Posted in

Why is my welding robot producing porosity?

If you’ve ever stood by a welding robot cell, watching a seamless automated weld only to spot tiny, gaping holes in the bead, you know the frustration. As a welding robot supplier who’s spent the last 12 years troubleshooting these exact issues for manufacturers across automotive, heavy equipment, and agricultural equipment, I’ve lost count of the times a production manager has called me at 2 a.m. panic-stricken because their robotic weld porosity has shut down a full shift. Porosity isn’t just a cosmetic flaw—those tiny voids weaken weld integrity, cause parts to fail structural load tests, and can even lead to costly field recalls. Today, I’m walking you through the most common, science-backed reasons your welding robot is spitting out porous welds, no fancy jargon, just real insights I’ve gathered from fixing thousands of robot cells. Welding Robot

First, let’s start with the most common culprit, and the one I see show up in 40% of all porosity cases I address: improper gas shielding. If you think your robot’s gas line is just carrying inert gas to the weld puddle, think again—this is the invisible barrier that keeps oxygen, nitrogen, and hydrogen (the big three porosity-causing gases) from contaminating the hot weld metal. I once worked with a small parts manufacturer in Detroit that thought their shielding gas setup was fine—until we crawled under their robot cell and found a 2-foot kink in the gas hose that only opened fully when the robot was at a specific welding angle. When the robot moved to the next seam, the kink pinched off, cutting shielding gas by 70%. The result? Porosity that popped up only on angular welds, and no one could figure out why because they only checked the gas flow at the start of the shift.

Here’s the science to back that up: weld pools are molten metal at around 3,000°F. At that temperature, oxygen and nitrogen from air are highly soluble in the molten metal, but when the metal cools and solidifies, their solubility drops sharply—like trying to hold water in a sponge that’s suddenly turned rigid. Those gases get trapped, forming tiny spherical voids: porosity. Hydrogen, meanwhile, comes from moisture—any moisture—since it dissolves easily in molten metal and forms bubbles as it solidifies. If your shielding gas is weak, or contaminated, or the gas nozzle is dirty, you’re not keeping those atmospheric gases out.

Wait, I know what you’re thinking: “I check my gas flow meter every week, it’s set to 25 cfh (cubic feet per hour) just like the manual says.” Yeah, that’s not enough. First, make sure your flow rate matches your nozzle size and the wire feed speed you’re running. For example, a 16mm nozzle on a robot welding thin 16-gauge steel needs around 20 cfh, but a 20mm nozzle welding ½-inch steel needs 30 to 35 cfh. Too low, and the shield is turbulent, letting air sneak in; too high, and the gas creates its own turbulence, pulling in surrounding air (that’s called gas entrainment, a trick I learned the hard way early in my career). Also, shielding gas purity is non-negotiable. I once had a client switch to a cheaper gas supplier to save $50 a month, and within two weeks, their porosity rate jumped from 0.8% to 6.2%. That cheap gas had 2% air contamination, which is enough to cause consistent porosity. And don’t forget the nozzle itself—spatter builds up on the tip, blocking gas flow, so you should be cleaning that every 4 to 6 hours of run time, not at the end of a shift when spatter has caked on.

Next, up next: material and preparation issues, which I blame for another 30% of robot-related porosity cases. This is where I see manufacturers skip the basics because they think robots will fix every flaw. Let’s start with base metal contaminants: oil, grease, rust, paint, mill scale, even condensation on the metal surface. All of these have hydrogen, oxygen, or nitrogen trapped in them, and when your robot welds over them, those contaminants vaporize into the weld pool, leaving bubbles as the metal solidifies. I worked with a farm equipment maker last year that was welding steel chassis parts that had been sitting outside in a rainy warehouse for two weeks—no one wiped them down before loading them into the robot cell. The resulting welds had so much porosity, they failed a pressure test for hydraulic lines. The fix? Just a quick wipe with isopropyl alcohol and a wire brush to remove surface contamination. Simple, but so many people skip it because robots are supposed to be hands-off.

Then there’s wire preparation and condition. Welding wire, especially solid wire, can pick up moisture from storage if it’s not sealed properly. If your wire spool is left open in a damp shop, the core wire will absorb moisture, and when it feeds through the robot’s liner, that moisture vaporizes in the contact tip, adding hydrogen to the weld. Even the liner itself matters: a worn, dirty liner can cause the wire to arc or burn unevenly, spitting small bits of molten metal into the weld pool and creating voids. I once had a client with a 5-year-old liner that was full of spatter buildup—every time the robot fed wire, the liner would catch, causing a 0.1-second arc spike that left tiny porosity on the root pass. Replacing the liner fixed it instantly. Also, check your wire diameter mismatched? If you’re using a 0.035-inch wire but your contact tip is for 0.045-inch, you’ll get inconsistent feeding, spatter, and porosity. That’s a common, easy-to-miss mistake.

Now, the third big category: robot parameters and programming errors, which make up the remaining 30% of cases. This is where I love working with clients because adjusting a robot’s program or tweaking settings can fix porosity faster than anything else. Let’s break it down. First, wire feed speed (WFS) and voltage: these have to be perfectly balanced. If voltage is too high for a given WFS, you get excessive spatter and a turbulent weld pool that traps gases. If voltage is too low, the weld pool is too cold, and the metal solidifies before gases can escape—trapping them and making porosity look like tiny, random holes. I’ve seen robot programs set up for a different material or thickness, so a client running 1/4-inch steel with parameters meant for 1/8-inch steel, leading to constant porosity. That’s a setting error that takes 2 minutes to fix.

