Really Bad Welds: Causes, Dangers, Identification, and How to Fix Them

Picture a critical support bracket failing under load because of a hidden crack no one noticed during inspection. The consequences of really bad welds aren’t just cosmetic—they can risk lives, projects, and reputations in an instant.

Introduction: What Exactly Are Really Bad Welds?

A crew in the Midwest once shipped a set of handrail brackets that looked perfect on the shop floor: bright, evenly rippled beads with clean toes. Three weeks after installation, a visitor leaned on the rail and one bracket cracked clean through at the root. The weld had zero penetration. It looked finished, but it was never actually joined. That is the honest definition of a really bad weld — not an ugly one, but one that cannot do the job it was made to do.

This is why appearance alone is a dangerous quality check. Almost every welder has laid a bead that was cosmetically rough yet structurally sound, and almost every welder has seen a visually tidy weld hiding a serious internal flaw. Welding quality is judged against acceptance criteria written into codes, and the industry standard most often cited for steel structures is AWS D1.1, the Structural Welding Code – Steel, published by the American Welding Society. Those codes spell out exactly how much undercut, porosity, or incomplete fusion a finished joint can carry before it stops being acceptable.

Discontinuity vs. Defect: The Line That Matters

Welding engineers draw a sharp distinction between a discontinuity and a defect. A discontinuity is any interruption in the normal structure of the weld — a small pores, a tiny slag trap, a shallow arc strike. It exists, but it does not automatically mean the part is bad. That discontinuity becomes a defect only when it exceeds the acceptance limits defined by the applicable code and compromises the joint’s fitness for service.

In plain terms: a weld can be imperfect and still be perfectly legal and safe. The same weld, in a pressure pipe or a dynamic load-bearing joint, may be completely unacceptable. Context determines everything — the code being followed, the service conditions, the load type, and the inspection level specified on the drawings.

Why “Really Bad” Goes Beyond the Weld Itself

A truly bad weld creates consequences that ripple outward along three axes:

• Structural integrity. Cracks, lack of fusion, and incomplete penetration act as stress risers. Under load, they grow. The joint may hold today and fail tomorrow with no warning, which is why hidden cracking is one of the most feared findings in inspection.
• Code compliance. Work that fails acceptance criteria cannot be certified. That means rejected welds, hold points on the schedule, mandatory re-inspection, and in worst cases, project shutdowns until repairs are proven sound.
• Real-world cost and trust. Rework is expensive — cutting out, grinding, re-welding, and re-testing — and poor welds increase liability for fabricators, contractors, and shops across the fabrication and repair industries. Beyond the invoices sits reputational damage: a shop known for porosity problems stops getting calls.

The safety dimension is even less forgiving. Defective welds in structural steel, piping, pressure vessels, and mobile equipment can lead to collapses, leaks, fires, and exposures. Regulatory bodies take this seriously: OSHA’s welding, cutting, and brazing standards and the health guidance published by NIOSH exist because welding failures hurt people. Even the PPE a welder relies on — covered in our guides to welding PPE and what a welder wears — only protects against the hazards of the process itself, not against a joint that fails once the structure is in service.

What This Guide Covers

This article walks the full arc of the problem. We will identify the major weld defects and how to spot them, dig into the root causes that produce them, and examine the safety dangers a bad weld creates. From there we will cover the standards and codes that define acceptable quality, the inspection methods used to catch problems early, and the repair and prevention strategies that keep work moving. Real-world cost examples, practical hands-on tips for welders, and a final look at building a quality-first culture round it out.

Whether you are a beginner still learning how to set up a MIG welder or an experienced hand troubleshooting parameters like 3/32 E7018 amperage, the goal is the same: welds that are sound, certified, and still standing years from now. Understanding what makes a weld really bad is the first step toward never making one.

Common Types of Really Bad Welds and How to Identify Them

A weld does not need to look rough to be dangerous, and it does not need to look polished to be safe. For this reason, spotting really bad welds requires more than a quick glance. The surface may be flawless, but issues like cracks, trapped gas or a lack of fusion can be hiding beneath it. By learning to recognize the most common defects, you can catch them before they compromise the strength of a joint.

Cracks: The Most Dangerous Defect of All

Out of all the flaws that can form in a weld, cracks are the most severe. Most welding codes treat them as automatic rejectable discontinuities, as their sharp tips create intense stress concentrations. Under load, that stress can quickly propagate, leading to a sudden failure. In cold conditions or under impact loads, this can even result in brittle fracture.

Cracks generally fall into three categories:

• Hot cracks form while the weld metal is still solidifying, often in the center of the bead.
• Cold cracks (hydrogen cracks) can appear hours after welding. These are often caused by hydrogen absorbed from moisture in electrodes or contaminated surfaces, paired with high restraint.
• Crater cracks form at the stop point when a bead is ended abruptly instead of being properly filled.

Visually, cracks appear as sharp, linear separations that can branch out. Since cold cracks can be delayed, a weld that looks fine at the end of a shift may crack overnight. For this reason, methods such as magnetic particle and dye penetrant inspection are essential for structural work.

