Oxyacetylene Welding Aluminum: A Complete Guide

The filler rod touched the puddle and nothing happened. It sat there like a wet match, a fat silver bead refusing to sink into the aluminum, while the torch flame danced happily over the top of it. Ten minutes earlier the same rod had flowed like honey into a scrap of 6061 plate. The difference wasn’t the metal, the rod, or the operator’s hands — it was the oxide. Aluminum grows a skin that melts at roughly 2,060 °C while the metal underneath melts at 660 °C, so you can heat the puddle right up to the edge of burning and still never crack the film that keeps the two halves apart. Every good gas welder learns the same lesson eventually: on aluminum, preparation and flux matter more than the flame.

Why Aluminum Fights the Torch: The Physics Behind the Process

Oxyacetylene welding was the first process ever used to weld aluminum. Long before TIG and MIG existed, fabricators joined aluminum with a torch, a flux, and a filler rod, and the practice was so demanding that it built the reputation the metal still carries in many shops. GMAW and GTAW now handle the overwhelming majority of modern aluminum work, which is why so many welders assume the gas torch simply cannot do the job. That assumption is only half right. The torch can do the job. What it cannot do is hide from physics.

Everything that makes aluminum difficult to weld comes down to four material properties: melting point, thermal conductivity, the oxide film, and hydrogen solubility. Once you know what each one does to the puddle, the rest of the technique stops being guesswork. The history of aluminum welding is essentially the story of engineers working around these four numbers.

1. A Low Melting Point With No Warning

Pure aluminum melts at 660 °C (1,220 °F), and common alloys weld across a fusion range of roughly 520–660 °C depending on composition. That sounds convenient until you realize there is no visual cue on the way there. Steel announces its temperature, moving from dull red to bright orange to straw yellow to white, giving you a generous window to pull the heat back before you burn a hole. Aluminum does not do this. The surface stays a flat, silvery grey right up until it suddenly liquefies.

2. Heat That Runs Away

Thermal conductivity compounds the problem. Aluminum conducts heat at about 237 W/m·K at 300 K, roughly six times the rate of steel, as documented in the European Aluminium fusion welding guidance. Heat poured into a puddle rushes out into the surrounding plate and is gone before you can use it. The AWS guide to welding aluminum notes the same effect: the joint has to be heated faster and more locally than steel to keep the weld pool alive.

3. The Oxide Skin That Refuses to Melt

Aluminum forms aluminum oxide, Al₂O₃, the instant it meets air. On technical aluminum surfaces the layer is thinner than 0.01 mm, with a typical fresh film only 25–50 angstroms thick. It looks like nothing. It behaves like a lid.

Aluminum oxide melts at about 2,060 °C (roughly 3,750 °F), around three times the melting point of the base metal. Your puddle is at 660 °C. The film is at 2,060 °C. It never dissolves. It sits on top of the molten metal and refuses to merge with it, and the weld ends up riding on the surface instead of fusing into it. This is the entire reason gas welding aluminumrequires flux: there is no arc here to spatter the skin away the way a TIG cathode does, so a chemical agent has to break the bond for you. Our overview of how aluminum flux actually works digs into that mechanism in detail.

4. Hydrogen: Dissolved Hot, Trapped Cold

Hydrogen is very soluble in liquid aluminum but has almost nowhere to go once the metal solidifies. Solubility falls from roughly 0.807 weight percent in liquid aluminum at the melting point to about 0.036 weight percent in the solid state. The metal releases its hydrogen as the pool freezes, and if the weld cools faster than the gas can escape, it stays trapped as porosity. That is the root cause of nearly every aluminum porosity problem, and it explains why aluminum welding demands meticulous preparation before the torch is ever lit. Moisture, oil, damp flux, and even a filler rod left open in a humid shop all supply the gas that ends up as voids in the bead.

Property	Aluminum	Steel (typical mild steel)
Melting point	660 °C (1,220 °F); alloys ~520–660 °C	~1,450 °C (2,640 °F)
Thermal conductivity	~237 W/m·K at 300 K	~45–50 W/m·K
Oxide behavior	Al2O3 skin melts at ~2,060 °C and will not dissolve in the puddle	Mill scale melts at or below the metal’s melting point and floats off as slag
Color change before melting	None. The surface stays grey until it liquefies	Strong glow progression from dull red to white

One more detail makes the oxide problem worse than it sounds. Aluminum boils at about 2,494 °C, which is below the melting point of the oxide clinging to its surface. The metal can be vaporized before its own skin gives up. And because aluminum oxide is denser than aluminum metal, any oxide that does get melted does not float to the top like slag in steel. It sinks into the pool and gets trapped as a solid inclusion. That inclusion is the seed of a crack.

Understand those four properties and the rest of the technique stops being guesswork. The metal gives no warning, the heat runs away, the skin will not melt, and hydrogen is waiting to be trapped. Every step that follows has a reason behind it.

Setup: Choosing Your Torch, Regulators, Filler Rods, and Flux

Most failed aluminum gas welds are gear failures disguised as technique problems. The operator blames the flame, then the rod, then their own hands, when the real culprit was a drifting regulator, a steel-contaminated wire brush, or a cracked torch tip.

The equipment list

More than a torch and two cylinders belongs on the bench:

• Two regulators with gauges — one for oxygen, one for acetylene. Gauges are not optional; pressure sets your flame.
• Separate hoses — red for fuel gas, green or blue for oxygen, rated for the service. Inspect for cracks.
• A torch with interchangeable tips — start small; #00 through #3 covers thin gauge. Oversized tips melt holes fast.
• A lighter — the spark-lighters sold for welding rigs, never a cigarette lighter.
• A dedicated stainless wire brush — aluminum only, brushed one direction. Steel contamination defeats it.
• A stainless chipping brush — for flux slag.
• Files and a beveler — plus a flat file for edge prep.
• Degreaser or solvent — acetone works well. Use it before brushing, every time.
• A preheat source — another torch, a warm oven, or a heat blanket.
• Backing strips and heat sinks — copper or brass bar, plus copper blocks to soak heat from the far side.
• A fire extinguisher — an ABC dry chemical unit, within reach, class rating visible.

Regulator pressures have a hard legal line, not just good practice. OSHA 1910.253(a)(2) prohibits acetylene above 15 psig (103 kPa gauge) or 30 psia (206 kPa absolute). If you remember one number from this section, make it 15. Above that, acetylene becomes unstable and can decompose on its own.

