Skip to content
Home Articles Standards About Contact
Aluminum Field Study

TIG Welding Aluminum Radiators: Setup, Steps & Fixes

NH
Nicholas Harper · Apr 2, 2026 · 45 min read
Table of Contents

A cracked aluminum radiator on a rebuilt engine is one of the most frustrating sights in any shop — the engine is done, the paint is fresh, and a pinhole the size of a grain of sand is dripping coolant onto the floor. Before you scrap an expensive custom core or pay shop rates for a replacement, TIG welding can often save it, but only if you understand why aluminum radiators are so unforgiving.

Can You TIG Weld an Aluminum Radiator? The Short Answer First

The engine is back in the bay, the paint is fresh, and there it is: a wet spot the size of a coin on the shop floor, traced to a hairline crack in an aluminum radiator. Every restorer meets this moment, and the first question is always the same. Can you actually TIG weld an aluminum radiator, or is that a job for the scrapyard?

Yes. Aluminum radiators are routinely TIG welded — across race shops, restoration garages and home workshops alike. But the honest answer has three conditions attached, and ignoring any one of them is how a repairable core becomes landfill. Success depends on material thickness, contamination and prep. The welding itself is the easy part.

Why TIG and Not Something Else

Radiator tube walls can be as thin as 0.020 in (0.5 mm), thinner than a dime. A MIG arc, a stick arc or an open flame will punch straight through that wall before you can react. TIG welding, formally gas tungsten arc welding or GTAW, gives the operator pedal-controlled heat input measured in single-digit amperes, which is exactly what a half-millimeter wall demands. You decide when the metal wets, when the filler dips in and when the arc comes off. That control is why aluminum radiator repair is a TIG job by default rather than by preference. Our guide to aluminum MIG welding explains where the easier MIG process runs out of capability on thin-wall work.

Other processes do exist. Flux-cored wire can work on heavier sections, and our piece on aluminum flux core welding covers what works and what doesn’t. Oxy-fuel brazing with zinc-aluminum rod and metal-filled epoxies are options too, each with real trade-offs. We compare all of them later, because the wrong repair path wastes more radiators than bad welding ever will.

What an Aluminum Radiator Repair Actually Involves

An aluminum radiator repair is not “weld the wet spot.” The real job looks like this:

  • Finding the true leak. A drip at the bottom of the core may originate a foot away at the header plate. Coolant wicks along fins and follows gravity, so the first pinhole you see is often not the first one you should fix.
  • Stripping the assembly. Fans, shrouds, hoses, petcocks and any rubber or nylon fittings come off. Plastic tanks, mounting tabs and cores are removed or shielded from heat.
  • Cleaning to welding standards. Not showroom clean, metallurgically clean. Coolant residue, oxidation and oil trapped in the metal outgas through the puddle and leave pinholes that refuse to close. This is where most DIY repairs fail, and why proper aluminum cleaner for welding prep matters more than torch technique.
  • Welding with restraint. Low amperage, tight arc, small filler rod, short stitched passes and constant attention to heat buildup in the surrounding fins.
  • Pressure testing and flushing. A repair that looks perfect under a light can still weep at 15 psi. Testing is part of the job, not an optional extra.

The most common failure mode deserves emphasis. Contaminated or coolant-soaked aluminum outgasses as the arc melts it, producing a frothy, porous weld full of pinholes. Experienced welders on WeldingWeb’s radiator welding thread describe a practical recovery tactic for a stubborn last pinhole: drill it out with a 1/16 in bit, clean the hole and re-weld it. Drilling converts a hidden gas pocket into open, clean metal you can actually fuse. Aggressive, perhaps, but it beats layering weld over weld and baking the whole core.

What Is Repairable, and What Isn’t

Set expectations honestly before you strike an arc:

Condition Repairable? Notes
Isolated pinhole in a tank or header plate Usually yes Best-case scenario for TIG
Crack at a tank-to-header joint with sound metal Often yes Clean thoroughly, watch heat
Cracked cast aluminum tank (A356) Sometimes Needs preheat and a cast-savvy filler
Salt-rotted, crumbling tube ends No Welding over corrosion just moves the leak
Multiple failures across a thin, fatigued core Poor value Re-core or replace instead

Quick Note: A single pinhole on an otherwise healthy core is a 30-minute TIG job. A core with a dozen weeping joints is a money pit — the labor of chasing every leak will exceed the price of a new radiator.

Who This Guide Is For

Three readers: the DIY enthusiast with a benchtop AC TIG machine and a project car in the garage, the restoration shop that needs repeatable procedure instead of guesswork, and the motorsport fabricator with a broken core between rounds and no time to waste. The physics is identical for all three; only the tolerance for risk and the available equipment differ.

What follows is the full sequence: how radiators are built and why they leak, how TIG compares to MIG, brazing and epoxy, the equipment and filler metal to load, the prep routine that decides success or failure, a step-by-step walkthrough, troubleshooting, pressure testing, safety and a repair-or-replace checklist. If you learn one thing, let it be the welder’s rule for aluminum: the arc is the easy part, and everything you do before striking it is the job.

How Aluminum Radiators Are Built and Why They Leak

Before you strike an arc on a leaky radiator, know exactly what you’re welding. Unlike a solid sheet of aluminum, an assembly is a mix of thin-walled tubing, delicate fins, and thicker castings joined together and sealed against pressure. Each piece responds differently to heat, which is why so many repairs go sideways.

The Anatomy of an Aluminum Radiator

An aluminum radiator is built to move coolant and shed as much heat as possible in the tightest space it can. Manufacturers use very specific alloys to balance strength, weight, and corrosion resistance.

  • Core Tubes: The small, flattened passages that carry coolant through the core, most commonly made from 3xxx series aluminum, specifically 3003, for its formability and corrosion resistance.
  • Fins: Sandwiched between the tubes, these razor-thin fins add surface area to pull heat out of the coolant. They are typically 3xxx series aluminum as well.
  • Header Plate: The transition piece connecting the many core tubes to the end tanks. It is a heavier sheet that carries the load of every tube attached to it.
  • End Tanks: These hold coolant before and after it passes through the core. Tanks may be brazed, welded, or plastic crimped over a stamped aluminum flange. Tank material is often wrought sheet such as 5052 or 6061, while many cast end tanks use A356-type alloys. Cast tanks behave drastically differently under the arc than their wrought counterparts.