Lack of Fusion and Incomplete Penetration

Lack of fusion occurs when the weld metal fails to bond to the base metal or to a previous pass. Incomplete penetration, on the other hand, happens at the root when the weld fails to melt all the way through the joint’s thickness. Both defects are incredibly deceptive, as the bead can look perfectly acceptable on the surface while drastically reducing the load the joint can carry. A flat or concave profile with no sidewall wash, an unfilled root gap or a hollow sound while grinding can all be red flags. Since these issues are internal, radiographic and ultrasonic testing are the most reliable ways to detect them.

Porosity: Trapped Gas Bubbles

Porosity takes the form of small pinholes or wormholes on the surface, or rounded dark spots on a radiograph. It is caused by gas becoming trapped in the solidifying metal, often from inadequate shielding gas coverage, contaminated base metal, oil, rust, paint or moisture in the electrode. While surface porosity is easy to spot, subsurface porosity often goes unnoticed. Some processes are more forgiving, and with the right precautions covered in our guide to flux core wire welding tips, many of these gas-related issues can be prevented.

Undercut

Undercut is a groove melted into the base metal along the weld toe that the filler metal never fills back in. This thins the plate at a critical point and acts as a stress raiser. It is especially common in high-amperage MIG welding, out-of-position stick welding or when travel speed is too fast. To learn more about its acceptance limits and repair methods, take a look at our full breakdown on undercut: causes, acceptance limits, and fixes.

Slag Inclusions and Spatter

Slag inclusions occur when flux or slag gets trapped between weld passes or inside the weld metal, most often in multi-pass stick and flux-core welds when interpass cleaning is skipped. They weaken the cross section and appear as dark, irregular shapes on X-rays. Spatter, on the other hand, is mostly cosmetic in nature. These small metal beads signal unstable arc conditions, incorrect parameters or a dirty contact tip, but they can still hide other flaws, so cleaning them is never a waste of time.

Distortion and Overlap

Distortion is the warping that comes from uneven heating and cooling. A part can be metallurgically sound, yet still be dimensionally useless and throw an entire assembly out of tolerance. Overlap, sometimes called “cold lap,” is filler metal that spills over the base metal without fusing to it. This is a common beginner issue, especially for those just getting started with stick welding, as we cover in Is Stick Welding Hard?.

A Quick Identification Cheat Sheet

The ability to spot these defects is only half of the battle. The next step is understanding why they happen in the first place, from machine settings to joint preparation, so that they can be prevented before the arc is ever struck.

Root Causes: Why Really Bad Welds Happen in the First Place

Nobody walks up to a joint intending to ruin it. Yet when a weld cracks days after inspection or a bracket fails in service, the cause almost always traces back to something preventable that happened before the arc was struck. Every defect has a fingerprint: porosity points to contamination, undercut points to technique, and cracks often point to hydrogen and restraint. Learn to read those fingerprints and you stop guessing and start fixing.

Here are the root causes that show up again and again in failure investigations.

1. Improper Welding Parameters

Amperage, voltage, wire feed speed, and travel speed are the four dials that control everything about a weld pool. Set them wrong and the metallurgy suffers before you ever see a crack.

• Too much heat input creates a wide, shallow bead with a coarse heat-affected zone (HAZ), reducing toughness and making the joint more prone to cracking under load. It also causes burn-through on thin material.
• Too little heat produces cold lap, lack of fusion, and incomplete penetration, where the filler sits on top of the base metal instead of bonding with it.
• Wrong travel speed directly changes penetration and bead profile. Move too fast and you get a tall, narrow bead with no fusion at the toes; move too slow and you flood the joint, causing overlap and distortion.

These settings are process- and position-specific. Stick electrode amperage for a 3/32 E7018 differs from a 5/32 E7018, and overhead positions demand their own adjustments, as covered in 3/32 E7018 Amperage: Settings, Polarity & How to Run It and Overhead Stick Welding. Ignoring those differences is a fast track to a rejected weld.

2. Poor Joint Preparation

Roughly half of all weld problems start at the joint itself. Incorrect bevel angles, insufficient root faces, wrong root openings, and skipped fit-up checks all distort heat distribution. A joint tacked unevenly or forced into alignment creates locked-in stress that shows up later as cracking. Even travel direction matters: push or pull stick welding changes penetration depth, so the wrong technique quietly undermines fusion.

3. Contamination: Oil, Rust, and Moisture

This is one of the most common and most underestimated causes. Oil, grease, mill scale, rust, paint, and moisture each attack the weld differently:

• Oil and grease vaporize in the arc and release gas that gets trapped in solidifying metal, causing porosity and spatter.
• Rust and mill scale introduce oxygen and inconsistent conductivity, producing unstable arcs, inclusions, and unpredictable penetration.
• Moisture is the classic hydrogen source. Hydrogen diffuses into the weld metal and HAZ, then gathers at stress points to cause delayed cracking, also called hydrogen-induced cracking. This is why low-hydrogen electrodes like E7018 must be stored in rod ovens and why understanding electrode classifications matters more than most beginners realize.

4. Operator Error and Training Gaps

Technique errors, such as the wrong arc length, improper stickout, erratic travel speed, or poorly overlapped weave beads, account for a large share of rework in fabrication shops. The uncomfortable truth is that operator technique and procedure adherence, not equipment quality, drive most quality variability, which is exactly why certification and requalification programs exist. If you are early in your journey, an honest look at is stick welding hard will save you from expecting mastery in a weekend.