Filler rod selection

Here is the pleasant part of aluminum gas welding: the rods are cheap, and the same rod you buy for TIG is right for gas welding. AWS A5.10/A5.10M classifies bare aluminum and aluminum-alloy electrodes and rods for gas metal arc, gas tungsten arc, oxyfuel gas, and plasma arc welding. No special “gas welding rod” exists.

Rod	Alloying element	Base alloys	Best use
4043	Silicon, about 5%	6061, 5052, 5xxx/6xxx	General-purpose, good crack resistance. Safe default.
5356	Magnesium with silicon	5052/5053, stronger 6xxx, marine alloys	Stronger deposit, good ductility. Preferred for marine and high-strength work.
4047	Silicon, about 12%	Thin sheet, cast aluminum, brazing repair	Lower melting range. Flows easily when filler must melt first.
1100	Commercially pure aluminum	1100, 3003, soft sheet	Matching pure-aluminum work. Soft, low-strength, very formable.

For deeper rod-to-base-alloy matching, our aluminum MIG welding guide covers the same logic from the arc side.

Why gas welding aluminum requires flux

Gas welding aluminum differs from gas welding steel in one big way. On steel the filler simply melts into the hot joint. On aluminum the oxide film stands in the way, and a torch flame will not remove it. TIG and MIG rely on an arc whose cathodic cleaning action blasts the oxide away. A gas torch has no arc, so flux is mandatory.

The standard choice is a fluoride-based flux, from the sodium fluoride or potassium fluoride families, sold as powders or pre-applied compounds for aluminum gas welding. They do the work the arc would do: dissolve the alumina film and let the filler wet into the base metal.

There is a separate family: chloride-containing fluxes, which AWS’s aluminum welding troubleshooting reference treats as its own category, mainly for brazing aluminum-copper and aluminum-zinc alloys. It also warns that leftover flux residue causes corrosion. You are not done when the bead looks good. You are done when the part is washed.

One shortcut: flux-cored aluminum brazing rod, with the flux inside. One has a solidus of 1,045 °F (563 °C), a liquidus of 1,110 °F (599 °C), and a brazing range of 1,080–1,120 °F (588–604 °C). Against aluminum’s 660 °C melting point, once the flux cracks the oxide, thin aluminum is nearly at welding temperature already. That is why thin-sheet aluminum gas welding is so fast, and so easy to blow through. Our guide to aluminum flux welding explains the chemistry.

A practical note on consumables

Aluminum filler rod and aluminum gas welding flux are cheap consumables, and quality pays off directly. Clean rod, dry flux, aluminum-specific product, sealed packaging. Each of those four words removes a known cause of porosity. Neither a spitting puddle nor a dirty rod is a torch problem.

Affiliate disclosure: some links in this article may point to equipment and consumable suppliers, and we may earn a commission if you buy through them. That never changes what we recommend.

Preheating, Joint Prep, and Cleaning: Getting the Base Metal Ready

A great torch, the right rods, and a clean neutral flame will not save a bad joint. Beads fail on aluminum because the metal was cold, the fit-up was wrong, or the surface was dirty. Preparation is the part of oxyacetylene welding aluminum that decides what happens next.

Aluminum pulls heat away from the joint roughly six times faster than steel, so the puddle shrinks and stiffens before filler can fuse into it. Preheat banks heat in the base metal. Good fit-up and cleaning handle the other half: bonding to clean metal, not floating on an oxide skin.

Why aluminum needs preheat

On carbon steel the torch heat stays in the workpiece long enough to fill. Aluminum behaves like a heatsink: the moment you stop sweeping the flame, the joint temperature races downward and the weld metal can freeze before the filler wets in. Preheat narrows that race and lowers the temperature difference between the pool and the surrounding metal, where residual stress and cracking begin.

Rough shop practice, though defer to the alloy supplier’s data sheet:

• Sheet under about 1/8 in. (3 mm): no preheat, or a light torch warming to take off the chill. More distorts the panel or burns through.
• Plate from about 1/8 to 1/4 in. (3 to 6 mm): roughly 200 to 300 °F (95 to 150 °C).
• Heavy plate and castings: roughly 300 to 400 °F (150 to 200 °C).

The Hobart Brothers aluminum welding guide makes the same point: preheat the whole part, not one hot spot. A torch sweep that only warms the joint leaves cold metal on either side that steals heat as fast as you add it. For larger work, a warm oven, heat blanket, or resistance heater plates beat a hand torch. Stage tools, rod, flux, and shield first so the weld can start as soon as the part reaches temperature. See the aluminum flux welding basics for how flux timing works with a preheated joint.

Joint prep: fit-up, bevels, and backing

Fit the joint close, but not with zero gap. A small opening, roughly a sixteenth of an inch for sheet and up to an eighth of an inch for thick plate, lets the filler flow so you get penetration instead of a bead on top. Leave a big gap and the filler drops through the joint, burns the edges, and undercuts the sides.

Bevel heavier sections rather than melting a thick edge through. A generous V-groove, roughly 60 to 75 degrees, gives the torch a path to the root and lets filler reach the bottom. On thin material, clamp a copper or brass backing strip behind the joint or fit a heat sink so you can push the puddle without burning through. Keep restraint light; aluminum’s high thermal conductivity already demands more heat input than steel.

Cleaning in the right order

1. Degrease with solvent first. Acetone or a dedicated aluminum cleaner, wiped on and wiped off. Oil and grease become hydrogen sources and cause porosity.
2. Remove the oxide mechanically next. Scrape or brush it with a clean stainless-steel brush dedicated to aluminum, used one direction only so it does not embed grit.
3. Weld immediately. Do not let the part sit.

Aluminum forms an oxide layer the instant a fresh surface is exposed, and the complete layer takes more than days to form, so a 20-minute break puts you back at square one. AWS troubleshooting guidance treats re-cleaning after any interruption as standard practice. Our walkthrough of how to prep aluminum for welding covers the products and techniques.

Cast aluminum gets special treatment

Castings carry more silicon, unpredictable grain structure, and often hidden thin sections, so they need more preheat than wrought material and slower, more even heating. Large castings with thin sections need more heat than small uniform ones, and preheat also reduces thermal stresses that cause cracking.

If preheating exposes a crack you had not seen before, treat it as an original crack, not a weld defect. Let the casting cool back down slowly to room temperature, because air cooling is too fast for a stressed casting. Once it is cool, cut the crack out to sound metal and weld it as a repair. For context on cast aluminum repair work, see the alloy wheel welding guide.