In short, aluminum radiator cores commonly use 3xxx series aluminum for tubes and fins with 6xxx series or cast aluminum for tanks. That jump in thickness and alloy means heat control is already a battle before contamination enters the picture. Filler choice matters even more on castings: 4043 is a 5% silicon alloy with a solidus of 1065°F and a liquidus of 1170°F, well suited to cast and heat-treatable alloys. As Hobart’s aluminum welding guide notes, 4043/4943 and 5356 account for over 85% of all aluminum weldments, with 4043 a go-to for the porous, unpredictable nature of cast tanks.

Why Aluminum Radiators Leak

These assemblies live in one of the harshest environments on a vehicle. Between heat cycles, pressure, and road abuse, a weak point eventually gives out. Common culprits include:

  • Vibration Fatigue: Constant engine vibration causes micro-movement between the tubes and the header plate. Over decades of cycles, hairline cracks form at those stress points.
  • Corrosion: Old or neglected coolant turns acidic and eats thin 3003 tubes from the inside out, creating pinholes with no outside impact at all.
  • Freeze Damage: Coolant that is not properly mixed or maintained can freeze, and ice expansion bulges and splits tubes, often stretching them past the point of repair.
  • Road Debris: Rocks and other projectiles puncture thin core tubes easily, especially on older or lower-mounted radiators.
  • Old Solder/Brazed Joints: Many older units were never welded. Decades of heat cycles make these brittle joints creep and weep, turning a tight seal into a constant drip.
  • Coolant Contamination: Mixing incompatible coolants or using tap water accelerates galvanic corrosion, rotting multiple tubes at once.

Quick Note: Internal corrosion is the sneakiest failure of them all. A tube can look pristine on the face of the core yet be paper thin inside, making it extremely prone to burn-through once you introduce arc heat.

The Three Leak Types That Matter For Welding

Not every leak is created equal. Where it comes from decides whether a TIG repair is worth attempting.

  • Leak at a Tube: Usually the smallest pinholes, found in the flattened 3003 core tubes. They are the most common and often the most realistic weldable failures on an otherwise solid core.
  • Leak at the Header-to-Tank Joint: Where the tube bundle meets the header plate, or the header meets the end tank. These joints see immense stress and are notorious for long, creeping cracks that run underneath fittings, making a continuous repair tricky.
  • Cracked Cast Tank: On radiators with cast A356 end tanks, these are almost always stress cracks originating at mounting ears, necks, or sharp corners. Casting porosity is rampant, and even after pristine prep, trapped gas can erupt under the arc. This failure demands the most patience of the three.

Knowing these differences saves hours of guesswork. Whether you are working around a delicate core or fusing a finicky casting, prep is non-negotiable. For the cleaning side of the job, follow a proper Aluminum Cleaner for Welding: How to Prep Aluminum Right before you lay a single tack. These fundamentals determine whether your repair ends with a solid seal or a brand new pinhole to chase.

TIG vs MIG vs Brazing vs Epoxy: Repair Options Compared

Walk into five different shops with the same leaking aluminum radiator and you’ll get five different answers: weld it, braze it, epoxy it, or replace it. All four approaches have closed a leak at some point, but they are not interchangeable. The right choice depends on where the leak is, how thin the surrounding metal is, what equipment you own, and how long the repair needs to last. Here’s how the realistic options stack up before you commit to one.

TIG Welding: The Professional Standard

Gas tungsten arc welding is the benchmark for aluminum heat-exchanger work, which is why professional radiator shops such as Affton Radiator Services hold TIG and MIG certifications for aluminum repair. TIG gives you precise heat control through the foot pedal, a clean non-consumable tungsten, and the ability to fuse a 0.020 in tube to a header plate without blowing through. The trade-off is skill: aluminum has no color change before it melts, and the oxide layer sitting on top melts at roughly 3,700°F while the base metal melts near 1,220°F. If your torch lingers a half-second too long, the puddle collapses.

For radiator work specifically, TIG is the default recommendation whenever the crack or pinhole sits on a visible, accessible seam: tank-to-header joints, end tank corners, and mounting tabs. The Hobart Brothers aluminum welding guide covers the fundamentals of AC balance, cleaning action and penetration that govern this kind of thin-section work.

MIG Welding: Faster, Harder on Thin Material

Metal inert gas welding on aluminum is entirely possible and considerably faster, but it is less forgiving on thin cores. A MIG arc is hotter and less concentrated than TIG, so the window between “good fusion” and “blown hole” is narrow on 0.020 in tubing. It works better on heavier cast tanks and thick mounting bosses where the heat sink helps you. If MIG is what you own, our Aluminum MIG Welding: The Complete Practical Guide walks through wire feed, push technique and gun maintenance specific to aluminum. And if you’ve only got a flux-core machine, read Aluminum Flux Core Welding: What Actually Works in 2026 first, because standard flux-core wire will not wet aluminum at all.

Oxy-Fuel Brazing with Zinc-Aluminum Rod

This is the lowest-skill entry point. Zinc-aluminum brazing rods melt around 700°F, far below aluminum’s roughly 1,220°F melting point, so the base metal never actually liquefies. You’re wetting a surface rather than fusing a joint, which means you can seal a leak with a torch and a inexpensive rod. The downside is mechanical: a brazed joint is a surface bond, not a metallurgical fusion, and it fails sooner under vibration and thermal cycling. It also leaves a zinc-rich deposit that is ugly, harder to inspect, and dissimilar to the base alloy. It’s a reasonable roadside or tractor-radiator fix; it’s a weak choice for a restored car going back into service. Related processes are covered in Aluminum Flux Welding: How It Works and When to Use It.

Two-Part Metal-Filled Epoxy

Epoxy is the tempting option because it needs no heat at all, and quality products do seal aluminum well. The catch is temperature. Many two-part epoxies are rated near 120°F in water service, which sounds adequate until you remember a modern engine coolant routinely runs well above that. A repair rated to 120°F can soften, creep and fail quickly on a hot running cooling system unless you specifically choose a high-temperature formulation and even then you’re buying time, not permanence. Treat epoxy as a diagnostic tool: seal the area, pressure test, confirm the leak location, then weld it properly.