5. Equipment Problems

Bad welds are sometimes purely mechanical. Worn contact tips, kinked or undersized cables, loose welding connectors, contaminated gas nozzles, and dirty filters all degrade performance in ways the welder may never suspect. Loose connections push resistive heat into the cable instead of the arc, delivering less current than the dial shows. Both Lincoln Electric and Miller publish troubleshooting guidance that puts maintenance first whenever weld quality suddenly drops.

6. Environmental Conditions

Wind, temperature, and humidity are silent saboteurs. Wind above roughly 5 mph can blow away shielding gas, causing porosity almost instantly in open-air MIG or TIG work. Cold base metal raises cooling rates, hardness, and hydrogen cracking risk. And running hot settings on light gauge leads straight to burn-through, as covered in stick welding thin metal. Enclose, preheat, or adjust accordingly.

7. Material Incompatibility

Finally, the wrong filler for the base metal is a defect waiting to happen. Dissimilar metals can form brittle intermetallic compounds, and carbon steels welded with incorrect alloy filler can suffer hot cracking. Welding copper, for instance, demands specific parameter control as outlined in spot welding copper. When in doubt, verify against a qualified welding procedure specification (WPS) before striking an arc.

Every one of these causes is detectable, correctable, and preventable with preparation and process discipline, which is exactly what the next sections on inspection and repair build toward.

Safety and Structural Dangers of Really Bad Welds

When a weld fails, the consequences rarely stay in the shop. From a cracked support bracket to a porous seam on a pressure vessel, really bad welds can lead to structural collapse, fires, toxic exposure, and life-changing injuries. These aren’t just cosmetic flaws. They’re safety hazards that demand immediate attention.

When Load-Bearing Welds Fail

Bridges, crane booms, building frames, excavators, storage tanks, and piping all rely on welded joints to transfer heavy loads. When those joints are compromised by undercut, lack of fusion, or cracking, they behave more like a pre-existing fracture than a solid connection. Under repeated or sudden stress, these defects will grow, and in many cases, they will fail with little to no warning.

These failures are often hidden from view. A truss, pipeline elbow, or hull can appear sound while a critical defect lurks below the surface. This is exactly why organizations like the American Welding Society treat weld quality as a safety-critical discipline. The structural requirements outlined in AWS D1.1 exist to prevent these brittle, sudden breaks in the first place.

Fire and Explosion Risks

Welding doesn’t just create a bond. It creates sparks, slag, and molten spatter that can travel surprising distances. Even a sound weld can be a fire risk if hot work is performed near flammable materials, but a defective one multiplies that danger. A cracked or porous weld on a fuel or natural gas line can leak for weeks or months before an ignition source turns it into a major event.

This level of risk is why services like gas line welding are held to some of the strictest safety standards. On these jobs, a marginal weld isn’t a production setback. It’s a public safety threat. Proper purging, joint prep, and inspection aren’t optional. They’re non-negotiable.

Toxic Fumes and Exposure Hazards

Poor welds don’t just endanger the end user. They endanger the welder. Bad technique often means excessive heat, longer arc time, and repeated rework, all of which drastically increase fume generation. These fumes can expose welders to metal oxides, hexavalent chromium from stainless steel, manganese, and flux decomposition products.

The NIOSH welding safety research program notes both acute effects, like metal fume fever, and chronic risks, including lung disease and neurological impacts. To combat these hazards, the OSHA welding, cutting, and brazing standards require ventilation, exposure monitoring, respiratory protection, hot work permits, fire watch, and proper personal protective equipment.

Rework is one of the biggest culprits. A welder forced to grind out and re-run a rejected joint three times breathes far more fume than one who gets it right the first time. From arc flash burns to eye and hand injuries, the risks only grow without the right gear. Our guides on weld burns: causes and first aid and what welders should wear go more in depth on these everyday protections.

Regulatory and Legal Consequences

When a bad weld causes harm, the fallout goes far beyond the cost of a repair. Failing an inspection under the ASME Boiler and Pressure Vessel Code can halt an entire project. Contractors or welders who work outside of approved procedures can face OSHA citations, stop-work orders, and in serious cases, criminal liability for negligence. These consequences aren’t reserved for mega-projects either. Mobile repairs, agricultural equipment, and handrails can all carry the same liability.

Codes like AWS D1.1, ASME, and API don’t exist to slow down work. They exist because experience, often through tragedy, proved exactly what happens when weld quality becomes optional. Every measurable limit for reinforcement, porosity, undercut, or testing is there to prevent the dangers listed above.

The Human Cost

Behind every failure is a person who trusted that joint. A rigger trusting a lifting lug. A technician opening a pressure vessel. A family driving across a bridge. For welders, that responsibility is at the heart of the trade. Quality isn’t bureaucracy. It’s care.

Whether you’re dialing in your E7018 amperage settings for structural stick, following proven pipe welding procedures, or laying down a clean MIG tack weld, your goal is the same: leave a joint that will never become the reason for an incident.

Industry Standards and Codes That Prevent Bad Welds

On any job, a guess can cost far more than a second look. The difference between a weld that passes inspection and one that fails comes down to an agreed set of rules. Industry codes and standards remove opinion from the equation by laying out exact limits for things like undercut, porosity, and penetration. These documents are what allow an inspector in one shop to judge a weld the same way as one across the country, and that consistency is what keeps really bad welds out of service.