Thickness	Preheat	Joint prep	Filler
Sheet under 1/16 in.	None or light torch warming	Close butt, small gap, backing strip	1/16 in. rod
Sheet 1/16 to 1/8 in.	Light warming	Close butt, small gap, backing strip	1/16 in. rod
Plate 1/8 to 1/4 in.	200 to 300 °F	Square butt, small gap, bevel heavy edge	3/32 in. rod
Plate over 1/4 in.	300 to 400 °F	60 to 75 degree V-groove, 1/8 in. root gap	1/8 in. rod or larger
Cast housing or cover	300 to 400 °F, whole casting, slow and even	Cut to sound metal, V-groove crack	1/8 in. 4047 or 4043 rod

Filler rod designations follow the AWS A5.10/A5.10M classification for aluminum electrodes and rods.

The Flame: Neutral, Carburizing, or Oxidizing?

The oxyacetylene flame is commonly cited at roughly 6,300 °F (3,500 °C) at the torch tip, and aluminum melts at 660 °C. Heat was never the problem. The problem is what the flame does to the puddle while it melts. Get it wrong and the bead floats on top of the base metal, the filler rod balls up instead of flowing, and the weld turns grey and brittle. Get it right and aluminum behaves like butter.

Flame adjustment is the skill that separates a good gas welder from a bad one, and the hardest one to learn from a book, because you adjust by ear and by sight at the same time.

Three Flames, Three Different Jobs

A neutral flame burns acetylene and oxygen in roughly equal proportion, about a 1:1 ratio, and shows a crisp, bright inner cone inside a softer pale blue envelope. It is the flame most people picture when they think of a welding torch.

A carburizing flame, also called a reducing flame, is acetylene-rich. The inner cone grows longer, ragged, and less sharply defined, and it typically pushes forward past the tip. A slightly carburizing flame is generally the right choice for aluminum gas welding, because aluminum has a strong affinity for oxygen. A mildly reducing flame limits further oxidation of the puddle, and any extra oxygen you keep out of the flame is oxygen that does not turn a clean aluminum surface into stubborn alumina.

An oxidizing flame is oxygen-rich. The inner cone turns short and sharp, the flame roars loudly, and the reaction temperature climbs. On steel that can help with speed. On aluminum it is a liability: it oxidizes the puddle, drives the alloy toward embrittling behavior, and produces the grey, dull, dirty-looking bead that shows up in just about every aluminum troubleshooting table ever written. You do not want an oxidizing flame when gas welding aluminum. Ever.

Because aluminum is so oxygen-hungry, the practical target sits just on the carburizing side of neutral, maybe 10 to 15 percent excess acetylene. If you are unsure, err toward slightly reducing. An oxidizing flame damages aluminum; a mildly reducing flame merely burns a little more fuel.

Flame Type	Inner Cone	Effect on Aluminum	When to Use
Neutral	Short, sharp, bright cone in a blue envelope; no roar	Clean bead, less oxidation protection	Thin aluminum needing a hot, tight puddle
Carburizing (reducing)	Longer, ragged-edged cone extending past the tip	Limits further oxidation; slightly cooler tip	Preferred for most aluminum gas welding
Oxidizing	Very short, sharp cone; loud roar	Oxidizes and embrittles; grey dull bead	Do not use on aluminum

Setting the Flame in Practice

Here is a repeatable procedure, whether you are setting up for the first time or fixing a flame that drifted mid-job.

1. Light with a slightly carburizing flame. Point the torch away from combustibles, crack the acetylene slightly and ignite it, then bring the oxygen up gradually.
2. Trim oxygen until the inner cone is clean and sharp and the flame stops roaring. The roar is your ears doing the job of a flame analyzer. Squeeze the oxygen lever and the flame will shorten, brighten, and quiet down.
3. Trim acetylene to the smallest tip that will do the job, until the cone is clean and well defined with a sharp white core. Too little acetylene gives a weak, wandering flame that will not concentrate heat.
4. Match the tip to the thickness. Small tips such as #00 through #3 suit thin gauge; step up for heavier plate. A tip too big for the tip holder is not just inefficient, it is a safety problem and a flashback risk.

Reading the Flame

The inner cone is the primary heating zone, where the actual combustion reaction happens, and it is where you melt the filler rod. Put the rod tip into that cone and hold it until it beads with molten metal, then move it down to the puddle and feed it in.

That distinction matters more on aluminum than on steel, because you get no color feedback. Aluminum does not glow before it melts, so the cone and the filler behavior are your only instruments. When the rod melts freely in the cone and drops into the puddle, your heat is right. When the rod sits in the flame and slowly smokes without melting, you are short of heat or the cone is not where you think it is.

Most beginners hold the rod in the puddle and hope, and that is the most common reason a filler rod “won’t flow.” The puddle sits at or below the melting point of the base metal, already at its own limit, so the rod needs the hotter inner cone to melt before the puddle will accept it. The flux behavior that confirms this is covered in how aluminum flux welding works.

Flashback and Backfire

A backfire is a brief pop or bang as the flame burns back toward the mixer. A flashback is a sustained flame traveling backwards through the torch into the fuel line, which can ignite the cylinder. Both are serious, and the usual causes are well documented:

• A cracked, damaged, or mismatched torch tip, or a tip too large for the tip holder
• Moisture or contamination inside the torch or the fuel line
• Excessive tip pressure, which can break the seal and let the flame back up
• Oxygen backing up into the fuel gas line, typically from excessive oxygen pressure or a faulty regulator
• An improperly installed or contaminated flash arrestor

Prevention is mostly mechanical discipline: correct tip for the holder, correct pressures, dry equipment, working flash arrestors. Never crack a fuel gas valve near a flame, because the escaping gas can flash back into the cylinder. This is not optional guidance. OSHA 1910.253 requires that fuel gas mixtures be guarded against flashback and that approved torches and regulators be used, and specifically in 1910.253(b)(5)(iii)(I), that fuel gas must never be taken from cylinders through torches or devices with shutoff valves without a suitable regulator attached to the cylinder valve.

For broader shop rules on torch and cylinder handling, 1910.252 covers general welding safeguards, and 1926.350 applies the same thinking on construction sites. For the case on why an arc process handles the oxide problem for you instead, see the aluminum MIG guide.