Side-by-Side Comparison

Method Heat Input Skill Relative Cost Permanence Best Use Case
TIG welding High, but precisely controlled High Moderate (machine + gas) Permanent, full fusion Tank/header seams, cracks, cast tanks
MIG welding High, harder to focus Moderate to high Moderate Strong if no burn-through Thicker tanks, bosses, brackets
Zinc-aluminum brazing Low, torch-level Low Low Temporary to serviceable Emergency fixes, older utility radiators
Metal-filled epoxy None Low Low Short-lived on hot systems Pinhole sealing for testing only
Full replacement / re-core N/A None High New-unit lifespan Widespread corrosion or multiple failures

Quick Note: Experienced welders advise practicing on aluminum cans before attempting 0.020 in radiator tubing. A soda can wall is roughly the same thickness as a core tube, and a stack of crushed cans will teach you pedal control and arc length faster than any YouTube video. If you can’t weld a clean bead on a can, you will open a hole in the header plate.

When Repair Becomes a False Economy

Repair only makes sense when the surrounding material is healthy. If the tubes are salt-pitted, the header plate is paper-thin from corrosion, or the tank is distorted and rocking against the core, welding is just an expensive delay. You’ll close one pinhole and open another on the next heat cycle. In those cases, re-coring or buying a quality replacement unit is the cheaper path over the life of the vehicle. The same logic applies to cracks that keep travelling: a fatigue crack is telling you the joint is being work-hardened by vibration, and welding it without addressing the cause invites a repeat failure. Discussion of the same problem appears in community threads such as this radiator welding thread on WeldingWeb.

One more variable: prep quality beats process choice every time. A flawless TIG weld on a greasy, oxide-contaminated seam will fail, while a careful braze on a surgically clean surface can outlast it. Whatever route you pick, the cleaning discipline in the next section is what decides the outcome, and Aluminum Cleaner for Welding: How to Prep Aluminum Right goes deeper on the chemistry of why.

Choosing Your TIG Machine, Tungsten, Gas and Filler Metal

Aluminum radiator work punishes wrong equipment choices faster than almost any other job in the shop. You are generally joining 0.9 to 1.2 mm (about 0.035 to 0.047 inch) tube wall to a header plate that may be twice that thickness, on a part you cannot easily reposition and cannot overheat without collapsing a tube. So the gear conversation is not about buying the most powerful machine. It is about control.

The Machine: AC/DC TIG With Real Adjustability

You need a TIG machine that does both AC and DC, with high-frequency start and, critically, adjustable AC balance and AC frequency. DC polarity handles steel and stainless, but aluminum requires alternating current to strip the oxide layer while depositing metal, and cheap AC machines with a single fixed setting will fight you on thin radiator work.

AC frequency controls arc width. Raising frequency narrows and focuses the arc, which is exactly what you want when you are chasing a pinhole in thin tube without heating the fins next to it. Lowering frequency produces a wider, softer puddle that spreads heat better across a heavy cast end tank. Most clean aluminum projects respond well to roughly 70 to 80 percent electrode-negative (EN) balance, and a practical baseline is 75 percent EN with the frequency set between 100 and 120 Hz. AC balance shifts how much of the cycle cleans versus how much it penetrates; more EN means deeper penetration and less cleaning, while more AC-positive time widens the cleaning zone but adds heat. Eastwood’s explainer on AC balance and frequency control walks through those trade-offs in more depth.

Amperage range matters too. A machine with smooth, low-end output at 20 to 40 amps and a fine-pitch foot pedal will serve you better than a 250 amp unit with coarse adjustment. If you are still weighing whether TIG is worth learning for this kind of work versus aluminum MIG welding, the pedal is the argument: it lets you modulate heat in real time around corners and thin sections.

Tungsten, Gas and Cup Selection

For general radiator work, a 2% lanthanated (blue band) tungsten in 3/32 inch (2.4 mm) is the modern default. It starts reliably on AC, resists balling better than pure tungsten, and tolerates inadvertent touch-ups better than the old pure or zirconiated (white/green band) electrodes, which were the traditional AC choices and still work but demand cleaner technique. Ceriated tungsten is a reasonable second choice. Grind the tip longitudinally, pointed, and keep a dedicated set for aluminum only.

Shielding gas is 100 percent argon, no exceptions, no helium mixes for this thickness. Flow rates of 15 to 20 CFH through a number 6 or number 8 gas cup cover most radiator seams; go to a number 8 when the joint geometry is awkward or you are welding near fins that disrupt gas coverage. An oversized cup on thin tubing just blows cooling air around the puddle and invites oxidation, so match the cup to the space you actually have. Your local supplier, whether an Airgas counter or an independent distributor, can set you up with a proper cylinder and flowmeter regulator rather than a cheap gauge-only setup.

Quick Note: Purge the inside of the tank where you can. Feeding argon through a hose and a small hole into the tank interior dramatically reduces sugary, oxidized backside welds that weep under pressure even when the visible bead looks perfect.

Filler Metal: 4043, 5356 or 4943?

This is where radiator work diverges from general aluminum fabrication, and where most beginners pick wrong.

Filler Typical as-welded tensile strength Character Best radiator use
ER4043 Around 27,000–29,000 psi Fluid, low smut, very crack resistant Tube-to-header seams, 6061 and cast A356 tanks
ER5356 Averages about 28,000 psi Stiffer, brighter, resists salt-water corrosion Marine and 5052 end tanks
ER4943 Comparable to 4043 Higher-strength 4043-type with better heat response Thin wall work where 4043 flows too freely

ER4043 is the default recommendation for a typical aluminum radiator. Its silicon content makes the puddle wet and forgiving, it resists solidification cracking in 6000-series base metal and in cast alloys, and it produces far less black smut than 5356, which matters when the finished part gets inspected and painted. Washington Alloy’s 4043 data sheet confirms the strength range and the crack-resistance rationale, and Hobart Brothers’ aluminum welding guide covers the selection logic in a shop context.