The Major Standards Bodies Behind Weld Quality

In North America, a handful of organizations set the rules most welders will run into:

• AWS (American Welding Society) — publishes the D1 series of structural welding codes and is widely considered the backbone for structural steel work. You can learn more at the American Welding Society.
• ASME (American Society of Mechanical Engineers) — focuses on pressure-retaining equipment, boilers, and power piping. Its scope covers materials, fabrication, and fitness for service through its codes and standards catalog.
• ANSI (American National Standards Institute) — approves and coordinates national standards, including safety documents like ANSI Z49.1, Safety in Welding, Cutting, and Allied Processes. You can browse approved documents at ansi.org.
• API (American Petroleum Institute) — sets prescriptive rules for pipelines, storage tanks, and refinery equipment. More details can be found at api.org.

AWS D1.1: The Structural Steel Benchmark

For structural steel, the most common reference is AWS D1.1, the Structural Welding Code — Steel. It defines base metal requirements, preheat, welding positions, visual acceptance criteria, and the limits for discontinuities such as porosity, undercut, and cracks. With those limits in place, an acceptable weld can be judged against the code, not a person’s opinion. For a closer look at how undercut is measured and judged, check out our breakdown of welding defects: undercut causes, acceptance limits, and how to fix it.

Any code-compliant job also relies on paperwork. A Welding Procedure Specification (WPS) acts as the recipe for the joint, while a Procedure Qualification Record (PQR) proves that recipe works through testing. Our pipe welding procedures guide explains how these documents, positions, passes, and inspection all tie together for pressure pipework.

ASME Section IX: Qualifying Procedures and Welders

While AWS D1.1 focuses on the finished weld, ASME Section IX focuses on the people and processes that make it. Section IX covers procedure qualification (PQR), welder and welding operator performance qualification, essential variables like heat input and position, and continuity rules. Under Section IX, a welder’s qualification lapses if they have not welded within that qualified process for six months. ASME Section V rounds this out by standardizing the nondestructive examination methods used to verify joints. For a broader technical overview, TWI’s technical knowledge base is an excellent reference.

API, ANSI, and Process-Specific Codes

Different industries layer in their own rules. API 1104 governs pipeline welding, while API 570 and 653 cover in-service piping and tank inspection. Gas transmission and distribution work comes with strict requirements for materials, purging, and testing, all of which we cover in our gas line welding codes, safety, and purging rules guide. On the safety front, OSHA’s welding, cutting, and brazing standard (29 CFR 1910.252) mandates ventilation, fire protection, and personal protective equipment. These requirements pair well with the gear outlined in our welding PPE complete guide.

Inspection Requirements That Keep Welds Honest

Codes don’t just say what to build, but also who can judge it and how. Most jobs follow a clear inspection path:

1. Visual inspection first — Surface flaws like porosity, undercut, overlap, profile issues, and identification marks are checked before advanced methods.
2. NDE by service duty — Radiographic, ultrasonic, magnetic particle, and penetrant testing are assigned based on the code’s examination category.
3. Qualified personnel — Examinations must be performed by certified individuals, such as AWS CWIs or those meeting ASNT SNT-TC-1A.
4. Traceability — Heat numbers, WPS references, and inspector sign-offs are retained for the life of the asset.

For more on how these expectations are framed, Welding Digest offers helpful background from AWS.

Why Compliance Matters

Codes are written with consequences in mind. A bridge, pressure vessel, or gas line can fail in service, and someone must answer for it. Following the applicable code is often the standard legal defense that a weld was designed, executed, and examined by competent people using accepted methods. Beyond liability, it means a predictable service life, insurable assets, and fewer catastrophic failures. That level of accountability is exactly what stops really bad welds from ever reaching the field.

Inspection Methods to Catch Really Bad Welds Early

The difference between a weld flagged in the shop and one that fails in the field usually comes down to inspection. Catching really bad welds early is far cheaper and safer than finding them after a structure is loaded or a pipeline is pressurized. Inspectors have a well-proven toolkit for finding defects on the surface and deep inside the metal.

Visual Inspection: The First Line of Defense

Every serious inspection program starts with the eyes. Visual inspection (VT) is the simplest, fastest, and most widely used method, and it catches more defects than all the advanced techniques combined. An experienced inspector checks weld profile, undercut, overlap, porosity, spatter, cracks, and incomplete fill, often with nothing more than a light, a mirror, a magnifier, and a simple gauge. That is why reputable codes make VT mandatory before any other test proceeds.

Good visual inspection is a skill, not a glance. Inspectors need proper lighting (typically at least 500 lux for close work), surfaces clean of slag and spatter, and knowledge of the acceptance criteria in the governing code. If you are still training your eye to catch defects before an inspector does, our guide to undercut causes and acceptance limits is a practical starting point.

Non-Destructive Testing: Seeing Inside the Weld

Surface inspection cannot tell you what hides in the root of a thick pipe weld. Non-destructive testing (NDT) examines welds without damaging the part, so the component can still go into service. Four methods cover most needs:

• Radiographic Testing (RT) uses X-rays or gamma rays to image the weld’s interior. Cracks, porosity, and lack of fusion appear as dark indications on film or a digital detector. RT finds internal discontinuities well but demands strict radiation safety controls and access to both sides of the joint.
• Ultrasonic Testing (UT) bounces high-frequency sound waves through the weld and reads the echoes. Flaws reflect sound differently than sound metal, so UT detects internal cracks and lack of fusion, including in material RT struggles with. Phased-array UT has made the method faster and more precise.
• Magnetic Particle Testing (MT) magnetizes the part and sprinkles fine iron particles over the surface, where they cluster at surface and near-surface cracks. It works only on ferromagnetic materials like carbon steel.
• Penetrant Testing (PT) applies a colored or fluorescent liquid that seeps into surface-breaking defects, then draws it back out with a developer. It works on any non-porous material, including stainless steel and aluminum.