The Bottom Line on Flame Control

Set a slightly carburizing flame, keep it quiet, keep the inner cone sharp, and melt the rod in the cone rather than the puddle. Everything else in aluminum gas welding depends on that. If you plan to practice on scrap before touching a real part, read our aluminum cleaning and prep guide in the same sitting, because a beautiful flame still cannot fuse a dirty joint.

Step-by-Step Technique: From Lighting the Torch to a Finished Bead

Everything before this was preparation. Now comes the actual weld, and on aluminum the details matter more than they do on steel. The biggest technical mistake is to dip a cold filler rod into the puddle and wait for it to melt. Gas welding runs the opposite way: the filler has to melt in the flame’s inner cone first, then be carried into the pool as a liquid. Drop cold rod into molten aluminum and you get a cold lap, a bead sitting on top of the metal instead of fused into it.

1. Check the work area

Look around before you touch a valve. Aluminum sheet, flux residue, rags, cardboard, and shop dust all burn fast and hot. Clear a radius of at least 10 feet, keep an extinguisher within reach, and set up a screen near combustibles. A solid grasp of oxy-fuel torch basics and shop safety pays off the first time you work next to a wooden bench.

2. Connect the regulators and open the cylinders

Open the oxygen valve first, then crack the acetylene valve before you attach the regulator, so acetylene pushes air out of the cylinder valve and outlet fitting. Do it the other way and that air ends up in your torch, turning the first ignition into a bang instead of a flame. Set the working pressures, red hose to acetylene, green or blue hose to oxygen.

3. Purge and light the torch

Point the tip away from your body and anything flammable. Open the acetylene valve and let gas flow for several seconds to purge, then ignite the acetylene stream. Open the oxygen valve only after you have a stable acetylene flame. Never light a mixture already fed with oxygen; it can flash back into the hoses and, worst case, the cylinder. OSHA 1910.253 requires approved apparatus and that fuel gas never be taken from a cylinder without a regulator.

4. Adjust the flame

Light with a slightly carburizing flame, then trim oxygen until the inner cone is clean and sharp and the flame stops roaring. Trim acetylene down to the smallest tip that will do the job. For aluminum you want neutral to slightly reducing. The inner cone is the primary heating zone, and it is where the filler rod has to sit to melt.

5. Heat the base metal

Bring the joint to preheat temperature before you touch the filler. Use a wide, sweeping flame over the entire part, not just the joint line. Aluminum dumps heat into the base metal faster than steel does, so a small hot spot under the puddle means a cold joint everywhere else. Thin sheet may need nothing more than warming; plate and castings need a real preheat range from the alloy data sheet.

6. Apply the flux

Flux goes on before heating, not after. Paint it along the joint, let it sit on the warm metal, and work it into the joint with the flame. Good flux melts, bubbles, and turns glassy, and the black oxide film visibly breaks down into the liquid. That bubbling and glassy flow is your first real temperature signal, because aluminum never tells you anything by color. How aluminum flux actually works is worth reading first.

7. Bring the joint to welding temperature

Keep heating until the flux is fluid and the base metal softens and shines. On steel you would watch the metal glow and the puddle form. On aluminum the metal just looks like metal, so you read the flux and the surface. A wet, slightly sweating sheen with the flux running freely means you are there.

8. Melt the filler in the flame, then feed the puddle

Hold the rod in the inner cone until a bead of liquid metal forms on its end, then move it down into the puddle so the molten filler flows in. Two techniques exist:

• Forehand (push): The torch travels ahead of the puddle and the rod is fed behind it, pushing molten filler forward into fresh metal. It is the more common choice on aluminum because the flame preheats the base metal ahead of the pool and keeps the puddle small.
• Backhand (pull): The torch trails the puddle and the rod leads it, so you drag the weld behind you. It gives better control of the pool edge, and some welders prefer it on castings and repair work.

Whichever you use, keep the rod in the flame between strokes. A rod held in the puddle stays cold, and cold rod does not fuse.

9. Control the puddle

Hold a small, consistent pool. Keep the torch at roughly a 45-degree drag angle and use a slight side-to-side or circular motion to spread heat evenly. Do not let the pool grow. A large molten pool on aluminum means burn-through on thin material, excessive heat input, and distortion you will spend an hour straightening out. Short, frequent passes beat one long soak.

10. Run out, finish, and clean up

Carry the puddle along the joint and let it freeze naturally rather than snuffing it. Then chip the remaining flux slag off with a stainless steel chipping brush, wash the part down with hot water and a stiff brush to remove residue, and let it cool undisturbed. Getting aluminum clean before and after welding covers the prep side of the same job. Flux residue left on a part will go on to corrode it.

Troubleshooting by what you see

What you see	What it means	What to change
Rod balls up and floats on the puddle	Filler is not melting before it reaches the pool	Hold the rod in the inner cone longer; check the flame; raise heat slightly
Puddle spits and pops	Moisture or contamination	Dry the flux, keep the rod sealed and clean, purge the torch properly
Weld floats on top of the base metal	Lack of fusion, oxide still in the joint	More flux applied before heating; re-clean the joint
Base metal melts before the filler flows	Too much heat, or the rod is too high-melting	Reduce heat input and travel faster, or switch to a lower-melting filler such as 4047
Weld turns grey or dull	Oxidizing flame	Close the oxygen slightly and add acetylene until the flame stops roaring

Filler Metal, Alloy Selection, and Joint Design

Before you light the torch, two decisions decide whether the job is worth attempting: can this alloy be welded at all, and will the joint you drew hold up? Aluminum punishes bad metallurgy faster than steel, giving no glow, no color change, and no second chance at a ruined heat-affected zone.

Weldability comes down to the alloy series

The AWS guide to welding aluminum sorts alloys by how they respond to fusion welding. The 1000, 3000 and 5000 series, plus some 4000 alloys, are generally easily weldable. The 2000, 6000 and 7000 series, and certain 4000 alloys, can be welded but demand more heat input, faster travel speed, and rigidly controlled procedures.

The 4000 series exists largely as welding filler rather than base metal. That is why 4043 and 4047 rods sit on nearly every shop shelf, yet you rarely see a 4043 plate. Rod selection is covered in how aluminum flux core welding actually works, and the practical guide to aluminum MIG welding covers the same filler choices for an arc process.