Reach for ER5356 when the radiator lives in a salt-water environment (boat radiators and some intercooler applications) or when the end tank is 5052, because matching the magnesium content preserves corrosion resistance. The trade-off is a tighter, less fluid puddle and a higher sensitivity to porosity if your cleaning was sloppy. In/Weld’s 5356 data sheet documents the approximately 28,000 psi average and its salt-water resistance. ER4943 has become a favorite for thin-wall work because it behaves like 4043 but holds strength slightly better. Use 1/16 inch (1.6 mm) rod for tube wall, stepping up to 3/32 inch for cast tank repairs. Cast aluminum has its own quirks, which we covered in more detail in our alloy wheel welding guide.

Starting Settings Cheat Sheet

Joint Approx. thickness Amps Frequency EN balance Filler
Tube to header 1.0–1.2 mm 35–55 A on pedal 120 Hz 75% 1/16″ ER4043
Header to cast tank 2.0–3.0 mm 60–90 A 100–110 Hz 70–75% 3/32″ ER4043
Pinhole in tube ~1.0 mm 25–40 A 130–150 Hz 80% 1/16″ ER4043

Treat these as dial-in points, not gospel. Back them up with proper joint preparation, which is where radiator repairs are actually won or lost, and skip flux-core aluminum wire entirely for this application. Real-world radiator threads on forums like WeldingWeb are full of the same conclusion: right machine, right filler, obsessive cleanliness.

Preparation: The Step That Decides Whether the Weld Succeeds

When it comes to tig welding aluminum radiator repairs, the outcome is rarely decided by the arc itself. It’s decided at the bench. Most of your time will be spent draining, stripping, cleaning, drying, and fitting the core. Aluminum is far less forgiving of shortcuts than steel, and the most frustrating leaks are the ones that appear after a weld that looked perfect.

Why “Clean” Means Something Different

A part can look spotless to the eye, but still contain contaminants that ruin a TIG weld. For radiators, three in particular create the majority of issues:

  • Trapped fluids. Coolant, glycol, and oil can wick deep into end tanks, header seams, and the base of tubes. When heated by the arc, these fluids vaporize instantly. With nowhere to escape, that gas travels through the puddle and solidifies as porosity. This outgassing is the reason many radiator pinholes refuse to close, no matter how much filler is added.
  • Oxide layers. Aluminum oxide melts at roughly 3,700°F (2,040°C), while the base aluminum melts near 1,220°F (660°C). Instead of burning off, the oxide gets stirred into the puddle as a stringy inclusion. Even after removal, a fresh oxide layer begins to form within seconds, so the final scuff should always be done immediately before welding.
  • Coatings and residues. Old paint, gasket material, underseal, and solvent films don’t just hurt the weld, but can release toxic fumes when heated. The OSHA welding fact sheet explicitly states that all welding surfaces must be free of these coatings and solvent residues prior to starting work.

Quick Note: If your cleaned area looks dull gray instead of a uniform, bright scratch pattern, it’s not ready. Shiny spots surrounded by dull areas are often fingerprints, and those spots will produce porosity every time.

The Prep Sequence, In Order

To avoid chasing leaks, follow these steps in order rather than cutting corners:

  1. Drain completely. Remove the lower hose and allow the core to drain fully. Invert it as well, as many radiators hold residual coolant in the tank corners.
  2. Flush. Run clean water through the inlet and outlet until it runs clear. If the engine was recently rebuilt, flush until all casting sand and machining residue is removed.
  3. Strip the hardware. Remove fans, shrouds, hoses, rubber isolators, plastic clips, sending units, and any wiring attached to the tanks. These materials can melt, smolder, or release irritating smoke when exposed to the arc.
  4. Dry and vent. Blow the core out with clean, dry air and let it sit. Any trapped moisture will instantly turn to steam during welding, causing porosity from within the assembly.
  5. Mechanical cleaning. Scrub the repair area with a dedicated stainless steel wire brush (never carbon steel) or a clean Scotch-Brite pad, working in one direction until the metal is uniformly bright. For a more in-depth breakdown, check out Aluminum Cleaner for Welding: How to Prep Aluminum Right.
  6. Solvent wipe. Wipe the joint and at least a two-inch margin with acetone on a lint-free rag. Change the rag frequently as it loads up, and allow all solvent to fully evaporate before striking an arc.
  7. Fit-up and tacks. Close any gaps as tight as possible, clamp securely, and tack both ends and the middle. For thin tube-to-header joints, keep your tacks small to prevent pulling the alignment out of square.

This level of discipline is standard for critical aluminum work, and is mirrored in practices outlined within Aerospace TIG Welding: Standards, Techniques, and Training.

Final Protections Before Welding

With the metal clean, a few preventative measures will safeguard the rest of the core:

  • Vent the core. Always leave the radiator cap off while welding. This prevents pressure from building inside the tanks and pushing gas through your solidifying puddle.
  • Protect the fins. The fins surrounding your repair are extremely thin. Clamp a copper or aluminum backing bar behind the seam, and drape a damp cloth over nearby fins. Be sure to keep any moisture far away from the puddle itself to avoid introducing hydrogen.
  • Aim your heat wisely. Direct the arc towards the thicker header or end tank, allowing the heat to soak into it. As the thick section reaches temperature, bring the puddle onto the thin tubing. This helps prevent burn-through without sacrificing fusion.
  • Consider preheat for cast sections. Cast A356 end tanks are porous and can hold residual moisture. A light preheat of 150–250°F (65–120°C), verified with a contact or infrared thermometer, can help drive off moisture and slow cooling to reduce cracking. As always, reference a recognized aluminum welding guide for your specific alloy. Never assume a warm tank is a dry one, as trapped moisture will blow out as porosity under heat.

None of these steps require specialty equipment, just patience. By locking in your preparation, you’ll find that the rest of your weld falls into place, and your repair will hold pressure for years to come.

Step-by-Step: TIG Welding the Radiator Seam

Everything up to this point has been preparation. Now we close the leak. Radiator TIG work isn’t dramatically different from other thin-aluminum TIG work, but the margin for error is smaller: most radiator walls run 0.040 to 0.063 in thick, and one moment of inattention burns a hole you’ll spend the next hour chasing. Work slowly, keep the pedal under control, and let the puddle tell you what it needs.

