence. On code work, that means following the WPS and documenting the welder and parameters. Pipe welding procedures show how tightly procedure, position, and inspection are linked once a joint carries pressure.
This is the step most often skipped, and it is non-negotiable: the repair must be re-inspected using the same method that rejected the original weld. A weld rejected visually needs a visual recheck. A weld rejected by radiography needs RT again. A weld rejected by ultrasonic testing cannot be signed off on a flashlight inspection. Re-running the same method is the only way to confirm the removal was complete and the new weld is sound through its full volume.
Repair Checklist:
Some defects should never be patched. Cracks in fatigue-critical areas, hydrogen-cracked joints (which require full removal, drying, and controlled preheat), load-bearing welds in pressure-containing parts, and base metal that has been burned through or significantly thinned all demand replacement of the component rather than a re-run. The same applies to vehicle crash structures and trailer frames where a previous repair has already removed original material. When in doubt, ask whether the part can be proven to its original design strength. If the answer is no, replace it, and keep the written warranty and inspection record on file — your proof that the second job was done right.
Everything in this guide boils down to one idea: bad welding jobs are preventable. Nobody wakes up planning to build a weld that fails under load. Failures happen because a step was skipped, a parameter was guessed, or nobody bothered to look before the paint went on. The good news is that prevention is cheap, and it almost always happens before the arc is struck. Below is the exact pre-job routine we recommend for homeowners, fabricators, and shop owners alike, followed by the shop-side controls that separate a professional operation from a driveway hustle.
Quick Note: OSHA’s 1910.252 tells management to establish designated cutting and welding areas and to insist that welders and their supervisors are trained. That is a buyer-verifiable item: ask where the hot work happens and who trained the person doing it. If the answer is vague, keep your wallet closed.
If you run a shop, the same discipline runs in the other direction:
Quality and safety are the same habit. NIOSH’s welding fume guidance sets the REL for manganese at 1 mg/m³ TWA with a 3 mg/m³ STEL, which means ventilation is engineering, not preference — pair it with the gear in MIG welding safety precautions. OSHA’s filter-lens shade table under 1910.252(b)(2)(ii)(H) calls for shade 12 in gas-shielded ferrous welding and shade 14 for large-diameter SMAW; an under-shaded welder gets fatigued, and fatigued welders drift out of parameter. Finally, OSHA’s hot-work rules require 35-foot clearance of combustibles and a fire watch for at least 30 minutes after welding stops. Non-negotiable finishing steps, every job.
Short Note: Nobody reads a checklist the first time it saves them. Print it, tape it to the bench, and use it until it’s habit.
Back to the hitch. It was never a metal problem or a design problem; it was a process problem. No clamping, no cleaning, no parameter, no inspection, no record — a weld that photographed beautifully and held nothing. That is what a bad welding job really is: a preventable failure wearing a decent coat of paint. Walk through the checklist, demand the paperwork, look at the raw bead, and release payment on acceptance. Do that, and you won’t be pulling anyone’s load out of a ditch.
A bad welding job is any weld that fails visual, code, safety, or service expectations — even if it looks neat under paint. Cracks, porosity, undercut, and lack of fusion all fall into this category because they weaken the joint and can cause failure under load.
Porosity (gas holes), undercut, overlap or cold lap, lack of fusion, lack of penetration, slag inclusions, hot and cold cracking, excessive reinforcement, misalignment, and burn-through are the ones inspectors see most. Lack of fusion and cracks are the most dangerous because they can be invisible from the surface.
Clean the weld, shine a bright light across it, and look at the toes of the weld for notches (undercut), holes (porosity), and any crack lines, plus check that both sides show a fused, consistent bead. If the joint is structural, pressure-bearing, or hidden after paint, hire a Certified Welding Inspector for VT, PT/MT, UT, or RT testing.
Yes. A failed weld can drop a handrail, split a trailer frame, rupture a pressure vessel, or release flammable product, and hot-work mistakes cause fires — OSHA requires 35-foot combustible clearance and a 30-minute fire watch after welding. Long-term fume exposure from poor ventilation also carries neurological health risks.
Repair usually costs several times the original weld because the bad metal must be ground or gouged out, the joint re-prepped, rewelded, and re-inspected. Replacing an entire part — a trailer frame, gate assembly, or vehicle section — can run into thousands of dollars, far more than paying for qualified work upfront.
Not for anything structural, load-bearing, pressure-related, or safety-critical. Certification shows the welder passed a performance test for a specific process and position, and professional shops can also supply a Written Procedure Specification and references.
Many bad welds can be repaired by grinding or gouging out the defective metal down to sound material and rewelding to a qualified procedure, then re-inspecting. Cracks in fatigue-critical areas, burned-through base metal, and pressure-boundary welds often require replacing the section instead.
Undercut is a groove melted into the base metal along the weld toe, which creates a stress concentration that can start a crack. Overlap (cold lap) is filler metal that rolled over the base metal without fusing — both are rejectable, but they have opposite causes: too much heat and travel speed for undercut, too little heat for overlap.
Ask for certification documents, a written quote, the process and filler metal being used, photos of the weld before paint, insurance, references, and a written warranty. Red flags include no paperwork, immediate painting, refusal to show intermediate photos, and pressure to skip inspection.
At minimum: a helmet with the correct filter shade (OSHA’s guide lists shade 10 for 1/8-inch SMAW and shades 11–12 for gas-shielded processes), gloves, fire-resistant clothing, eye protection for helpers, and adequate ventilation or respirators. OSHA requires local exhaust or general ventilation to keep toxic fumes below allowable limits, with airline respirators and an outside attendant in confined spaces.