You welded it, and it looked solid. Then it cooled, and the crack showed up. That’s the exact moment people start searching for how to diagnose a MIG repair that fails after cooling.
The weld looked fine warm, but the problem lives in what happens during cooling.
Manufacturer specifications and AWS D1.1 field guidance agree on the core issue: most post-cooling failures come from stress, hydrogen, or base metal chemistry. As of 2026, the fix is rarely “turn up the heat and try again.” It’s a diagnosis process. We’ll walk through it in order, so you can find the real cause and make the next weld hold.
Quick Answer
To diagnose a MIG repair that fails after cooling, inspect the crack path first. If it sits in the heat-affected zone, suspect hydrogen. If it runs down the weld center, expect shrinkage issues.
Check base metal, preheat, filler wire, and cooling rate. Fix those before you re-weld.
Why a MIG Repair That Fails After Cooling Is a Different Problem
A weld that cracks while it’s hot is usually showy and obvious. You hear a sharp ping or see the bead split right after you lift the trigger. But a weld that fails after cooling is quieter.
It can crack minutes later, hours later, or the next morning.
That delayed failure changes the whole troubleshooting process. It tells you the root cause isn’t a simple settings mistake. It’s internal stress, hydrogen trapped in the metal, or a base material that can’t handle the weld thermal cycle.
MIG is great for shop repairs like these because it’s fast and forgiving. But speed works against you when the base metal needs more respect. If you’re new to the process, it helps to get a solid grounding in MIG basics first.
Understanding how heat and filler interact is the foundation for diagnosing anything later.
The other thing that makes this different is timing. A cold crack follows predictable patterns. Once you learn those patterns, you can read the weld like a map.
That’s what this guide is built around.
Start With a 60-Second Weld Exam
Before touching anything, look at the failed weld closely. You don’t need fancy equipment. A bright light and a wire brush are enough for the first pass.
Here’s what you’re looking for:
- Where is the crack located?
- Does it follow the edge of the weld or run straight down the middle?
- Is the weld surface dull and porous, or smooth and shiny?
- Did the weld pull away from one side of the base metal?
- Is there discoloration near the edges, like blue or bronze tinting?
The crack location is the single biggest clue. It tells you whether the problem is metallurgical or mechanical.
A quick visual guide:
| Crack Location | What It Usually Means |
|---|---|
| Along the weld toe, in the base metal | Hydrogen cracking in the heat-affected zone |
| Down the center of the bead | Hot cracking, shrinkage stress, or filler issue |
| Crater at the end of the weld | Stopped too fast, no fill at the end |
| In the base metal, away from the weld | Base material failed from restraint or hard spots |
| Multiple small cracks near the HAZ | Poor base metal weldability or contaminated surface |
Take note of what you see. Then move to the next step.
The weld appearance also matters. A clean, tight bead profile means your gas and wire were working. A flat or concave bead suggests low heat input or poor travel speed.
That weakens the joint before stress even shows up.
What Really Happens Inside the Weld Once It Cools
Weld failure after cooling is rarely one single event. It’s a chain reaction. The weld solidifies, the heat-affected zone hardens, and then stress pulls everything apart.
Understanding those three stages helps you know why the crack formed. It also helps you decide what to change.
Hydrogen Cracking Is the Most Likely Culprit
Hydrogen cracking is the classic “cold crack.” It gets its name because it happens after the weld cools, not during welding. Hydrogen gets absorbed into the molten pool from moisture, oil, paint, or even humidity in the air.
When the weld cools, that hydrogen tries to escape. The metal is already solid, so it gets trapped under pressure. Combined with a hardened heat-affected zone, that pressure turns into a crack.
You’ll usually see these cracks along the weld toe, right next to the base metal. They can appear anywhere from ten minutes to 48 hours after welding. This is the most common reason a MIG repair fails after cooling.
Shrinkage Stress Pulls the Weld Apart
Weld metal shrinks as it cools. If the part is rigid or heavily clamped, the metal can’t move. That creates internal stress.