Positioning and Clamping Before the Arc

Aluminum pulls hard as it cools, and a radiator is a box of thin connected panels that warps easily. Support the tank or seam fully and let both members expand freely. Prop the radiator on a fixture or V-blocks rather than clamping the core, which will crush tubes. Orient the joint so you can see the leading edge of the puddle: a tank-to-header joint welds best slightly uphill, letting the puddle sit back while you add filler.

Tack before you weld. Place tacks every 1 to 1.5 in, alternating ends so heat doesn’t pile up in one area; on a long header joint, work from the center outward in a skip pattern. Tacks also confirm fit-up. If the joint opens or shifts as you tack, stop and fix it now, because full welding heat will only make it worse.

Torch Settings and Starting Amperage

Published GTAW tables give a useful anchor: for 1/16 in aluminum with 1/16 in filler and about 20 CFH of argon, most references recommend roughly 60–80 A, rising to about 125–175 A at 1/8 in, according to the Hobart Brothers aluminum welding guide. Radiator walls sit below that 1/16 in threshold, so you’ll likely start well under 60 A, sometimes as low as a couple dozen amperes on 0.040 in tube with a 0.040 in electrode. The pedal, not the dial, is your real current control: set the machine a little high and modulate down.

Material (approx.) Electrode Filler Starting current
0.040 in tube wall 0.040–1/16 in, 2% lanthanated 0.045 in 4043 25–40 A, pedal modulated
1/16 in sheet (tank) 1/16 in, 2% lanthanated 1/16 in 4043 45–60 A
1/8 in tank or cast A356 3/32 in, 2% lanthanated 3/32 in 4043 or 4943 70–110 A, plus preheat

Two AC settings change the arc’s character more than raw amperage. Frequency (100–200 Hz) tightens the arc cone for precision; as Eastwood explains on AC balance and frequency, higher frequency concentrates the arc without necessarily increasing penetration. Balance matters even more on radiators: shifting toward more cleaning (DCEP) increases the cathodic action that strips oxide, but it also cuts net heat input into the part, giving you cleaner oxide removal while cooking the thin wall less.

Arc Length, Travel Speed and Filler Technique

Keep the tungsten as close to the work as you can without touching it, roughly one electrode diameter. A long arc on AC is a wide arc, and a wide arc on 0.040 in wall is a blow-through. Point the torch so 10–20 degrees of the arc’s force pushes toward the thicker member: on a tank-to-header joint, aim heat at the tank or cast section and let the puddle wet into the thinner tube. That one habit fixes a huge share of fusion problems.

Dab filler into the leading edge of the puddle, not into the arc: dip, withdraw, advance, repeat in a steady rhythm. If the puddle turns shiny and fluid without rod, ease off the pedal before it runs away. If the bead piles up and won’t wet out, you’re moving too fast or the joint isn’t hot enough — add pedal travel, not more rod.

Tacking Sequence, Back-Step Welding and Corners

Never run one continuous bead the length of a radiator seam. Use back-step welding: short 1 to 1.5 in segments, each running opposite the direction of travel, toward cool already-welded metal. This keeps heat concentrated in a small zone at any moment and dramatically reduces the warping that pulls tanks out of shape. Between segments, brush away the fresh oxide film — aluminum re-oxidizes almost instantly, and proper aluminum cleaning for welding applies between passes, not just before the first arc strike.

Corners and tank-to-header transitions are where radiators die. Corners concentrate heat, so shorten segments to 0.5 in, ease off the pedal, and reposition so you’re never welding a corner blind. At the header joint, put roughly 60–70% of the heat into the header plate and less into the thin tube edge; that melts the tube into the plate without vaporizing it. If a spot refuses to fuse, let it cool completely, then attack it fresh instead of pouring amps into cold metal.

Finishing the Weld

As you complete each segment, ease the pedal down over two or three seconds while adding a final dab of filler — crater cracks in aluminum almost always start in an abruptly extinguished arc. Keep postflow running over the finished bead; shielding hot aluminum from oxygen as it cools avoids the gray sugar coating that signals fresh oxide. Leave the bead slightly convex rather than washed flat — reinforcement costs nothing and buys insurance on a pressure-tested seam.

Quick Note: If TIG still fights you on this job, the practical fallback for thin, uncontaminated tank sections is a spool-gun MIG setup, as covered in our aluminum MIG welding guide. It’s less precise but far more tolerant of marginal prep.

One last discipline tip: if a segment goes badly, grind it out and redo it while the repair is still local. Aluminum repairs breed — porosity in one spot means contamination nearby, and aerospace welding practice treats every suspect area as suspect until proven otherwise. Fixing it now takes ten minutes; fixing it after the radiator is painted, mounted, and pressure-tested takes an afternoon.

Common Problems and How to Fix Them

Aluminum radiator cores are thin and heat sensitive, and it’s rarely the first pass that causes the most grief. More often it’s one of the small, sneaky defects that shows up after hours of careful work. The key to saving your repair is spotting those defects before you pressurize the core.

Porosity and Blowholes

Cause: The number one offender for these pin-sized voids is contamination. Trapped coolant residue left in the tubes is especially deceptive, as post-weld leaks frequently trace back to subsurface porosity from trapped coolant residues rather than the visible bead. Moisture, oils, and failure to break up the aluminum oxide layer can also trap gas in the puddle.

Fix: Flush the radiator thoroughly before welding and confirm every passage is clear of coolant. During prep, pay close attention to your AC cycle. Black contamination or a dull, dirty puddle indicates insufficient cleaning action on the AC cycle. If that happens, shift your AC balance toward more cleaning, and if the surface is heavily oxidized, re-prep it properly using a method from a guide such as Aluminum Cleaner for Welding: How to Prep Aluminum Right before striking another arc.

Burn-Through

Cause: Radiator tubes are paper thin compared to a header or tank. Holding a stationary arc on these passages for too long will almost certainly open them up, especially on older, fatigued cores.

Fix: Use your foot pedal to modulate amperage heavily, keeping puddles small and travel speed up. Add filler only when the leading edge is a true puddle, and never “soak” a pinhole with the arc. On ultra-thin tube repairs, bridging with minimal heat is far safer than holding still to fill a crater.

