Diagnosing MIG and Flux-Core Weld Defects

You can already name porosity, undercut and lack of fusion. A mid-level welder goes further and finds out why the flaw appeared, without waiting for someone else to figure it out. That matters because the fix for one cause can make another one worse. Turning up the gas to cure porosity, for example, can pull in more air. Diagnosis is a method you can learn: read the pattern, form a theory, test one thing, and prove the result.

What You Will Be Able to Do

• Use where and when a flaw appears to narrow down its cause.
• Work through porosity, fusion and cracking problems in a logical order.
• Tell a machine problem from a technique or material problem.
• Prove a fix with a fillet break test or a bend test.
• Know when to stop welding and call your lead or inspector.

A Method That Works

1. Stop and look. Do not weld over a flaw to see if it goes away.
2. Get the facts. When did it start? Does it happen on every weld, or only on some? Is it at the start, the end or all along? Did anything change, such as a new spool, a moved machine or a different welder?
3. Check the simple things. Most problems come from gas, contact tips, connections, cleanliness and stickout.
4. Form one theory. Pick the most likely cause, based on the pattern.
5. Change one thing. Make a test weld on scrap of the same metal, in the same position.
6. Prove it. A nice-looking bead is not proof. Break it, bend it or inspect it.
7. Write it down. Note the cause and the fix, because it will happen again.

Stay inside your Welding Procedure Specification (WPS) while you do this. If the fix needs something outside the WPS, take your findings to your lead.

Read the Pattern First

Where and when a flaw shows up is your best clue.

Porosity

Porosity means gas got trapped in the weld. The gas has to come from somewhere, so you are looking for a source or a loss of shielding.

Work through it in this order, cheapest first:

1. Look for wind. A fan, an open door or a fume hood too close to the arc can strip away shielding gas. Remember that raising the flow does not fix a draft, and it can make turbulence that pulls in air.
2. Check the cylinder and flow. Is the cylinder near empty? Is the flowmeter reading what the WPS says? Flux-cored gas flow commonly runs around 35 to 50 CFH, and solid-wire flow is often lower.
3. Check the gun end. Look for spatter in the nozzle, a loose nozzle, or a cracked diffuser.
4. Check for leaks. A leak in the hose or gun can pull air into the gas stream. Spray soapy water on the fittings with the gas on. Do not use oil-based products on gas fittings.
5. Check the metal. Oil, paint, rust, galvanizing and moisture all make gas in the arc. Cutting fluid from the saw is a common cause. Clean back from the joint.
6. Check the wire. Flux-cored wire can pick up moisture when it sits out, and one maker's data sheet says flat out that redrying is not recommended. Keep wire covered and replace a spool that has been left in damp air.
7. Check your stickout. Too long a stickout moves the nozzle away from the arc. On many self-shielded wires, too short a stickout causes porosity as well.

Reading Flux-Core Surface Marks

On gas-shielded flux-cored wire, wormlike marks on the bead surface, often called gas marks or worm tracks, usually mean too much gas or the wrong voltage. They differ from pores. Lower the flow to the right range and check your voltage. If you see tiny holes as well as marks, treat it as porosity and work through the list.

Lack of Fusion

This is the most dangerous flaw on this list to miss, because it hides. Solid-wire MIG leaves no slag to warn you, and a cold bead can look smooth and neat.

Common causes:

The trap is the puddle. If you slow down to make a bead look better, the puddle can run ahead of the arc and cushion it, so the arc never reaches the steel. A bigger, slower bead can be worse than a smaller, faster one.

Slag Problems on Flux-Core

Slag inclusions on flux-cored welds nearly always come from one of a few things:

• Slag running ahead of the arc. Check your gun angle and your travel speed.
• Convex beads with deep toes. The next pass cannot wash out the notch. Grind the high spots first.
• Poor cleaning. Chip and brush every pass, and look along the toes.
• Wide weaves. Slag gets left at the edges, so tighten the motion.
• Too little heat. A cold bead traps slag. Check your voltage.

Cracks

Cracks get different treatment. Do not weld over them or try to diagnose them while you keep producing welds. Stop and tell your lead. Most codes reject cracks of any size.

Centerline cracks form while the metal is still hot. A bead that is deep and narrow is more likely to crack, since its last metal to freeze sits at the center and takes all the shrinkage. A thin, concave root pass does the same thing. Check for a tight joint, a thin root bead, and too much restraint.