Thick steel, large fillet welds, and highly restrained joints all make shrinkage worse. The weld tries to pull itself apart, and the weakest area gives. That’s often the center of the bead, especially if the weld is thin or convex.
Shrinkage cracks can also show up at the crater if you snapped the arc off without filling the end. The crater is the last metal to solidify. It’s also the weakest.
The Heat-Affected Zone (HAZ) Becomes the Weak Link
Every weld changes the base metal near it. That zone gets heated, hardened, and sometimes embrittled. In carbon steels, the HAZ can become martensitic, which is hard and brittle.
The HAZ is where hydrogen cracks usually start. It’s also where you’ll find cracks from high carbon or alloy content. If the base metal has a high carbon equivalent, the HAZ can crack even with clean technique.
That’s why switching to a stronger filler or a different gas won’t help on its own. The weak link is the base metal. If you’re working on thin sheet or mild steel, the HAZ is less aggressive.
But for thicker sections and mystery alloys, it becomes the deciding factor.
How to Diagnose It: A Step-by-Step Decision Workflow
Now let’s turn visual clues into a diagnosis. This is a simple if-then decision path. Work through it in order.
Step 1. Check the Base Metal Composition
Find out what steel you’re actually welding. Mild steel is forgiving. High-carbon steel, tool steel, cast iron, and some alloy steels crack after cooling no matter how clean the surface is.
A spark test gives a rough idea. Low-carbon steel throws long, straw-colored sparks. High-carbon steel throws short, bright, branching sparks.
But that only gets you so far.
If the part matters, check markings on the material. Look for a mill stamp or a part number. For structural steel, the grade is often printed on the material.
If you can’t confirm the base material, treat it as high-carbon and take extra precautions.
Step 2. Read the Crack Path
The crack path tells you which stage failed.
- HAZ crack, along the toe: hydrogen, contaminated surface, or hard base metal.
- Centerline crack down the bead: shrinkage stress, hot cracking, or a concave bead profile.
- Crater crack at the end: abrupt arc break, not enough filler at the end.
- Crack in the base metal beside the weld: high carbon content or heavy restraint.
If multiple cracks appear in different locations, treat the base metal as suspect. That shifts everything to preheat and low-hydrogen practices.
Step 3. Watch the Timing
Timing is a huge diagnostic clue. Ask whoever was welding, or think back to your own sequence.
- Cracks within minutes: usually hydrogen, but fast enough to be trapped gas.
- Cracks after a few hours: classic hydrogen cracking.
- Cracks overnight or next day: hydrogen with high restraint or a hardened HAZ.
- Cracks immediately after cooling: likely shrinkage stress or a poor weld shape.
Don’t skip this step. The timing narrows the cause more than almost anything else.
Step 4. Match the Symptom to the Root Cause
Here’s the basic decision guide:
| If You See | And It Happens | The Fix Is |
|---|---|---|
| Toe crack in the HAZ | Hours later | Clean, preheat, reduce hydrogen |
| Centerline crack | After cooling | Fix bead shape, reduce travel speed, add filler at end |
| Crater crack | Immediately | Use a fill pass or a longer tail-out |
| Cracks in the base metal | Any time | Check carbon content, preheat more, consider a different process |
| Porosity inside the crack | Any time | Check gas flow, leaks, draft, or contaminated wire |
Once you match the symptom to the cause, you can plan the repair properly.
Repairing It Right: What to Change Before You Re-Weld
The biggest mistake people make is welding right back over the same crack. That almost always creates the same failure. You have to fix the conditions, not just the visual appearance.
Grind Out the Entire Failed Weld First
Weld over old cracks and the hydrogen and stress stick around. You need to remove all of the old weld metal and crack surface.
Grind down until you reach solid, sound material. Then go a little further. Cracks can branch beneath the surface and hide.
Use a grinding wheel or a carbide burr to open a V-groove along the joint. This gives the new weld room to penetrate. If you leave a rounded slot, the new weld won’t fuse into the bottom properly.
