Industry
Insights

Factor one

Where is the crack or damage?

Location matters more than the size of the damage. A crack on a flat, non-critical surface is often straightforward to weld. A crack running through a threaded bore, a bearing seat, or a sealing face needs more thought — it can still be welded and machined back, but the repair is more involved. Damage in a structural stress zone needs a welder who understands load paths, not just someone who can run a bead.

Factor two

What is the part made of?

Most steel and iron parts can be welded. Cast iron is trickier — it's brittle and needs careful preheat and the right filler rod — but it's very much weldable in skilled hands. Hardened steel, stainless, and some aluminum alloys each have their own rules. The material dictates the process, not the other way around.

Factor three

How long does a new part take to arrive?

For industrial equipment, lead times on replacement parts can run 6 to 20 weeks — sometimes longer for imported or obsolete components. A weld repair done in-house can often be completed in days. If your operation can't afford weeks of downtime, repair is almost always the faster path, even when replacement might seem simpler on paper.

Factor four

What does it cost to replace vs. repair?

For a $200 part, replacement is usually the right call. For a $15,000 pump housing or a custom-machined component that no longer exists in a catalogue, weld and machine repair is almost always worth it. The math changes dramatically once you factor in downtime, shipping, and reinstallation costs on both sides.

Process one

MIG welding — fast, versatile, high volume

MIG stands for Metal Inert Gas. A wire electrode feeds automatically from a spool through the welding gun, and the arc melts both the wire and the base metal together. A shielding gas (usually a mix of argon and CO₂) protects the weld pool from contamination in the air. MIG is fast, relatively easy to learn, and works well on most mild steel and stainless. It's the go-to for fabrication work where speed and production volume matter.

Process two

TIG welding — precise, clean, high quality

TIG stands for Tungsten Inert Gas. Instead of a consumable wire electrode, TIG uses a non-melting tungsten electrode to create the arc. The welder feeds filler rod into the puddle by hand with the other hand — it requires both hands and a foot pedal for heat control simultaneously. TIG is slower than MIG but produces cleaner, more precise welds with less spatter. It's the choice for stainless steel, aluminum, thin materials, and any job where weld appearance and quality are critical.

Process three

Stick welding — tough, portable, works anywhere

Stick welding (officially called SMAW — Shielded Metal Arc Welding) uses a coated electrode rod that melts into the weld as it burns. The coating on the rod burns off as flux, creating its own shielding gas and slag layer. No external gas bottle needed. Stick welding is slower and less clean than MIG or TIG, but it's extremely versatile — it works outdoors in wind, on rusty or dirty metal, and in tight spots where other processes can't reach. It's also the most common process for heavy structural work and pressure vessel welding.

Step one

Engineering drawing or sketch

Everything starts with a drawing — either a full CAD drawing from the client, a hand sketch, or even a broken part we're reverse-engineering. This tells us dimensions, material, tolerances, and what the part needs to do. The better the drawing, the faster and more accurate the result. Rumix works directly with clients to develop drawings when they don't have them.

Step two

Material selection and procurement

The right steel for the job isn't always the cheapest or the easiest to find. Mild steel, stainless, structural steel, tool steel — each has different strength, corrosion resistance, and weldability. Rumix selects material based on what the part needs to survive: temperature, pressure, chemical exposure, and mechanical load all play a role.

Step three

Cutting and forming

Raw steel stock is cut to size — by saw, plasma torch, or other cutting methods depending on thickness and precision required — then formed into shape. Forming includes bending, rolling, and pressing. This is where flat steel plate starts becoming a recognizable shape.

Step four

Welding and assembly

Cut and formed pieces are welded together. Fit-up — how well the pieces align before welding — is critical. A poor fit-up means more filler metal, more distortion, and a weaker joint. Good fabricators tack-weld first to hold position, check alignment, then complete the welds in a sequence that minimizes distortion from heat.

Step five

Machining and finishing

Welded fabrications rarely come out at exact final dimensions — there's always some distortion. Critical surfaces go to the machine shop for turning, milling, or grinding to hit final tolerances. Then finishing: grinding welds flush, deburring edges, applying primer or paint if needed. The part is inspected against the original drawing before it ships.

Reason one

Defects hide inside the weld

The surface of a weld can look perfect — smooth, consistent, no visible cracks — while hiding porosity (gas bubbles), lack of fusion, or slag inclusions trapped inside. These internal defects weaken the joint significantly. They don't show up until the part is under load, pressure, or vibration. A rushed welder skips the preparation and interpass cleaning that prevents these defects from forming.