Next, travel speed and angle. Robots move consistently, but if the travel speed is too fast, the weld pool doesn’t have time to degas—gases can’t rise to the surface and escape, so they get stuck inside. If travel speed is too slow, you overheat the weld, causing more spatter and turbulent flow that traps voids. I once watched a robot welding a long seam at 15 inches per minute, which was too fast for the ½-inch steel. The porosity was so uniform along the seam, it looked like someone had poked it with a pin. Cranking the travel speed down to 9 inches per minute let the gases escape, and porosity vanished. Also, the work angle and contact tip to work distance (CTWD) matter. CTWD should be ½ inch to ¾ inch for most robot GMAW (MIG) welding. If it’s too long, the gas shield is weak, and if it’s too short, the contact tip spits onto the workpiece, contaminating the weld. Robots can drift in CTWD over time as the wire wears, so you should calibrate CTWD every month, not just when you set up a new program.

Wait, I can’t forget about root pass issues if you’re welding joints that require a root run. A lot of robot cells use a backing gas for TIG or MIG root passes to prevent atmospheric contamination on the back side of the weld. If the backing gas line is blocked, or the gap between the two base metal parts is too wide, air gets into the root and creates porosity from the inside out. I worked with a pipe fabricator that had 10% porosity on their pipe roots until we found that the gap between pipes was 1/8 inch, when the robot was calibrated for 1/16 inch—air was getting sucked into the root gap, causing porosity. Tightening the gap fixed that.

Now, let’s talk about something I don’t hear enough: equipment wear and calibration. Robots are precise, but their parts wear out, and if you’re not doing routine maintenance, porosity is going to pop up. For example, the robot’s positioner, which holds the workpiece, can drift over time—if it’s not calibrated, the weld seam might be slightly off, so the torch is at the wrong angle, or CTWD is inconsistent, leading to variable gas shielding. I once had a client whose robot’s axis calibration was off by 0.5 degrees, which meant that on the far end of a long seam, the torch was angled so that half the gas shield was pointing away from the weld pool. The porosity was only on that far end, and it took a full shift of troubleshooting to realize it was a calibration issue. Also, check for loose connections in the power cables—loose connections cause inconsistent current, leading to variable weld pool temperature, which can trap gases.

Here’s a pro tip I’ve developed over the years for robot-specific porosity testing: do a simple “bead-on-plate” test every week. Take a scrap piece of the same material and thickness you’re welding, set the robot to run a 6-inch straight seam, and then etch or bend that bead to check for porosity. This is way more effective than waiting for a full production part to fail a test. I tell all my clients that this 5-minute test catches 90% of porosity issues before they cause downtime.

Now, let’s address the elephant in the room: sometimes, it’s not a mistake, it’s just the limits of your process. For example, if you’re welding very thin gauge steel (under 16 gauge) with a robot, the weld pool is so small that gases can escape before the metal solidifies, leading to porosity. Or if you’re using a flux-cored wire in a high-humidity environment, even with perfect shielding, the flux can pick up moisture. The fix there might be switching to a solid wire with better gas shielding, or adding a dehumidifier to your shop if humidity is over 60%, which is another common hidden cause of hydrogen-related porosity.

I want to make this real, so let’s recap a recent example from a client: a heavy equipment manufacturer that was getting 5% porosity on their bucket hinge pins. We ran through the steps: first, checked gas flow (it was set to 30 cfh, correct for the 20mm nozzle), then checked material prep (the pins had mill scale, so we tested a bead after wire brushing, porosity dropped to 1%), then checked robot parameters: travel speed was 12 ipm, which was perfect, but CTWD was 1 inch, which was too long—calibrating CTWD to ¾ inch brought porosity down to 0.2%, which is well under their quality limit. That took two hours of troubleshooting, not two days, because we followed this exact list of causes.

So, what should you do next if you’re dealing with welding robot porosity? Start with the basics: check your gas lines for kinks, check your nozzle for spatter, wipe down your material, run that bead-on-plate test. If that doesn’t fix it, it’s likely a parameter or calibration issue, and that’s where working with someone who knows robotic welding, not just manual, makes all the difference.

As a welding robot supplier, our job isn’t just to sell you a cell—it’s to make sure it runs efficiently, with zero defects. We’ve helped hundreds of clients troubleshoot porosity, from the most simple gas line kink to complex parameter mismatches, and we’ve refined this process over 12 years to get to the root of the problem fast.

If you’re tired of porosity shutting down shifts, wasting material, and delaying shipments, we can walk through your specific cell, material, and production needs to diagnose the issue and implement a long-term fix. No generic manuals, no long wait times, just targeted solutions tailored to your robotic welding setup.

CNC Robot References
Lancaster, J.F. (1986). Metallurgy of Welding. Allen & Unwin.
ASME Boiler and Pressure Vessel Code, Section IX: Welding Qualifications (2023).
Davis, J.R. (Ed.). (2003). Welding, Brazing, and Soldering. ASM International.


Robotic Technology (GD) Co., Ltd.
Robotic Technology (GD) Co., Ltd. is one of the most experienced welding robot manufacturers and suppliers in China, also supports custom service and one year warranty. Please feel free to wholesale CE approved welding robot for sale here from our factory. Contact us for quotation.
Address: Room 102, No. 30,Hengfeng 2nd Road,Daguo, Changping Town, Dongguan City,Guangdong,China
E-mail: admin@iroboticplus.com
WebSite: https://www.rb-tic.com/