Lack of Fusion

Cause: Oxidized material left underneath a pinhole is the greatest culprit. Running a bead over an old, corroded leak will rarely bond to the parent metal, because the oxide layer insulates itself from the puddle.

Fix: Drilling a contaminated pinhole out to 1/16 in before re-welding removes oxidized material that otherwise prevents fusion. Then use a fresh, dedicated stainless brush to gently re-clean the exposed edge, wipe it down with acetone, and tack only after verifying a shiny, uncontaminated surface. This one step alone fixes the vast majority of “it looks welded but still won’t seal” repairs.

Cracking at the Crater

Cause: When the arc is abruptly shut off, the molten puddle cools at an incredible rate. In the high-heat-input zones of a radiator, that sudden contraction is the perfect recipe for a hairline crater crack.

Fix: Crater cracks are prevented by tapering current down and adding filler at arc shutdown. Slowly ease the pedal down toward idle while feeding a small dab of filler into the tail of the puddle so it solidifies gradually. Never simply lift the torch off a freshly formed keyhole; that is the most common way to create a crack that will weep days later.

Sugar, Oxide Contamination & Tungsten Contamination

Cause: “Sugar” is the grey, crystalline, frosted oxidation that forms when a hot weld is exposed to atmosphere, most often from insufficient shielding gas coverage or drafts. A discolored, balled, or dipped tungsten electrode introduces microscopic inclusions into the puddle.

Fix: Use a cup large enough for the joint, keep gas flow steady, and work in a draft-free area. If sugar appears, do not grind it off and seal over it; it is oxidation beneath the surface, so remove it entirely down to clean metal. For a contaminated tungsten, shut down, re-grind to a clean point appropriate for AC, and re-establish the arc. That level of cleanliness matters just as much on radiators as it does on more cosmetic repairs, such as those covered in Alloy Wheel Welding: A Comprehensive Guide to Safe and Effective Aluminum Wheel Repair.

The “Last Pinhole Won’t Close” Syndrome & Weeps With No Visible Defect

Cause: This is the most infuriating failure. The visible bead looks perfectly sound, yet micro-porosity runs through a network of fatigued tubes. Each time you seal one pinhole, pressure finds the next weakest interconnected passage.

Fix: Stop chasing pinholes individually. Isolate the area by testing in segments and inspect adjacent tubes under bright light for hairline splits. If you find multiple weeps clustered together, or the core feels brittle under the slightest manipulation, the repair has reached its limit. No number of additional weld passes will restore long-term reliability to a fatigued core.

Short Diagnostic Flow

  1. Observe the puddle: If it’s dull and black, fix your AC cleaning action before anything else.
  2. Inspect the start: Find the leak’s origin. If it’s an old pinhole, drill to 1/16 in and re-expose clean metal.
  3. Monitor heat: If the tube glows cherry red or warps, drop amperage and increase travel speed.
  4. Mind the end: Always taper and fill your crater to prevent a delayed crack.
  5. Test smartly: If a seemingly perfect bead weeps under pressure, assume subsurface porosity and inspect for interconnected splits before attempting another pass.

Quick Note: Troubleshooting is only ever as effective as your prep. Before blaming your machine settings for any of the issues above, verify your cleaning and flushing first. Nine times out of ten, the defect is contamination, not skill.

Testing, Finishing and Protecting the Repair

A repaired radiator isn’t finished when the arc stops. It’s finished when it holds pressure, runs clean coolant, and won’t stain white powder down the side of the tank a month later. Skipping the verification step is how a “fixed” radiator ends up failing 50 miles from the shop, which is exactly the reputation you don’t want attached to your work.

Pressure and Leak Testing the Repair

The standard shop method is still the simplest and most reliable: air-under-water testing. Cap or plug the inlet and outlet, introduce low-pressure air (roughly system operating pressure, around 15 psi, is plenty), then submerge the radiator in a tank of clean water and watch. A pinhole gives itself away as a steady string of bubbles rising from the weld. Mark the spot with a paint pen, drain and dry the core, dress the weld, and repeat. It’s slow, it’s unglamorous, and it finds leaks no visual inspection ever will.

Two alternatives exist for cores that won’t fit a tank. The pneumatic cap test seals the openings and sprays the welds with soapy water, looking for growing foam. A dedicated shop pressure rig with a gauge and regulator does the same job on a bench. Either approach is easier with proper fittings, caps and gauges, which is where a supplier like Airgas Welding Supplies is worth a call rather than improvising with hardware-store plugs.

Safety Caution: Never pressure test with compressed air on a hot radiator, and treat any air pressure test with respect. Air is compressible, so a failed plug or cap releases stored energy violently, and trapped air can turn a small split into a shattering one. Many shops fill the core with water instead of air (hydrostatic testing) precisely because water doesn’t store energy. Let the radiator cool fully, wear eye protection, and stand to the side, not in front of the caps.

Flushing and Cleaning After the Weld

Once the weld passes, flush the core thoroughly with clean water to remove coolant residue, grinding dust and any solvent left inside the tanks. A hot-water flush helps dissolve dried coolant salts that would otherwise circulate into the fresh engine. Then deal with the outside: aluminum welds leave smut, a dark or chalky oxide film that is not just ugly but corrosive under paint.

Cleaning up properly matters more than most people expect, and Aluminum Cleaner for Welding: How to Prep Aluminum Right explains the chemistry behind it. Two practical notes here:

  • Filler choice affects the finish. 4043 welds produce less smut and better anodized color consistency than higher-magnesium fillers like 5356, especially on cast A356 tanks.
  • If the radiator was previously patched with epoxy, solder or a flux-based repair, that material must be mechanically removed with a burr or wire wheel before any TIG work begins. Welding over an old repair traps contamination and guarantees porosity.

Scrub the weld zone with a stainless brush or Scotch-Brite, wipe with acetone, and inspect for the white oxide bloom that indicates incomplete shielding.