RT and UT reveal internal discontinuities without destroying the part, which makes them the workhorses of pipeline, pressure vessel, and structural inspection. The American Welding Society publishes guidance on these methods, and its AWS D1.1 Structural Welding Code – Steel is the reference most structural inspectors work from.

Destructive Testing: When Learning the Truth Means Cutting

Sometimes you need to know exactly how strong a weld is, and that means sacrificing a sample. Bend tests, tensile tests, and macro-etch examination fall into this category. In procedure and welder qualification, coupon specimens are welded under production conditions, then cut and tested to prove the process performs as claimed. You cannot do this to every production weld, but a good quality program uses coupons to validate the procedure before production starts. For more on what qualification involves, see our overview of pipe welding procedures, WPS, and inspection requirements.

Qualified Inspectors and Written Procedures

Tools alone do not catch bad welds. Qualified inspectors do, working from written procedures aligned with recognized standards. Most U.S. certifications follow the American Society for Nondestructive Testing (ASNT) framework, with personnel qualified to Levels I, II, or III depending on responsibility. AWS and ASME both specify which NDT methods are acceptable for given applications, while API standards govern inspection in petroleum and pipeline work. An unqualified inspector following an undefined process is a recipe for a really bad weld slipping through.

When to Inspect: Timing Matters

Timing matters as much as technique. A crack in a root pass can be repaired cleanly before later passes cover it; the same crack found after the joint is finished may require gouging out nearly the whole weld. Smart quality programs call for in-process inspection between passes on critical joints, especially in pipe welding, overhead positions, and heavy plate work where rework gets ugly fast.

Building an Inspection Routine You Can Trust

You do not need a laboratory to start catching defects early. Build the habit in layers: clean the weld, look at it critically under good light, question anything that looks off, and escalate to NDT when the code demands it. Inspection is also part of a broader safety culture. Inspectors in fabrication shops should carry proper protective gear since they work alongside live welding operations, and radiographic crews must follow radiation safety protocols far beyond standard OSHA welding, cutting, and brazing requirements.

The bottom line: an uninspected weld is an unverified weld. Whether the job calls for a visual check or a full multi-method NDT program, building inspection into the workflow is the most effective way to stop really bad welds before they carry a load.

Repair, Remediation, and Prevention Strategies

A bad weld found during inspection is rarely a death sentence for the joint. Most defects can be removed and re-welded until the connection passes code again. The catch is that a repair is itself a welding procedure, and improvising one turns a single defect into two. A sound repair follows the same discipline as the original weld: an approved procedure, a qualified welder, controlled heat, and documented inspection.

Assess and Approve Before You Strike an Arc

Before anyone picks up a grinder, a qualified inspector needs to define what is wrong and how far it extends. Visual inspection catches surface problems, while dye penetrant or magnetic particle testing reveals cracks the eye cannot see. The type, size, and depth of the defect decide whether the area can be repaired at all. Cracks in members under dynamic load, for example, frequently need an engineer’s sign-off before work begins.

Repairs must also be documented against the governing standard. Structural steel repairs typically fall under the AWS D1.1 Structural Welding Code, pressure piping and boilers under ASME codes and standards, and pipeline or hazardous-service work under API standards. Whichever applies, approve the repair procedure before work starts and hold the fix to the same acceptance criteria the original weld had to meet. For regulated piping, the purge, fit-up, and documentation rules stay in force during a repair exactly as they do on first pass, as our gas line welding codes, safety, and purging rules guide explains.

Remove the Defect Down to Sound Metal

Light grinding suits shallow surface flaws such as a minor undercut, but deeper problems need carbon-arc gouging or careful arc-air cutting to reach sound base metal. The defect has to be eliminated, not buried under a cap pass. Feather the edges so the new bead ties in gradually, and chase any associated indications until they disappear. A crack left at the root of a repair will simply restart under load.

Verification after gouging is quick and cheap: a shot of magnetic-particle or penetrant testing confirms you are back to clean metal before the welder repositions.

Reweld With the Heat Under Control

Preheat and interpass temperature control are the two levers that stop repair welds from cracking, especially on restrained joints, heavy sections, and carbon steels sensitive to hydrogen-induced cracking. Preheat drives out moisture and slows the cooling rate; interpass limits prevent the next bead from quenching the one beneath it. Both values must come from the written procedure, not from habit or rule of thumb, because they shift with thickness, steel chemistry, and ambient temperature.

Hydrogen control matters just as much. Use low-hydrogen consumables stored in a dry oven or holding quiver, and confirm your settings against a proven reference such as our 3/32 E7018 amperage, polarity and technique chart or the broader guide to welding electrodes. Where the procedure calls for it, post-weld heat treatment relieves residual stress and helps diffusible hydrogen escape. PWHT is not optional when the code or the WPS requires it, and it must follow an approved thermal cycle with recorded temperature readings.