The alloys that will fight you

Two families deserve a hard warning. The 2000 series (2024, 2014) and 7000 series (7075, 7010) are high-strength alloys that suffer hot cracking in the weld metal and heat-affected-zone softening beside the bead. The European Aluminium fusion welding guidance treats them as high-risk for any fusion process. In practice they are usually joined by rivets, brazing, or friction stir welding.

5xxx versus 6xxx, and why 6061 is the default

The 5xxx alloys (5052, 5053) have a lower melting range than the 6xxx series, so they weld more easily: the filler fuses over a wider window and the puddle is forgiving. The trade-off is distortion. Magnesium alloys show no heat-affected-zone strengthening loss, but they move more than 6xxx grades when they heat up.

For general structural work, 6061-T6 remains the most common general-purpose alloy and a sensible default for trailer frames and machine guards. It is heat treatable, so weldments normally need re-solution heat treatment and aging afterward to restore strength. That work belongs to a heat treat shop, not the person holding the torch.

Matching filler rod to base metal

Most shops follow these rules of thumb:

• 4043 (aluminum-silicon) for 6061 and general 5xxx work. Its silicon content improves fluidity and crack resistance in the puddle, which matters when you manage a small pool by feel.
• 5356 (aluminum-magnesium-silicon) for higher-strength 6xxx work and marine applications, where matching 5xxx base metals are common.
• 4047 (aluminum-silicon, lower melting range) for thin sheet, hard-to-reach joints, and brazing-style repairs where the filler should flow at the lowest possible temperature.
• 1100 for pure aluminum base metal.

Rod specification is standardized under AWS A5.10/A5.10M, which classifies bare aluminum and aluminum-alloy electrodes and rods for gas metal arc, gas tungsten arc, oxyfuel gas, and plasma arc welding. Your rod answers to that same standard whether you feed it through a torch or a MIG gun.

Joint design is metallurgy too

Design for welding before you cut metal. Prefer simple butt and lap joints, and avoid heavy restraint, because aluminum contracts as it cools and a joint that cannot move will crack. On heavy sections, use a generous bevel: a commercial flux-cored aluminum brazing rod specification calls for a 60° to 75° V-notch bevel with a gap not exceeding 1/8 inch, keeping the arc or flame focused on the puddle.

Aluminum’s high thermal conductivity forces higher heat inputs to reach fusion temperature, and that extra heat raises distortion risk. Parts can shift mid-weld, so plan a sequence, use strong fixturing, and back-step long seams to control heat.

Base alloy	Weldability	Recommended filler	Notes
1100 / 1350 (pure)	Easily weldable	1100	Soft, low strength; keep heat minimal
3003 / 3005	Easily weldable	4043 or 5356	Good corrosion resistance
4043 / 4047 (4000)	Mostly filler alloys	4043 / 4047	Rarely base metal; 4047 flows lowest
5052 / 5053 (5000)	Easily weldable	5356 or 4043	Lower melting range; distorts more
6061-T6 (6000)	Higher heat, faster travel	4043	Common structural alloy; re-heat treat after welding
2024 / 7075 (2000, 7000)	Poor: hot cracking, HAZ softening	4047 under a qualified procedure	Prefer rivets, brazing, friction stir welding

Troubleshooting: Porosity, Lack of Fusion, Inclusions and Cracks

A bead that looks fine can hide a honeycomb of holes, and a joint can sit on the plate like bad solder, refusing to fuse. Nearly every ugly aluminum weld traces back to four causes: hydrogen gas, trapped oxide, wrong filler, or heat that went wherever it wanted.

Porosity: the invisible gas problem

Hydrogen is the biggest destroyer of aluminum welds, and it starts with a simple asymmetry: it is extremely soluble in molten aluminum but barely soluble in solid metal. As the puddle solidifies, dissolved hydrogen is squeezed out. Cool slowly and it escapes, giving a sound bead; chill too fast and it stays trapped, giving you porosity.

That is why the same rod and joint can give a clean weld in the afternoon and a bubbly mess the next morning, and why cooling rate matters as much as cleanliness. Any hydrogen-containing compound on the rod or the work surface causes porosity.

Real-world hydrogen sources, in the order you will most likely meet them:

• Moisture in the air. Aluminum oxide is porous and soaks up contaminants, a ready-made hydrogen source at the joint.
• A hydrated oxide film. The film holds water, and improper cleaning right before welding is a recognized cause of porosity.
• Oil and grease on the plate, the filler rod, or on gloves that touch the rod.
• Damp or contaminated flux. Flux stored open in a humid shop can be worse than no flux at all.
• Rod left out in the open. Wire left on the bench overnight absorbs moisture into its oxide layer, and you carry that into the puddle.

The fixes are unglamorous but reliable. Degrease with solvent, then brush the joint with a clean stainless brush immediately before welding, not twenty minutes before. Wipe the rod with acetone, keep it sealed, keep the flux dry, and where the joint allows, slow the cooling down.

Lack of fusion: the weld that sits on top

When the bead rolls off the puddle like mercury and leaves a clean gap at the toe, the metal never melted into the base material. Four causes account for nearly all of it:

1. Oxide not removed. Insufficient flux, or flux on a cold joint that powder-sits there instead of flowing.
2. Not enough heat. The base metal never reached welding temperature, so the filler melted but the parent metal did not.
3. Wrong flame. An oxidizing flame burns the film faster than you can clear it.
4. Torch angle pushing oxide ahead of the pool. Drag the flame with the direction of travel and you sweep the film ahead of the puddle.

Inclusions: oxide that sank and stayed

Aluminum oxide is denser than aluminum. Melt a bit of the film and it does not float, it sinks into the puddle and stays trapped, leaving unmelted gray specks and dark streaks inside the bead, plus reduced strength and poor fatigue life. Because oxide is also an insulator and a poor wetting surface, it doubles as a lack-of-fusion site.

Common causes: an oxidizing flame, a joint that re-oxidized while you set up, flux heated too hot so it burned instead of dissolving the film, and a dirty torch tip dropping scale into the pool.

Cracking, burn-through, undercut and distortion

Hot cracking in aluminum usually means the joint was too restrained for the shrinkage the alloy wanted to make. It also shows up when the filler is wrong for the base alloy, when heat input is excessive, and when the weld cools too fast for the alloy’s solidification range. Severe restraint on a thick section is the classic trigger, so plan your flux practices and joint fit-up with that in mind.