Corrosion Protection and Touch-Up

Bare aluminum protects itself with a thin oxide layer, but the heat-affected zone and any remaining smut will chalk and pit under road salt. Clean the area, apply an epoxy or weld-through primer, then a color-matched topcoat rated for higher underhood temperatures. If the radiator is bolted to steel brackets, use isolating washers or dielectric grease to prevent galvanic corrosion between the two metals. Expect some cosmetic color difference in the weld area; a Alloy Wheel Welding: A Comprehensive Guide to Safe and Effective Aluminum Wheel Repair covers why blended aluminum repairs rarely match anodized finishes perfectly.

When to Re-Core Instead of Reusing

Some radiators shouldn’t be saved. Re-core or replace when:

  • pinholes keep appearing along the same header run,
  • tubes feel chalky, thin or spongy under a pick,
  • the tank is distorted, salt-rotted or previously welded more than twice,
  • the brazed core joints themselves are weeping.

A repair that costs half a new core but fails in one season is a false economy. On a rare or custom core with solid material around the joint, TIG repair remains the better investment.

Quick Note: Track your exposure while grinding and welding the repair. Fume limits are written as total particulate, and NIOSH recommends a 5 mg/m³ 8-hour TWA for aluminum welding fumes, so keep ventilation running during post-weld cleaning too.

Everything else in this article — the settings, the prep, the back-step technique — is worthless if you skip this section. TIG welding an aluminum radiator puts three hazards in front of you at once: toxic fume, blinding radiation, and an invisible gas that quietly pushes oxygen out of the air you breathe. None are exotic. All are manageable with ordinary shop discipline, which is exactly why they get ignored until something goes wrong.

What Is Actually in Aluminum Welding Fume

The good news first: aluminum TIG welding produces comparatively little fume. There is no flux coating burning off (unlike stick or flux-cored aluminum processes), and the arc melts clean base metal plus clean filler rod. What does come off the puddle is mostly aluminum oxide — extremely fine particulate that lodges deep in the lungs — plus whatever was on the metal before you struck the arc.

“Whatever was on the metal” is the part that should worry you. Old radiators carry years of coolant residue, silicate scale, oil film and previous brazing flux. Heat any of those and you aerosolize them. OSHA lists aluminum among the metals found in welding fume and states plainly that prolonged exposure may cause lung damage and cancer. The agency treats welding fume as a hazard controlled by engineering controls, not by hoping you won’t inhale enough.

The practical translation: scrub the joint the way covered in aluminum cleaning prep, because a degreased, brushed, solvent-wiped seam produces a fraction of the fume of a dirty one. Most of the “welding smell” you notice on a radiator job is contaminated metal, not the arc itself.

Ultraviolet Radiation and Ozone

A TIG arc on aluminum is an intense source of ultraviolet light — enough that an unprotected arc viewer in the same room can get a corneal burn, the “welder’s flash” that feels like sand in your eyes the next morning. NIOSH notes that aluminum welding arcs generate relatively little fume but produce intense radiation capable of generating ozone. Ozone is that sharp, clean-smelling metallic odor you notice after a few minutes of AC TIG work. It irritates the airways, aggravates asthma, and in poorly ventilated shops it builds up fast enough to cause headaches.

AC TIG on aluminum is harder on your eyes and lungs than DC steel work at the same amperage, so treat every crack in your curtain, every gap under your helmet shade, and every bystander walking past the bay as a real exposure. Shade 10 to 13 is typical for aluminum TIG at radiator amperages; auto-darkening lenses should be checked for battery health before you start a long seam.

Argon: The Oxygen Displacer

This is the hazard most hobbyists never think about. Argon is heavier than air, odorless, and chemically inert — it does not warn you. Spill enough of it and it pools in low areas: the floor of a service pit, under a lift, in a cramped corner, or simply the bottom of a basement shop. The OSHA welding guidance for shipbuilding and repair is explicit that argon and other inert gases displace oxygen and can cause suffocation, particularly in enclosed or partially enclosed spaces.

A radiator repair is low-amperage but slow, and a slow leak from a pinhole-sized defect can displace a lot of gas over an hour. Never work under a vehicle in a pit or on a lift with a flowing TIG torch and poor floor-level ventilation. If you feel lightheaded, short of breath, or oddly euphoric, get into fresh air immediately — that is oxygen deprivation, not fatigue. For cylinder handling, flow rates, and regulator basics, gas supplier safety guides are worth a read before your first tank swap.

Ventilation: The Numbers OSHA Actually Uses

You do not need a certification to judge your shop — you need two measurements. OSHA requires mechanical ventilation where the welding space is less than 10,000 cubic feet per welder, or where overhead height is under 16 feet. That is the trigger point. A two-car garage is roughly 3,000 to 5,000 cubic feet with an 8- to 9-foot ceiling: it always requires ventilation for any meaningful welding, no exception for “I’m only doing a short repair.”

For a radiator on the bench, a capture hood or fume extractor arm positioned within a few inches of the arc does most of the work. A fan blowing across the room is not ventilation; it just redistributes the fume and can disrupt your argon shield, causing porosity you’ll then waste an hour chasing. If you’re considering process alternatives, flux-core aluminum work and other processes carry different fume profiles worth understanding before you choose.

PPE Checklist for Radiator Work

  • Helmet: auto-darkening or fixed shade 10–13, with a properly fitted headgear seal.
  • Skin: jacket or shirt with no open cuffs or collar — UV burns the back of your neck and wrists just as fast as your arms.
  • Hands: lightweight TIG gloves, not heavy stick gloves, which kill dexterity on a 1/16-inch rod.
  • Respiratory: an N95 minimum for general shop work; a half-face P100 respirator for enclosed spaces or radiators with heavy scale and old flux.
  • Feet and legs: closed shoes, no shorts, no rolled-up pant cuffs that channel sparks.
  • Fire: coolant-soaked rags and nearby hoses moved before the arc starts, plus a charged extinguisher within arm’s reach.

Note: Compressed air leak testing belongs in the outdoor or open-shop category, never in a closed bay with the arc lit. Purge the radiator of coolant, vent it fully, and keep your face out of the direct line of any plug or cap that could let go under pressure. The same discipline applies to the gas side: check your regulator, hoses and fittings for leaks with soapy water, never a lighter.