Note: Repairs often land in awkward positions. If the defect sits overhead or out of position, refresh your technique with our pipe welding procedures guide on WPS, positions, passes, and inspection before committing to the joint.

Prevention Costs Less Than Any Repair

Every repair bill points back to the same root causes, and a strong quality control program is what keeps them from recurring. The practical building blocks are:

• A written WPS. Qualified procedures with supporting procedure qualification records give welders the correct range of amps, voltage, travel speed, heat input, and preheat, so operators never have to guess at parameters.
• Verified machine setup. Dial in the feeder, gas flow, liner, and ground before the first bead. Our MIG welder setup guide and flux core welding tips cover the settings that most often drift, while welding connectors address the loose connections that cause erratic arc behavior and hidden porosity.
• Qualified, trained welders. Performance qualification ensures the person holding the torch can make the joint in the required position, and continuing practice keeps the skill current.
• In-process inspection. Inspect between passes rather than only at the end, when rework means cutting out good metal. Log welder IDs, consumable batches, and parameters for traceability.
• Root-cause analysis on every rejection. Record why the defect happened, then fix the process, not just the weld.

None of this works without the right protective equipment. Repair work involves gouging, grinding, and extra arc time, meaning more spatter, UV, and fumes than a routine pass, so review welding PPE and protective gear and your MIG welding safety precautions before starting. Skin injuries are common during clean-up; our weld burns on arms guide covers treatment and prevention. Employers also owe workers a program meeting OSHA’s welding, cutting, and brazing requirements that addresses fume exposure identified by NIOSH welding health research.

The bottom line is straightforward: repair to an approved procedure, verify the removal, control the heat, then invest in the written procedures, training, and QA/QC routines that stop the same defect appearing on the next joint. Rework always costs more than doing it right the first time, and the reputation damage of a failure travels far beyond the cost of the consumables.

Real-World Examples, Cost Impact, and Lessons Learned

Abstract warnings about weld quality rarely change behavior. Money, downtime, and near-misses do. Across industries, the pattern is consistent: rework and scrap from weld defects significantly increase fabrication costs, and the projects that recover fastest are the ones that treated inspection and training as investments rather than expenses. Let’s walk through some realistic scenarios and pull out the lessons that actually stick.

Scenario 1: The Structural Bracket That Failed a Mill Test

A small fabrication shop took on a rush order for support brackets welded from 3/8-inch plate. Production moved fast, but the root passes were run cold to avoid burn-through on a joint that was actually over-designed for the heat input. The first batch looked acceptable after a quick visual check. Then a load test on the customer’s side failed, and ultrasonic testing revealed lack of fusion along several root passes. The brackets had to be cut out, scrapped, and re-fabricated.

The material cost was minor compared to the labor: roughly three times the original welding hours, plus expedited shipping, plus a customer audit that forced the shop to document every procedure. The shop’s fix was unglamorous but effective: written weld procedure qualification, mandatory fit-up checks, and operator training on pipe welding procedures and inspection sequencing. Defect rates dropped, and so did the stress.

Scenario 2: Porosity in a Sheet Metal Run

Consider a thin-gauge sheet metal job where a team chased speed over gas coverage. Wind from an open shop door disrupted the shielding gas, and the welds came out porous and sputtered. The root cause was mundane: a compromised gas line connection and open-air draft, not welding skill. It’s the kind of failure covered in our welding connectors guide, where a loose or leaking connection quietly ruins an entire afternoon of work.

The lesson here is that defect causes are often upstream. Checking hoses, regulators, and draft control before striking an arc costs minutes. Chasing porosity after the fact costs a full day of grinding and re-welding, plus the risk of a customer rejecting the batch. On thin material, that grinding often causes burn-through anyway, creating a second defect from the repair itself. Our stick welding thin metal guide covers exactly that trap.

The Cost Math Nobody Budgets For

Industry surveys have long shown that fabrication shops spend a meaningful share of total labor hours on rework, and weld defects are consistently among the top contributors. The costs stack up in layers:

• Direct labor: cutting out, grinding back, and re-welding defective joints, often in overhead or out-of-position positions that are slower and harder.
• Inspection overhead: re-inspection fees, additional NDT passes, and documentation for code compliance.
• Downtime: a single rejected weld on a critical line can idle fitters, painters, and erectors who were scheduled downstream.
• Reputation: the quietest cost, and the largest. One structural failure traced to a shop’s welds can close doors that no marketing budget reopens.

Defect prevention, by contrast, is comparatively cheap. A well-written procedure, proper parameter settings (as detailed in our 3/32 E7018 amperage guide), and consistent MIG welder setup cost hours, not weeks.

Scenario 3: A Pipeline Repair That Became a Shutdown

In pressure-bound work, a bad weld isn’t just rework, it’s a forced outage. A fabricated piping spool that fails hydrostatic testing halts commissioning entirely. According to TWI, defects such as lack of fusion, cracking, and slag inclusions are among the most common reasons welds are rejected during NDT in oil, gas, and power work. Each rejection triggers a documented repair procedure, often requiring preheat, controlled grinding, re-welding, and sometimes post-weld heat treatment. A repair that could have been prevented by correct root gap and current settings can cost tens of thousands once the outage clock is running.