Delayed cracks appear hours or even days after welding. If one appears on a job that looked good yesterday, think hydrogen. That points to moisture in wire or gas, oil on the metal, or preheat that was too low, and it needs your lead and the inspector right away.

Arc and Feed Problems That Look Like Weld Problems

Many weld flaws begin as equipment trouble. Here are the most common ones.

A contact tip costs very little, and a worn one causes more confusing problems than almost anything else on the gun. Swap it early.

Worked Example: Tracking Down a Repeat Defect

A shop runs 0.045 inch E71T-1M on frames. On Tuesday, a welder finds porosity at the first half inch of each weld and nowhere else. The rest of each weld is clean.

Pattern: Only at the start of the weld, and on every weld.

Theory: Gas is not reaching the arc when it strikes. Porosity from wind or dirty metal would not stop after half an inch.

Facts: The flowmeter reads 40 CFH, which is fine. There is no draft. The cylinder is half full. The feeder was swapped on Monday.

Test 1: Check the machine's preflow, which is the short burst of gas before the wire starts. It is set to zero on the new feeder. The welder sets it to a small value and runs a test bead on scrap. The start is clean.

Proof: Three test welds in a row with clean starts. The welder notes the setting on the machine, and tells the lead so that the affected welds can be checked.

Notice what the welder did not do. They did not raise the gas flow or change the voltage, and they did not grind the starts and hope. One change was enough, and the pattern pointed straight at it.

Proving the Fix

A weld that looks better is not always a weld that is better. Use a test that shows what is inside.

Fillet break test. Weld a short fillet on one side of a tee joint, then force the upright plate over toward the weld until it breaks. Look at the broken surface for lack of root fusion, porosity and slag. It is fast and cheap, and it is a good check when you are chasing fusion problems. Wear eye protection, and hold the part in a press or a heavy vise. Metal can fly when it snaps.

Bend test. Cut a strap from a groove weld and bend it in a jig around a mandrel. Open defects show on the stretched surface. This is the same kind of test used on qualification coupons.

Macro etch. A section is cut, polished and etched so that the weld profile shows up. It reveals penetration and fusion depth. Etching uses acids, so only trained people should do it, with the right ventilation and protection.

When to Stop and Escalate

Stop welding and tell your lead or inspector when:

• You find any crack.
• The same flaw returns after you have tried the obvious fixes.
• The fix needs settings outside the WPS.
• A flaw may have gotten into welds that were already finished or painted.
• You find the wrong wire or gas has been in use.

Raise it early. Finding that twenty frames share one flaw is bad news, and finding it after shipping is far worse.

Common Mistakes

• Raising the gas to cure porosity. It can pull in air and make things worse.
• Slowing down to fix a thin bead. The puddle can roll ahead and cause lack of fusion.
• Changing several things at once. You will not know what worked.
• Trusting appearance. Prove the fix with a break test or a bend test.
• Ignoring the pattern. Where and when the flaw appears is your best clue.
• Forgetting earlier welds. If you find a cause, the welds made before the fix may be affected too.

Check Your Understanding

Porosity appears along the full length of every weld at one station, starting right after a large exhaust fan was installed nearby. What is your first theory?

The fan is stripping away the shielding gas. Block or redirect the airflow, then test. Raising the flow is the wrong fix.

A fillet looks smooth, but it breaks open with little force and shows a shiny, unfused root. The welder used short-circuit settings on 1/2 inch plate. What happened?

That is lack of fusion. The cool arc did not melt the root. Ask about a process or settings that give more heat on that thickness, and check where the arc was aimed.

You see a straight crack down the center of a root pass, right after welding. What might have caused it, and what do you do first?

A deep, narrow or thin concave bead, or high restraint, can cause it. Stop and tell your lead first. Do not weld over it.

Your flux-cored bead has worm-shaped marks on its surface, but there are no holes. What do you check?

Check the gas flow and the voltage. Too much gas or the wrong voltage is the usual cause of those marks.

Key Takeaways

• Read where and when a flaw appears before you change anything.
• Start with cheap, simple checks: gas, tip, connections, cleanliness and stickout.
• Change one thing at a time, and test on scrap in the same position.
• Lack of fusion hides under a neat bead, so prove your fix by breaking or bending.
• Stop at any crack, and tell your lead right away.
• When you find a cause, think about the welds already made.