On thin material the opposite failure shows up: burn-through when the puddle grows too large, undercut when the torch angle is too steep, and distortion when the heat is unbalanced. Cure them with a smaller tip and puddle, a backing strip or heat sink, a lower torch angle, even heating across the part, heat sinks, back-stepping and strong fixturing. Aluminum’s thermal conductivity pulls heat away fast, which tempts you into cranking up the input, and the extra heat bows the part.

The defect table

Defect	What it looks like	Root cause	Corrective action
Porosity	Round pinholes, sugary or foamy bead	Hydrogen from moisture, oil, wet flux or damp rod; fast cooling	Degrease, brush clean just before welding, wipe rod with acetone, keep rod sealed, dry the flux, slow the cooling
Lack of fusion	Bead on the surface, clean gap at the toe	Oxide not removed, low heat, oxidizing flame, torch pushing film ahead	Flux a warm joint so it flows, correct the flame, hold a proper drag angle
Oxide inclusions	Dark specks and streaks in the bead, gray unmelted patches	Oxidizing flame, re-oxidized joint, overheated flux, dirty torch tip	Neutral-to-slightly-carburizing flame, re-clean, keep flux below burnout, clean the tip
Hot cracking	Crack along or through the bead, often at the toe	Excessive restraint, wrong filler, high heat input, rapid cooling	Better fit-up and fixturing, match filler to the alloy, lower heat input, cool slowly

What to do before you weld aluminum again

A few habits catch most of these problems before they start: keep the joint and rod clean, keep the flame neutral, control the heat. If a joint keeps giving you trouble, check what process the job really calls for. When porosity is persistent and the joint is not a simple butt, aluminum MIG welding will usually outrun a torch, and aluminum flux core welding has its own process logic worth learning.

I’ll rewrite this section to fit the word count while keeping the heading, facts, and all links.

Let me first check the current word count of the fragment.

Safety: Cylinders, Torches, Fumes, and Flux

Oxyacetylene welding puts a small chemical plant in your shop: two pressurized cylinders, a torch hotter than most metals, and, for aluminum, a fluoride flux that fumes. Most accidents start with a skipped step.

The rules behind this section: OSHA 1910.253 for oxygen-fuel gas welding and cutting, OSHA 1910.252 for general welding requirements, and 1926.350 plus shipyard rule 1915.55. The AWS Welding Aluminum guidance and the OSHA welding fume fact sheet cover aluminum hazards: flux, fume, and molten metal.

Cylinder Handling: The Rules That Prevent Most Accidents

Acetylene is not a stable gas at cylinder pressure. Under 1910.253(a)(2) it must not be generated, piped, or utilized above 15 psig (103 kPa gauge) or 30 psia (206 kPa absolute), and liquid acetylene is prohibited. Every acetylene regulator is a two-stage design that bleeds pressure to a safe working range.

1910.253(b)(5)(iii)(K) is specific about opening the cylinder: the acetylene valve shall not be opened more than one and one-half turns of the spindle, preferably no more than three-fourths of a turn. These cylinders are packed with a porous filler, and a wide-open valve lets gas bleed out faster than it can be replaced, creating a void that can collapse the cylinder. Open slowly, standing aside.

• Store fuel-gas cylinders valve end up; acetylene is lighter than air and drifts toward sparks.
• Keep oxygen 20 feet (6.1 m) from fuel gas and combustibles, or use a noncombustible barrier, per 1910.253(b)(4)(iii).
• Keep cylinders away from heat, sparks, open flame, and traffic that could topple them.
• Keep cylinders and equipment free of oil and grease; oxygen aimed at an oily surface ignites explosively, and 1910.253 forbids oxygen-cleaning.
• Never take cylinders into a confined space; 1926.350 and 1915.55 require this for oxygen and acetylene.

Torch, Hoses, and Shift Inspections

Under 1910.253, only approved apparatus may be used, and fuel gas must never leave a cylinder without a regulator attached. Hoses and torches are inspected at the start of each shift; defective gear comes out of service.

That inspection is quick. Run hoses end to end for cracking, blistering, chafing, and damaged collars, and check the handle, tip holder, and tip. Clean clogged tip openings with proper tools, never a wire wheel or blowtorch: a blocked tip changes the flame shape and flashes back. Confirm the regulators read correctly, hose colors match contents, and flashback arrestors are installed. Change tips with the valve closed and the pressure bled down. A cracked tip, an oversized tip, moisture, a tight tip, or oxygen backing up into the fuel line can each flash back into the hose.

The Aluminum Hazards: Fumes, Flux, and Molten Metal

Flux fumes. Gas welding aluminum without flux does not work, and the fluoride flux that makes it work gives off fumes that irritate eyes, nose, and throat. 1910.252(c)(1)(vi)(C) requires fluxes containing fluorine compounds to carry the wording “CAUTION CONTAINS FLUORIDES.” Without that label, it is not the flux you think it is. Residue must be removed, since flux left on a part causes corrosion. Our guide to aluminum flux welding covers the chemistry, including why chloride-bearing fluxes are a separate category for brazing aluminum-copper and aluminum-zinc alloys.

Welding fume. The OSHA welding fume fact sheet lists aluminum among metals found in welding fume and names process type, base metal, filler metal, location, and work practice as exposure drivers. Use local exhaust ventilation or a fume extractor; a respirator is the last step. For confined spaces, 1910.252(c)(2)(1) requires ventilation with clean, respirable replacement air, and 1910.252(c)(4)(v) bars oxygen for ventilation, since it enriches everything, flux fumes and flammable residue included.

Molten aluminum and water. Molten aluminum reacts violently with water. Never quench a glowing weld, never drop wet scrap into molten metal, never set a hot part where condensation or a wet rag can reach it. Let it cool undisturbed. This is the rule aluminum workers most often break. Add ordinary heat and eye protection: flame-resistant clothing, leather gloves, and the correct eyewear shade. Because aluminum shows no color change before melting, you cannot see the joint approach welding temperature, and a white-hot aluminum weld looks more dangerous than a dull red steel weld of the same size.