Lessons Learned From the Field

Looking across these cases, three lessons repeat themselves:

1. Inspect early, not after assembly. Catching a lack-of-fusion root pass before the cap goes on saves a complete tear-out. Visual inspection as defined by AWS D1.1 is cheap and catches more than people expect.
2. Train for the specific joint, not general “welding.” Overhead positions, thin sheet, and dissimilar metals each have their own failure modes. Practitioners should review technique-specific resources like our overhead stick welding guide before tackling unfamiliar work.
3. Protect the operator while fixing defects. Repairs mean extra grinding, fumes, and burn risk. Proper welding PPE and attention to OSHA welding, cutting, and brazing standards are non-negotiable during remediation.

The shops with the best reputations don’t avoid defects through luck. They measure rework, fix root causes, train deliberately, and treat every rejected weld as a free lesson. Prevention always costs less than the failure it prevents.

Practical Tips for Welders to Avoid Really Bad Welds

Ask any seasoned welder what separates a career of clean, passing welds from a career of rework, and talent rarely comes up. It comes down to a short list of habits repeated on every joint: set the machine honestly, prep like the weld depends on it, control your heat, and inspect your own work before the inspector does.

Set Up the Machine Before Striking an Arc

Many defects start before the hood goes down. Loose or oxidized cable connections steal amperage, so the number on the dial is not what the arc actually receives. Clean your threads, check the ground clamp, and inspect your liner and contact tip. On MIG, repeat the same startup sequence every time: wire feed speed, voltage, stick-out, gas flow, drive roll tension. A step-by-step resource like How to Set Up a MIG Welder turns that sequence into muscle memory.

Joint Preparation and Cleaning: The Highest-Return Habit

Porosity, lack of fusion, and slag inclusions almost always trace back to metal that was not ready. Grind away mill scale, rust, oil, paint, and moisture at least an inch back from the joint face. Cut bevels to the specified included angle, land, and root opening, and check them with a gauge rather than your eye. Tacks must be sound, since a cracked tack becomes a crack initiator when the main bead runs over it. On structural work, the AWS D1.1 Structural Welding Code requires base metal cleaning in the weld area, and inspectors do check it.

Heat Input and Travel Speed: Control Them on Purpose

Heat input sits on three variables you hold: amperage or voltage, travel speed, and pass count. Run hot and slow on thin material and you blow through; run cold and fast and you get cold lap and lack of fusion. TWI defines heat input as energy delivered per unit length of weld, which is why identical settings behave differently the moment you change travel speed. Know your consumable’s window before you strike up. A 3/32 E7018 amperage guide gives the correct range for a given diameter and polarity, keeping you out of the too-cold zone where hydrogen cracking starts.

Read the puddle instead of the clock. A narrow puddle that will not wet the sidewalls means you are moving too fast or running too cold; an orange, sagging puddle means too much heat or lingering in one spot.

Small Technique Details That Decide the Result

• Arc length: keep it roughly the diameter of the electrode core. Long arcs pull in atmosphere and cause porosity and spatter.
• Work and travel angle: hold them consistently, and decide before you start whether you will push or pull.
• Consumable choice: match the rod or wire to the steel, position, and thickness. A guide like What Are Electrodes in Welding? explains how to read the AWS classification.
• Pass sequencing: balance heat from side to side, and back-step where distortion matters.

Tack Smart, Then Troubleshoot Before the Next Pass

Skipping or misplacing tacks quietly causes distortion and fit-up drift. Space them so the joint cannot pull open, and place them where sound weld metal will consume them; the MIG Tack Weld Guide covers spacing, settings, and sizing. Between passes, clean every bead and stop at the first sign of porosity or undercut, fixing the cause rather than the symptom. Undercut is one of the most common acceptance failures, so read Welding Defects Undercut: Causes, Acceptance Limits, and How to Fix It before the inspector writes you a rejection slip.

The Pre-Flight Welder’s Checklist

Welders who consistently pass inspection work from a short written checklist instead of memory:

1. WPS or parameter sheet in hand, machine settings verified.
2. Joint cleaned, bevels and gap measured, fit-up confirmed.
3. Consumables correct, dry, and within shelf life.
4. Cables, connectors, gas lines, and PPE inspected before the arc starts.
5. First pass visually checked before the second pass is laid.
6. Finished bead cleaned and graded against the code’s acceptance criteria.

Self-inspection catches most defects while they are still cheap to fix, and it builds the habit of grading your own work by the standard an inspector will use. Pair it with proper welding PPE so discomfort never tempts you to rush a pass you should be watching.

None of these habits need a better machine or a steadier hand. They need consistency treated as the actual skill. Do the small things in the same order every time, and really bad welds stop being something you find at inspection.

Conclusion: Building a Culture That Prevents Bad Welds

Every weld tells a story. Sometimes it’s the story of a welder who checked the procedure, cleaned the joint, dialed in the machine, and laid a bead they’d be proud to sign. Other times it’s the story of a rushed Saturday night job, a contaminated plate, and a crack that nobody caught until the bracket buckled. Throughout this guide, we’ve walked through what really bad welds actually look like, why they happen, how they get discovered, and what it costs when they don’t. The closing lesson is simple but powerful: preventing bad welds is never just one person’s job. It’s a culture.

The Three Pillars of Sound Welds

If there’s one idea to carry into every job from here on out, it’s this: strong, code-compliant welds rest on three pillars that must work together.

