Hobby welding

Welding joins metal by melting the parent material so that two pieces become one, usually with additional metal added into the joint. That is what separates it from brazing and soldering, where a lower-melting filler sticks two pieces together without melting them, and the distinction is not pedantry: it is why a weld can be as strong as the metal around it and a braze generally is not.

What it gives a workshop is a change of category. Without it, metal is something you cut, drill and bolt; with it, metal is something you can make into any shape you can hold together long enough to tack. Gates, trailers, brackets, furniture, sculpture, repairs to things that were never meant to be repairable — the range is the reason people learn.

The question every beginner arrives with is which process to buy, and it is very nearly the right question. The reason it produces such bad answers online is that it gets answered as a ranking, and it is not one. MIG, TIG and stick are answers to different situations, and the situation is set by two things you already know: what you intend to weld, and where. The equipment section below is organised around that, because getting it right is worth more than any amount of technique advice.

The real first question

What, and where — not which machine

The hazard that hurts most

Fume, over years. Then the arc

What this page publishes

Decisions, hazards, diagnosis

What it does not

Settings, technique, structural work

1. 1

   Step 1: Do a day or an evening course first

   Colleges, makerspaces and a growing number of independent workshops run introductory welding courses for £80–£200. You will use good equipment, be told immediately what your hands are doing wrong, and — most valuable — find out which process suits you before spending anything.

   You will know it worked when: The purchase decision made from experience rather than from forums.

2. 2

   Step 2: Sort the space before the machine

   Ventilation, a non-combustible surface, somewhere sparks cannot reach, and a way to keep other people out of the arc. If the honest answer is that your garage cannot provide these, that is worth knowing before the delivery rather than after.

   You will know it worked when: A place you can actually weld in.

3. 3

   Step 3: Buy the helmet properly, and the machine adequately

   A good auto-darkening helmet transforms the experience and a cheap one causes headaches and eventually injury. The machine can be mid-range; the helmet should not be the place you save.

   You will know it worked when: The protective purchase made first.

4. 4

   Step 4: Weld scrap for several hours before you weld a project

   Pads of beads, then joints, on offcuts of the thickness you actually intend to use. Nobody’s first welds are any good and doing them on something you care about turns a learning exercise into a disappointment.

   You will know it worked when: A hand that has some idea what it is doing.

5. 5

   Step 5: Cut every weld you make in the first month

   Grind or saw through your practice joints and look at the cross-section. This is the single most useful habit in learning to weld, because the entire difficulty is that a weld’s quality is inside it and the surface tells you much less than you want it to.

   You will know it worked when: Feedback that a photograph of the bead cannot give you.

6. 6

   Step 6: Pick a first project that is allowed to be ugly and holds nothing up

   A workbench, a frame, a log rack, a fire pit. Not a towbar, not a trailer, not anything that goes on a car or over a person.

   You will know it worked when: Something useful, and no consequences from the learning curve.

What can be said usefully without teaching: welding is far more about preparation than about the arc. Clean, bright metal; joints that fit before you start rather than being pulled together by the weld; and enough tacking that the work does not move as it heats. The great majority of bad welds by beginners are bad joints welded competently, and a fit-up that needs the weld to close a gap is the commonest example.

The other honest thing to say is that heat distortion surprises everybody. Steel moves as it cools, a frame welded square comes out not square, and managing that — sequencing, tacking, letting things cool, welding a little on one side and then the other — is a real part of the craft that has nothing to do with how a bead looks.

Answer these two and the process mostly picks itself. What are you welding — mild steel of a few millimetres, thin sheet, thick section, aluminium or stainless? And where — an enclosed workshop, or outdoors and on site?

• MIG — the default, and usually the right one — A wire fed continuously through a torch under shielding gas. Easiest to get a usable weld with, quick, and well suited to the mild-steel fabrication that is most of what hobbyists do. Its weakness is that it is easy to make a weld that looks acceptable and has not fused underneath, which is exactly why the structural refusal above matters. Needs gas, so it dislikes wind and therefore dislikes outdoors.
• Gasless MIG — the caveat version — Flux-cored wire that needs no cylinder, which makes it portable and outdoor-capable and removes the rental contract. It spatters heavily, smells worse, and the results on thin material are poorer. A reasonable answer for occasional outdoor repair work and a poor one for a workshop.
• Stick (MMA) — the outdoor and thick-metal answer — A consumable coated electrode that generates its own shielding as it burns, so wind does not matter and neither does a slightly rusty surface. Cheap machines, robust results on thicker material, and genuinely harder to learn — the arc has to be struck and held at a distance you cannot see well. The traditional farm and site process for good reasons.
• TIG — the precise and slow one — An arc from a tungsten electrode with filler added by the other hand, like gas welding with electricity. It gives the most control and the cleanest results, it is the practical route to aluminium and thin stainless, and it is the hardest to learn because both hands and often a foot are doing separate jobs. Not a first process unless precision is the actual goal.
• The multi-process inverter — Machines doing MIG, stick and often lift-arc TIG in one box are now cheap and are a sensible hobby purchase — with the caveat that a machine that does three things adequately does none of them as well as a dedicated one, and the TIG on most of them is a compromise.
• What you buy alongside, and must — An auto-darkening helmet worth wearing, gauntlets, a natural-fibre or leather jacket, an angle grinder with cutting and grinding discs, clamps — many more than you think — a steel-topped bench, a wire brush, and a fire extinguisher. The grinder gets used more than the welder.
• What can wait — A chop saw, a bench vice worth having, magnets, a bandsaw, a plasma cutter. All genuinely useful and none of them the reason a first project fails.