The End-of-Job Safety Checklist

Phase	Check Before You Proceed
Pre-job	Combustibles cleared or shielded; extinguisher in reach; ventilation running; correct eyewear shade
Cylinders	Chained upright, valve end up; oxygen 20 ft (6.1 m) away or barriered; away from heat; valve opened slowly, 3/4 to 1.5 turns; regulators on both
Torch	Inspected at shift start; correct tip size; tip unclogged; arrestors fitted
Flux	Aluminum flux, dry and sealed; marked “CAUTION CONTAINS FLUORIDES”; fume extraction running
During and after	Small puddle; never quench with water; cool undisturbed; chip slag, wash off residue, dry the part

Two documents belong on the shop wall. The OSHA welding fume fact sheet lays out the hierarchy of control for fume: source control, then ventilation, then respiratory protection. And ANSI Z49.1, Safety in Welding, Cutting, and Brazing, is the consensus standard behind the training, eye protection, ventilation, and cylinder handling practices. For the equipment side, read our air acetylene torch basics and safety guide and the complete aluminum MIG welding guide on the shielded alternative.

Oxyacetylene vs TIG vs MIG for Aluminum

There is no best aluminum welding process. There is only the best process for a specific job, and the mistake most shops make is choosing by habit instead of by the part in front of them.

Oxyacetylene gas welding earned its reputation the hard way. It was the first method used to weld aluminum, decades before TIG and MIG existed, and it still holds one advantage no arc process can match: it needs no electricity. A torch and two cylinders work on a farm gate, a rooftop, a trailer bed, or a pipeline field site where a power source cannot go.

Where Oxyacetylene Earns Its Keep

• No power source required. Field repair, farm equipment, rooftop flashing, remote structural work — the uncontested win for gas welding.
• Lowest capital and consumable cost. A torch, regulators and cylinders cost a fraction of a machine package, and consumables are a few dollars of rod and flux.
• Very thin sheet and tiny repairs. Managing a TIG arc on 0.4 mm aluminum is fussy work. A small tip, a slightly carburizing flame and a 4047 rod handle it faster with less burn-through risk.
• Brazing, heating and annealing. For building up a worn surface or softening a heat-treated part, the torch has no competition.
• Non-structural repair work. Cracked wheel spokes, brackets and housings are the kind of job a repair shop takes on without a qualified code procedure.

Where Oxyacetylene Falls Short

• With no shielding gas, flux is mandatory, and flux residue must be chipped and washed off or it will corrode the part. That is why surface cleaning and prep matter so much in this process.
• Heat control is crude. A flame is either there or it is not, so puddle size and dilution are far harder to manage than with an arc.
• There is no built-in cleaning action. The oxide is broken chemically rather than by the arc, so porosity risk is higher.
• It does not qualify for code work in critical structures. Structural aluminum falls under AWS D1.2, hull work under D3.7 and pipe work under D10.7, and none of them call for gas welding of aluminum.
• It is slow. Thick plate and production schedules belong to GMAW.

TIG (GTAW): The Clean, Code-Ready Answer

The arc does the work an oxyacetylene torch cannot. Cathode cleaning breaks the stubborn oxide film, so the weld needs no flux and leaves no residue. AC GTAW accounts for the vast majority of GTAW welding of aluminum because AC balances cleaning and penetration in one pass. TIG is the process referenced in structural, hull and pipe codes, and it stays the best choice for thin sheet where heat input must be controlled precisely.

The trade-offs are real: a machine, a power supply, argon, and a steeper learning curve in AC arc control. If you already run a TIG set up, aluminum makes it worth owning.

MIG (GMAW): The Production Standard

For thicker sections and repeat work, GMAW wins on speed and penetration. Because aluminum’s high thermal conductivity pushes heat out of the puddle fast, spray and pulsed-spray are the two recommended modes of metal transfer: spray lays down a deep, fast bead and pulsed-spray keeps the arc stable on thinner material. This is the process behind most structural aluminum fabrication. Our aluminum MIG welding guide covers the setup in depth.

And the Processes That Do Not Apply

Stick (SMAW) electrodes for aluminum exist, but they are not approved for code welding because the process is prone to porosity and flux inclusions that promote internal corrosion. Flux-cored arc welding and submerged arc welding have no commercial fluxes developed for aluminum and are simply not used on it.

Job Type	Recommended Process
Field or rooftop work with no power	Oxyacetylene
Thin sheet, non-structural patch	Oxyacetylene (small tip) or GTAW
Brazing, build-up, annealing, straightening	Oxyacetylene
Cracked wheel, bracket or housing repair	Oxyacetylene 4047 or GTAW
Boat hull seam	GTAW per AWS D3.7
Structural aluminum frame or stair	GMAW (spray / pulsed-spray)
Aluminum pipe or tube	GTAW or GMAW per AWS D10.7
Production shop, heavy plate	GMAW

So the honest verdict is this. Oxyacetylene is the right tool for thin material, brazing, field repair and non-code work, and the wrong tool for anything a code weld inspector will look at. If the part is structural, marine or safety-critical, spend the money on a TIG or MIG machine and a qualified procedure.

Real-World Projects, Pro Tips, and the Bottom Line

Four real repairs, the habits behind a clean bead, and when the oxy setup is right.

Project 1: Patching 1/16 in. Aluminum Sheet

The oxyacetylene sweet spot: a trailer skin hole, a torn fender lip, a cracked toolbox lid, thin enough that a TIG arc is overkill.

• Cut out the damage; leave about 1/16 in. between patch and sheet, close enough to hold heat, open enough for filler to flow.
• Bevel patch edges 15 to 30 degrees so filler has a path in.
• Back the joint with copper or brass: it holds heat and catches the puddle if you punch through.
• Flux both sides; on thin sheet the underside oxidizes as fast as the top.
• Smallest torch tip, sweeping flame, pool no bigger than a dime, travel slowly enough that filler melts in.

Chip the slag, wash with hot water and a stiff brush until nothing soapy comes off. Proper aluminum cleaning and prep matters, because residue left now starts a future crack.

Project 2: Adding a Tab or Bracket to a Hull or Trailer Panel

A bracket on a boat hull or trailer rail looks like a simple fillet weld: flux, small puddle, slightly carburizing flame. The complication is the paperwork, not the weld.

Treat it as a non-structural, non-code repair. Hull construction falls under AWS D3.7, and critical seams need a qualified GTAW procedure with a written WPS. Read our guide to aluminum boat welding and strong hull joints first.

For a non-loaded trailer tab the oxy route is defensible: use 4043 or 5356, keep heat input low to limit distortion, and clamp the panel to a fixture.

Project 3: Repairing a Cracked Cast Aluminum Housing

Cast aluminum was welded with oxy-fuel because it was the process available; after GTAW, oxy-fuel is rarely used for critical welds, and the AWS guidance on welding aluminum reflects that shift.