Reason two

Distortion from rushing costs time in machining

When a welder goes too fast and skips proper sequence or interpass temperature control, the heat input becomes uneven. The part warps. What should have been a flat surface ends up bowed. What should have been a round bore ends up oval. Correcting this in the machine shop takes far longer than doing the welding carefully in the first place.

Reason three

Codes and standards exist for a reason

Pressure vessels, structural steel, and piping systems are all welded to codes — ASME, CSA, AWS — that define minimum quality standards. These aren't suggestions. They exist because weld failures in those applications put people at risk. A certified welder working to code spends more time on each pass, but the result is a weld that can be tested, certified, and trusted.

Reason four

A quality weld lasts — a fast weld doesn't

Industrial parts live hard lives. Pressure, vibration, thermal cycling, corrosive fluids. A quality weld is fully fused, properly stress-relieved, and made with the right filler metal for the application. It can outlast the surrounding base metal. A rushed weld develops fatigue cracks at the toes, peels at the fusion line, or fails outright when conditions get tough.

Cause one

Hot cracking — happens right after welding

Hot cracks form while the weld metal is still cooling from liquid to solid. They happen when the weld pool solidifies with too much tension — either because the joint was restrained, the filler metal composition was wrong, or the cooling was too fast. They're usually visible along the center of the weld bead.

Cause two

Cold cracking — appears hours or days later

Cold cracks (also called hydrogen-induced cracking) are sneaky. The weld looks perfect when it cools, then a crack appears 24–72 hours later. The culprit is hydrogen — absorbed into the weld metal from moisture in the electrode coating, the base metal surface, or the air. Hydrogen diffuses through the metal as it cools and causes cracking in the heat-affected zone.

Cause three

Underbead cracking — hidden under the surface

Underbead cracks form just below the weld in the heat-affected zone — the area of base metal that got hot but didn't melt. This zone becomes harder and more brittle than the surrounding metal, and if there's hydrogen present and residual stress, it cracks. This type is particularly dangerous because it's invisible without NDT testing.

Cause four

Fatigue cracking — develops over time in service

Fatigue cracks aren't caused by the welding process itself — they develop in service, usually at the toe of a weld where the weld meets the base metal. Every cycle of loading and unloading is a tiny stress event. Over thousands or millions of cycles, that stress concentrates at the weld toe and a crack initiates. Poor weld profile makes this worse.

Step one

Preheat the metal

Before striking a single arc, we heat the entire pump body — or at least the area around the repair — using a torch or an oven. This brings the metal to a specific temperature called the preheat temperature, which depends on the type of metal the pump is made from.

Why? Because cold metal next to a hot weld is like pouring boiling water into a cold glass — it can crack. Preheating slows the cooling rate so the weld and the base metal shrink together at the same pace. For cast iron pumps, we typically preheat to 300–600°F (150–315°C). For carbon steel, 200–400°F is the usual range.

Step two

Weld the repair

Now we lay down the weld — filling the crack, building up worn surfaces, or joining a broken section back together. Our welders use short passes called stringer beads and select filler rods matched exactly to the pump's metal. For cast iron, nickel-based electrodes are the go-to because pure nickel is stretchy enough to handle the stress without cracking.

The pump stays warm the entire time. Letting it cool mid-weld is a classic mistake. On large pump housings, we wrap the part in insulating blankets between passes to hold the heat in.

Step three

Stress relieve (or slow-cool)

After welding, the metal is full of internal tension — invisible forces locked inside from the heating and cooling. Left alone, those stresses can cause warping, cracking later, or premature failure under pressure. The fix is post-weld heat treatment (PWHT), also called stress relieving.

The pump goes back into the oven — or we use ceramic insulating pads with electric heating elements — brought up to around 1,100–1,250°F (595–675°C) for carbon steel, held there for an hour or so, then cooled very slowly. On simpler jobs, wrapping the pump in insulating blankets overnight can be enough.

Step four

Machine it back to spec

Welding adds metal and distorts dimensions — the pump is no longer the right size or shape. This is where Rumix Machining and Fabrication earns its name. The pump goes to our lathe, mill, or boring machine and the weld is cut, ground, or bored back to the original engineering drawings. Bearing seats, seal faces, impeller fits, and flange surfaces are all brought back to exact measurements.

This final step turns a rough weld repair into a pump that works as well as — or better than — the original. Machining removes any surface porosity and leaves a clean, precise surface for seals and bearings to run on.