• Skilled welders. Technique matters. Whether you’re running a 3/32 rod or dialling in 3/32 E7018 amperage settings, operator skill determines penetration, bead profile, and defect rate.
• Approved procedures. A Welding Procedure Specification (WPS) isn’t paperwork for its own sake. It locks in the parameters that are known to produce sound welds for that exact joint, material, and position.
• Rigorous inspection. Visual checks, non-destructive testing, and hold points catch what the eye misses. As we covered in pipe welding procedures, inspection is baked into quality code work from the first root pass to final sign-off.

Remove any one of those pillars and the structure fails. A brilliant welder working without a WPS is guessing. A perfect WPS ignored by an untrained operator is a document no one reads. And inspection without skilled tradespeople just produces a long list of rework nobody wants to pay for.

Why Culture Beats Checklists

Checklists help, and we gave you a solid one in the practical tips section. But checklists only work when people actually believe in them. A welder who sees quality as “the inspector’s problem” will cut corners the moment supervision walks away. A welder who understands that a hidden crack in a support bracket can injure a colleague, a customer, or a stranger will stop, clean the joint, and check their settings without being asked.

Think about the safety side of the equation for a moment. Welding carries real hazards beyond the weld itself: arc radiation, hot metal, electrical risk, and toxic fumes. The same discipline that produces clean welds also protects people. Wearing proper protective equipment, following the complete welding PPE guide, and practising MIG welding safety precautions aren’t separate concerns from weld quality. They’re the same mindset applied twice.

Standards Are Your Safety Net

Nothing in this article is a substitute for the codes themselves. The AWS D1.1 Structural Welding Code remains the backbone for steel fabrication in the United States, and the wider American Welding Society catalog covers everything from sheet metal to bridges. ASME codes govern pressure-containing work where failure is catastrophic by definition. ANSI and API standards round out the framework for everything from pipeline service to equipment safety.

Regulators back those codes up. OSHA’s welding, cutting, and brazing standards set the legal floor for workplace safety, while resources from NIOSH keep the science on fumes and exposure current. Manufacturers like Lincoln Electric, Miller Welds, and Hobart Welders publish application guides, and technical resources such as TWI and Welding Digest are excellent places to keep learning.

Your Path Forward

Here’s a practical closing checklist to carry off this page:

1. Write or verify your WPS before the arc ever strikes, especially for structural or code work.
2. Prep like it matters — clean metal, correct bevels, dry consumables, and proper storage.
3. Know your equipment inside out, from MIG welder setup to welding connectors that won’t fail mid-job.
4. Inspect your own work first. Grinding and re-welding beats explaining a failure later.
5. Keep learning. Whether it’s flux core technique, TIG sheet metal settings, or overhead stick fundamentals, every skill you sharpen reduces the odds of a defect slipping through.

Really bad welds are rarely caused by one dramatic mistake. They’re the compound interest of small oversights: skipped cleaning, unchecked settings, unverified consumables, and unasked questions. The good news is that the same compounding works in reverse. One welder who insists on a clean joint, one inspector who refuses to wave through a suspect bead, one shop that budgets time for proper fit-up. Stack enough of those together and bad welds don’t just decrease. They become the exception, and eventually, they disappear from your shop floor entirely.

That’s the culture worth building. It protects the people doing the work, the clients trusting the result, and the public who never sees the weld until it’s holding something they depend on. Start with the next weld in front of you. Prep it, run it, check it, and own it.

Frequently Asked Questions

What makes a weld considered really bad?

A weld is considered really bad when discontinuities exceed acceptance criteria in applicable codes, compromising structural integrity or safety. Visual flaws alone may not be defects, but cracks or lack of fusion often are.

What are the most dangerous welding defects?

Cracks are typically the most dangerous due to their tendency to propagate under stress. Incomplete penetration and lack of fusion also critically reduce joint strength in load-bearing applications.

How can I visually identify bad welds?

Look for cracks, undercut, porosity, excessive spatter, overlap, or inconsistent bead profiles. Compare against visual acceptance criteria in standards like AWS D1.1 for guidance.

What causes porosity in welds?

Porosity is often caused by contamination from oil, rust, moisture, or improper shielding gas coverage. Proper cleaning, gas flow, and dry electrodes help prevent it.

Can really bad welds be repaired safely?

Many bad welds can be repaired by removing defective material and rewelding using an approved procedure. Critical repairs should follow code requirements and be inspected by qualified personnel.

How do industry codes prevent bad welds?

Codes like AWS D1.1, ASME Section IX, and API standards set requirements for procedures, welder qualification, and acceptance criteria, ensuring consistent, safe weld quality.

What safety hazards come from poor welding?

Hazards include structural failure, fire risks, and exposure to welding fumes and gases. OSHA and NIOSH provide guidance on controls to protect workers and the public.

When should I use NDT instead of visual inspection?

Use NDT (UT, RT, MT, PT) when internal defects are possible or when code requires it for critical joints. Visual inspection alone can’t detect subsurface discontinuities.

What role does training play in preventing bad welds?

Proper training ensures welders understand procedures, parameters, and defect prevention. Ongoing education reduces errors and supports a culture of quality and safety.

How much can bad welds cost a project?

Costs include rework, scrap, downtime, inspection, and potential liability. Prevention through QA/QC, training, and inspection is far more cost-effective than correcting failures.