Build the Habit Once

None of this takes long. Ventilation on, extractor arm in position, respirator on, curtains closed, fire watch set, gas lines checked. That sequence becomes automatic after a handful of jobs and costs you maybe five minutes. What it buys is a career’s worth of clear lungs and unburned corneas, and a radiator repair you can finish without wondering later whether the headache was the fume or the frustration. Aerospace welding standards and formal aluminium welding training both hammer the same point: safety is a system, not a checklist you read once.

Repair or Replace? Final Verdict and Practical Checklist

After everything in this article, the decision still comes down to one honest question: is this radiator failing in one spot, or is it failing everywhere? An isolated pinhole in a sound tank, welded to clean material with the right filler, is a repair worth making. A core that has started weeping in three places is telling you its service life is over, and no amount of TIG work will change that.

When TIG Repair Makes Sense

  • The leak is isolated. One pinhole, one cracked seam, one damaged tank corner, with solid metal surrounding it.
  • The core is otherwise healthy. Tubes are straight, fins are undistorted, and there is no white oxidation powder or pitting along the tube ends.
  • There is clean material to weld to. You can brush back to bright aluminum at least a few millimeters beyond the defect on both members.
  • The job makes financial sense. A custom or hard-to-find core, or a fresh restoration where a replacement would mean re-fabricating mounts and hose routing, almost always justifies a repair.

When You Should Replace Instead

  • Repeated pinholes in the same area. This is the biggest warning sign. Popping one hole only to find another two a week later usually means internal corrosion, not a one-off defect. You are chasing a symptom.
  • Salt exposure. Marine and road-salt environments attack aluminum heat exchangers far faster than a sealed, well-maintained system. If the tubes are chalky, thinned, or perforated in clusters, the wall thickness simply is not there for a durable weld.
  • Multiple failures or distorted tanks. Warped header plates, crushed tanks, and leaks at several joints point to a unit that has been overheated or stressed repeatedly.
  • Repair cost approaches replacement. If a straight fix costs 60–70% of a quality new unit with a warranty, replace it and keep the TIG work for something worth saving.

Quick Note: Filler choice matters more than most people expect. 5356 filler is the better pick where salt-water corrosion resistance matters, while 4043 flows more easily and is preferred for cast A356 sections and crack-sensitive joints.

The Shop Checklist

  1. Drain, flush and fully dry the radiator; coolant residue causes porosity.
  2. Mark the leak, then inspect the surrounding area for pitting and nearby hairline cracks.
  3. Scrub mechanically with a stainless brush, then solvent-wipe. Follow a proper sequence like the one in Aluminum Cleaner for Welding: How to Prep Aluminum Right.
  4. Decide: isolated defect = repair; clustered corrosion = replace.
  5. Fit up, tack, and back-step weld with pedal modulation, feeding filler into the thicker member.
  6. Pressure test, flush again, and touch up corrosion protection.
  7. If you are new to thin aluminum, build control first with practice runs on scrap or flattened aluminum cans; structured training such as an Aluminium Welding Course shortens that learning curve considerably.

Final takeaway: aluminum radiators are routinely saved with TIG, but only clean prep, correct filler, and restrained heat input decide whether the fix lasts. Know the difference between a bad spot and a bad radiator, and you will never waste a weekend chasing a repair that was never going to hold.

Frequently Asked Questions

What amp setting do I use to TIG weld an aluminum radiator?

Start low and build up. Thin tube walls around 0.020–0.050 in often need only 25–60 A with a 1/16 in or 0.040 in tungsten, while tank and header material of 1/8 in may take 125–175 A. Use the pedal to keep the puddle fluid without blowing through, and aim heat at the thicker tank or header so it sinks heat away from the tube.

Should I use 4043 or 5356 filler for an aluminum radiator?

4043 is usually the better choice: its silicon content makes it less crack-sensitive on 6000-series sheet and cast aluminum tanks, and it flows and cleans up with less smut. Choose 5356 when you need higher strength or better salt-water corrosion resistance, such as on 5xxx-series components.

Why does my weld keep making pinholes when the radiator had coolant in it?

That is outgassing. Trapped coolant, oil and combustion residue heat up and release gas through the molten puddle, so the last pinhole will not close. Flush and dry the core thoroughly, leave the cap off so it can vent, mechanically clean the area, and if a pinhole persists, drill it out to about 1/16 in to remove contaminated metal before re-welding.

Do I need an AC TIG machine with adjustable balance and frequency?

You need AC for aluminum, because the electrode-positive half of the cycle strips the oxide layer. An inverter machine with adjustable balance and frequency gives much better control: roughly 75% EN and 100–120 Hz is a solid starting point, with more EN for penetration and higher frequency for a tighter arc on thin material.

Can I MIG weld or braze an aluminum radiator instead of TIG?

You can, but each trade-off matters. MIG puts in more heat and is harder to control on very thin tubing, though it works on thicker tanks and seams. Oxy-fuel brazing with zinc-aluminum rod melts around 700°F and is gentler and more forgiving, but the joint is not as strong as a fusion weld. On thin, visible repairs TIG gives the cleanest result.

How do I test the radiator after welding?

Plug the outlets and pressurize the core lightly with air, then submerge it in water or spray the seams with soapy water and watch for a steady stream of bubbles. Never pressure-test with high-pressure air on a hot system, and always release pressure slowly before handling fittings.

Is welding an aluminum radiator safe from a health standpoint?

It produces aluminum welding fume plus ozone from the intense arc radiation, and argon can displace oxygen in enclosed spaces. Use local exhaust ventilation or work in a well-ventilated area, wear appropriate respiratory protection when engineering controls are not enough, and use a properly shaded welding helmet — OSHA’s welding fact sheet outlines the main controls.

When should I just replace the radiator instead of welding it?

Replace it when corrosion has thinned multiple tube walls, the core is heavily scaled, the tanks are distorted, or repairs keep popping up in different spots — that points to end-of-life rather than a fixable defect. An isolated pinhole or a cracked tank-to-header seam on an otherwise sound core is usually worth welding.

NH

Nicholas Harper

Editorial Team

Covering welding processes, equipment selection, and shop-floor practice for WeldZen — with every claim checked against the governing code.

Share this technical standard

Related Studies

Technical Dispatch

Get The Next Benchmark First

New field studies and calibration sheets, once a week. No spam.