• Preheat heavily, roughly 300 to 400 °F. Large castings lose heat into every branch at once.
• Use a low-melting filler: 4047 flows at lower temperature and forgives a moving puddle better than 4043.
• Keep heat input low. Internal geometry is unpredictable, so a hot spot in one arm can crack another. Use stringer beads, let each cool, and crack the casting up before you start.
• Treat it as a field repair: it restores function, but it is not a qualified production method.

Project 4: Brazing-Style Repair With a Flux-Cored Aluminum Rod

Sometimes the joint needs wetting, not melting. A flux-cored aluminum brazing rod carries its own flux, so it stays protected in storage. Aluminum flux core welding has its own rules, and brazing rods rank among the most versatile torch rods for cracked extrusions and castings.

• Gap: no more than 1/8 in. Wider gaps need a different filler.
• Heavy sections: 60 to 75 degree V-notch bevel.
• Flame: slightly carburizing, torch held 1 to 2 inches from the work.
• Technique: let the filler flow without melting the base metal. Melt the parent and it becomes a welding problem.
• Working range: roughly 1,080 to 1,120 °F, solidus near 1,045 °F, liquidus near 1,110 °F.
• Cleanup: wash residue away with warm water and a stiff brush.

Building up a worn part is the same idea: the rod rebuilds a scarred bracket or a stripped flange face at far lower heat than a full weld.

Pro Tips From the People Who Burned Their Fingers

1. Practice on scrap of the same alloy: learn 6061 on 6061 before meeting 5052 or a casting.
2. Keep the rod in the flame, not the puddle. A cold rod dipped in a hot pool gives a bead sitting on top of the metal.
3. Heat the whole part, not just the joint. Aluminum pulls heat away fast.
4. Keep the puddle small and weld in short stringers. Heat input you can control beats heat input you can stop.
5. Clean immediately before welding. Twenty minutes is enough for a fresh oxide skin.
6. Chip and wash the flux off. The residue is hygroscopic and corrosive.
7. Never quench aluminum with water. Molten aluminum reacts violently with it.
8. Use heat sinks and backing bars; on thin material they separate a bead from a hole.
9. Wear proper eye protection. The flame is around 6,300 °F, and OSHA requires suitable protection: follow 1910.252 with ANSI Z49.1.

The Bottom Line

Oxyacetylene welding of aluminum is legitimate and learnable, and most failures are prep failures, not torch failures. Clean the metal, apply the right flux, heat the entire part, hold a small puddle, feed the filler from the flame. Those habits carry you through thin sheet, brazing, field repair, and non-code work with two cylinders and a torch.

But be honest about its lane. There is no shielding gas, so flux is mandatory and residue must be cleaned. There is no arc to clean the oxide, so the job leans harder on preparation. It does not qualify for code work in critical structures, so for structural or code aluminum, GTAW or GMAW with qualified procedures is the right answer, every time.

That is why many shops keep the oxy setup: torch for tacking, heating, straightening, and brazing, TIG for the structural weld.

Frequently Asked Questions

Can you oxyacetylene weld aluminum?

Yes. Oxyacetylene gas welding of aluminum is one of the oldest aluminum welding methods and is still widely used for thin sheet, brazing, heating and field repair work. It requires a flux because, unlike arc welding, there is no cathode cleaning action to break up the oxide film. It is not a substitute for TIG or MIG on code-governed structural work.

What flux is needed for oxyacetylene welding aluminum?

A fluoride-based aluminum gas welding flux, applied to the joint before heating so it melts and dissolves the oxide film. Chloride-containing fluxes are a separate category generally used for brazing aluminum-copper and aluminum-zinc alloys. Whatever flux you use, the residue must be chipped and washed off afterwards or it will corrode the joint.

What flame should I use when gas welding aluminum?

A neutral to slightly carburizing (reducing) flame. Aluminum has a strong affinity for oxygen, so an oxidizing flame will burn oxides into the puddle and embrittle the weld, while too much acetylene can leave a brittle carbon-rich deposit. Adjust until the inner cone is sharp and the flame is silent rather than roaring.

Do I need to preheat aluminum before gas welding?

Usually yes for anything thicker than about 1/8 inch, and always for castings. Aluminum conducts heat away from the joint roughly six times faster than steel, so the puddle cools before the filler can fuse. Thin sheet generally needs only warming, while heavy plate and castings are commonly preheated to roughly 300–400 °F — follow the alloy supplier’s data sheet for your specific material.

Why are my aluminum gas welds porous?

Almost always hydrogen. Hydrogen is very soluble in molten aluminum but barely soluble in solid aluminum, so it is expelled as the pool freezes and trapped if the cooling is fast. The usual sources are oil or grease on the plate or rod, moisture, a hydrated oxide film, and damp flux. Degrease with solvent, brush with a clean stainless brush right before welding, keep rod sealed and dry, and wipe it with acetone-dampened cloth before use.

Which filler rod should I use for oxyacetylene welding aluminum?

4043 (aluminum-silicon) is the general-purpose choice for 6061 and most 5-series work. 5356 (aluminum-magnesium-silicon) suits higher-strength 6-series and marine applications, and 4047, with its lower melting range, is a good choice for thin sheet and brazing-style repairs. AWS A5.10/A5.10M is the classification standard for aluminum filler rods used with oxyfuel gas welding.

Is an oxyacetylene aluminum weld as strong as a TIG weld?

No. A properly made TIG weld is cleaner, more consistent and code-qualifiable under documents such as AWS D1.2, D3.7 and D10.7, and it needs no flux cleanup. A good gas weld on suitable thin material can be perfectly serviceable for non-structural and field repair work, but it does not carry the same qualification or the same proven quality record.

Can you weld cast aluminum with an oxyacetylene torch?

Yes, and it was historically the standard method for cast aluminum before GTAW existed. Castings need even, generous preheat — commonly 300–400 °F — and a lower-melting filler such as 4047. It is best treated as a field repair rather than a qualified production process, and any pre-existing cracks found during preheating should be addressed as original cracks.

Why won’t the filler rod melt into the puddle?

Most often the puddle is too cold or the flame is oxidizing, and sometimes the oxide film has not been broken. Heat the base metal until it softens and the flux runs glassy, melt the rod in the flame’s inner cone so it arrives as a liquid droplet, and never dip a cold rod straight into the pool — a cold rod in a cold puddle just beads up and floats.