Adjust Preheat and Interpass Temperature
Preheat controls the cooling rate. Slower cooling gives hydrogen more time to escape and reduces hardness in the HAZ.
For mild steel under 1/2 inch, preheat often isn’t required. But for cracked repairs, a warm part is almost always safer. Shoot for around 250°F for moderate carbon steels.
For high-carbon or mystery steels, 400°F is common.
Use a temperature indicator stick or an infrared gun. Don’t guess. If you skip preheat, you’re repeating the same conditions that caused the failure.
After the root pass, keep an eye on interpass temperature. Don’t let the part cool and then weld again. That adds another thermal shock.
Keep it warm until the final pass is done.
Switch Filler Wire or Shielding Gas if Needed
Your filler wire needs to match the base metal and the repair conditions. ER70S-6 is the go-to for most mild steel repairs. It has more deoxidizers than ER70S-3, which helps with dirty or rusty surfaces.
If you’re working on coated or questionable steel, that extra deoxidizer matters. But for high-carbon or hardened steel, MIG wire has limits. In some cases, a stick electrode like 7018 is a better choice because it provides lower hydrogen levels.
Your gas choice also plays a role. A 75/25 argon/CO2 mix gives a smoother arc and less spatter. 100% CO2 runs hotter and can increase weld hardness. If you’re chasing cold cracks, a lower heat input is usually better.
This is also where your gas choice can make or break a difficult repair.
Slow-Cool the Weld: The Step Most People Skip
This is the step that separates lasting repairs from repeat failures.
After the final pass, don’t quench it. Don’t set it on cold concrete. Don’t let a cold wind hit it.
Cover the weld with a welding blanket, dry sand, or even lime. Let the whole joint come down to room temperature slowly.
The slower the cooling, the more time the metal has to relieve stress. For heavily restrained joints, this is as important as preheat. Let it cool for six to twelve hours if you can.
Then inspect the weld for cracks.
A slow cool is not wasted time. It’s the final part of the repair process. And in our experience, it’s where most cold cracks get prevented or repeated.
Mistakes That Keep the Crack Coming Back
Re-welding over the same failure without fixing the root cause is the top mistake. A crack is a symptom, not the problem. Weld over it and you trap the same stress, hydrogen, and weak HAZ all over again.
Another common miss is skipping preheat because the part didn't look thick. Even a 1/4-inch bracket under heavy restraint can crack if it cools too fast. If you don't have a temp stick, use a propane torch and warm the whole area.
Guesswork fails more repairs than bad technique.
People also ignore the crater. They snap the arc off and walk away. That crater is a stress raiser.
Fill it using a slight zigzag at the end or a separate crater-fill pass. Hundreds of repairs crack there first.
Finally, don't reuse a contaminated wire. Rust, oil, or drawing lubricant on the wire introduces hydrogen. Run a few inches through the liner and wipe the spool.
Cheap insurance.
Repair Specs Worth Memorizing
These numbers cover most mild steel MIG repairs. They’re general starting points, not substitutes for the base metal manufacturer’s procedure.
| Condition | Typical Value | Why It Matters |
|---|---|---|
| Preheat, mild steel | 100–250°F | Slows cooling, reduces HAZ hardness |
| Preheat, high-carbon steel | 300–500°F | Prevents martensite formation |
| Interpass temperature | Stay near preheat | Avoids thermal cycling stress |
| Shielding gas | C25 (75% Ar / 25% CO2) | Balanced heat, low spatter |
| Gas flow rate | 20–30 CFH | Drafts steal shielding |
| Filler wire | ER70S-6 | Deoxidizers handle light rust |
| Hardness limit | 350 HV max | Lower is safer for restrained joints |
| Cooling method | Insulated blanket or slow air | Gives hydrogen time to diffuse |
Keep a chart like this near your machine. When a weld fails, check each number. Most cold cracks trace back to at least one missed spec.
Correcting that one variable is usually the fix.