• Consistent beads on flat plate — The first milestone and much less interesting than it sounds, because almost nothing real is welded flat and downhand.
• Joints, and fit-up — Butt, lap, tee and corner joints, and the realisation that preparing the joint decides more than welding it does. The step where most people’s work visibly improves.
• Position welding — Vertical, horizontal and overhead, where gravity stops helping. Genuinely harder and where hobbyists most often stall.
• Controlling distortion — Sequencing and tacking so the finished thing is the shape you drew. The skill that separates a competent frame from a good one.
• Thin material — Sheet and bodywork, where the fault is burning through rather than not fusing. A different discipline from thick-section work.
• A second process — Usually TIG after MIG, and usually because of aluminium or because of how the result looks. It feels like starting again, briefly.

Welding’s hazard list is unusual because the thing that frightens beginners is not the thing that harms welders. New welders worry about burns. What actually accumulates is fume, and what actually causes the memorable injuries is the arc and the fire.

The rest, in the order they arrive:

• The arc, and your eyes — A welding arc emits ultraviolet intense enough to burn a cornea, and the injury — arc eye — is delayed by hours, so the first you know is waking at two in the morning with what feels like sand under the lids. It is usually temporary and it is entirely preventable. The part beginners get wrong is that it does not require looking at the arc: reflected and glancing UV from across a workshop does it, which is why bystanders and anyone in the space need protection too, and why welding bays have screens.
• Skin, which is the same injury — The same UV sunburns exposed skin in minutes, including through a thin synthetic t-shirt. Covered arms and a proper jacket are not fussiness. Natural fibres or leather, because synthetics melt.
• Fire, and where the sparks actually go — Grinding and welding both throw sparks a surprising distance and they travel further than the arc suggests — under benches, into corners, into a rag you forgot. Most workshop fires from welding start after the welding stops, smouldering in something nobody checked. Clearing the space, having an extinguisher, and coming back after twenty minutes are the ordinary practice.
• Electric shock — Lower risk than the others and not zero, and the conditions that raise it are the ones a hobbyist creates: damp, sweat, a wet floor, welding while sitting on the work. Dry gloves, dry footwear, and not welding in the rain.
• Compressed gas — A cylinder is a stored-energy device. Chained upright, valve protected, moved with the regulator off. The failure mode of a knocked-over cylinder with a snapped valve is the one everybody has heard about and almost nobody has seen, which is exactly why the handling rules are treated as absolute.
• Containers that used to hold something — Never weld or grind a sealed container, a drum, or a tank that has held fuel, solvent or anything unknown. This is the classic fatal accident in amateur metalwork, the residue is the hazard rather than the contents, and cleaning it out is not a hobbyist procedure. If a job involves one, it is a job for somebody with the equipment to purge it.

• Fabrication — Gates, frames, brackets, furniture, trailers, workshop equipment. The mainstream of hobby welding and where most people stay, because the range is enormous.
• Restoration — Vehicles, machinery, agricultural equipment. Thin, rusty, awkwardly positioned metal — much harder than it looks and the reason patch-panel welding has its own following.
• Sculpture and art metalwork — A substantial tradition, with its own shows and a completely different relationship to tidiness. TIG and MIG both feature, and so does deliberate rust.
• Aluminium and stainless — Different metals, meaningfully different problems — aluminium’s oxide layer and heat conduction, stainless’s distortion and its more serious fume. Both are usually TIG territory and both are a step up rather than a variation.
• Blacksmithing and forge work — The adjacent hot-metal craft, working by heat and hammer rather than by arc. Shares a workshop culture and almost none of the technique.
• Coded and qualified welding — The professional end, where welds are made to a procedure and tested. Worth naming because it is the honest destination for anybody who wants to weld things that hold loads, and because the existence of that qualification is the reason this page declines that work.