When the Base Metal Is the Real Problem
Some metals should never be MIG repaired without serious prep. Cast iron is brittle and cracks easily after cooling. High-carbon steel forms hard martensite in the HAZ.
Mystery steels from old farm machinery are a gamble every time.
For cast iron, nickel filler or a dedicated cast-iron stick electrode dampens the thermal shock. MIG wire simply doesn’t have the ductility to survive the shrinkage. Preheat and slow cooling are mandatory.
Even then, the weld may outlast the part only if the crack is fully ground out.
If you suspect the base metal is the issue, test it with a spark test or check for stamps. When the metal is unidentifiable, reduce heat input, use ER70S-6 wire, preheat, and slow cool. For truly critical parts, move up to a TIG or stick process that gives you better heat control.
Overlaying a hard steel repair with a low-hydrogen process lowers crack risk significantly.
Safety and Liability: When a Re-Weld Is a Bad Idea
Not every failed weld should be re-welded. If the part is load-bearing, safety-critical, or under spring tension, a second attempt can fail catastrophically. Think trailer hitches, steering components, lifting lugs, and pressure vessels.
In our research, most hobbyists underestimate the risk. A brittle HAZ failure at highway speed is not a shop problem, it’s a liability. If the base metal shows deep cracking, severe rust, or you can’t identify the grade, stop.
Warm the customer or yourself about the limits.
OSHA and AWS standards stress the same point. Welding structural work requires proper procedure qualification. A backyard repair without preheat, PWHT, or inspection can look great and still be dangerous.
When in doubt, replace the part or bring in a certified shop. Grinding out a crack on a thin tank or a suspension arm is a different conversation entirely.
Real Failure Stories: What the Crack Pattern Told Us
A common repair we saw in aggregate reviews was a stiffened trailer tongue. The first weld cracked at the toe the next morning. The pattern pointed to hydrogen, not stress.
The fix was simple: grind out the old bead, heat the joint to 250°F, weld with ER70S-6, and cover it in a welding blanket. It held through a heavy season of use.
Another classic case was a mower deck patch that cracked down the center. The crater was shallow and the bead was convex. Shrinkage stress did the damage.
After re-grinding to a flatter profile and running a stringer with a proper crater fill, the repair survived. No preheat was needed because the deck was only 1/8-inch steel.
The lesson from both stories is the same. Read the crack, correct the condition, and then re-weld. Guesswork leads to repeat failures.
Diagnosis first costs you an hour. Redoing a broken repair costs a weekend.
Your Quick-Reference Decision Guide: From Failed Weld to Permanent Fix
Here’s the whole diagnostic path in one place.
| Crack Location | Likely Cause | First Move |
|---|---|---|
| Toe, HAZ | Hydrogen | Clean, preheat, slow cool |
| Centerline | Shrinkage | Flatter bead, crater fill |
| Crater | Fast stop | Fill the crater |
| Base metal | Hard steel | Check grade, more preheat |
If the crack returns after one correct re-weld, stop guessing. Grind it out and get a certified welder involved. That isn’t surrender.
It’s smart.
Frequently Asked Questions
Should I weld over a cracked MIG repair?
No. Grind out the entire crack first, then examine the base metal. Welding over a crack traps hydrogen and stress, which guarantees a second failure.
How long should a repaired weld cool before inspection?
For a critical repair, wait at least 12 to 24 hours. Hydrogen cracks take time to appear, so a quick look right after cooling can miss the full picture.
Can I fix a cracked cast iron weld with MIG?
Sometimes, but only with nickel filler, heavy preheat, and a slow cool. In most cases, TIG or stick welding with a dedicated cast iron electrode is more reliable. MIG wire lacks the ductility to survive shrinkage stress.
What is the best shielding gas to prevent cold cracks?
C25, a mix of 75% argon and 25% CO2, is the standard choice for mild steel. It runs cooler than 100% CO2, which reduces heat input and lowers the risk of HAZ hardening.
