Reading welds like topographic maps of the Black Hills — a field guide by Bernardo Zubko, Pine Ridge Reservation
Fig. 1 — The arc is the only sun that matters in the puddle. TIG on 304L stainless, 120A DCEN. Photo: Pexels.
I grew up reading the hills around Pine Ridge. Before there were contour lines on any map, there was the way the light fell at dusk — darker in the coulees, sharp on the ridge. My grandfather pointed with his walking stick and said that slope will hold water, that one won't. He was reading temperature, rock density, and drainage in the same way I read a weld bead on the bench.
Every weld is a landscape. The fusion zone is the fault line. The heat-affected zone is the weathered plateau. The base metal beyond is the untouched prairie. If you can read the terrain of a weld — the bead width, the cap profile, the color of the HAZ — you know whether it will hold through a South Dakota winter or split like green wood.
This is not a spec sheet. It is a field guide. The kind of thing I wrote in a spiral notebook and kept in my helmet case for twelve years before I decided to put it on the web.
The fusion zone is the narrowest part of the weld — the actual line where base metal and filler became one substance. On A36 steel, it's about 0.2 to 0.6 millimeters wide per side, depending on travel speed and heat input. It looks like a fine root growing into both plates.
| Parameter | Typical Value | What It Means |
|---|---|---|
| Penetration depth | 3–6 mm | How deep the molten pool ate into the base. Shallow = lap weld. Deep = burnthrough. |
| Fusion width (each side) | 0.2–0.6 mm | The actual bonded interface. If this is inconsistent, the joint is weak. |
| Peak temperature | ~1510°C | Melting point of low-carbon steel. Above this, the metal is liquid. |
| Wettable angle | 20°–45° | Where the fusion line meets the base. Too acute (<20°) means lack of fusion. Too blunt (>45°) means excessive heat input. |
I learned to read the wettable angle by breaking my own bad welds. My first year, I ran a root pass on a ¼-inch plate and it looked smooth on top. Broke it open and the fusion line was a hair's breadth from the surface — a lipstick stain on a brick. The filler sat on top without eating in. Too much travel speed, not enough voltage. The puddle was a lazy pool, not a hungry one.
Just beyond the fusion zone, the metal never melts but it remembers the heat. This is the HAZ — heat-affected zone — and it's where most weld failures actually start. The microstructure changes without the metal ever turning liquid. Grain structure grows, hardness shifts, residual stress locks in.
In carbon steel, the HAZ has three sub-zones I watch for:
| Zone | Peak Temp | Grain Size | Mechanical Effect |
|---|---|---|---|
| Coarse-grained HAZ | 1100–1510°C | Large, equiaxed | Highest hardness. Most brittle. Cracks start here. |
| Fine-grained HAZ | 900–1100°C | Small, uniform | Tougher. Often stronger than base metal. The "good" HAZ. |
| Subcritical HAZ | 600–900°C | Variable | Partial annealing. Softens hardened steel. Can weaken it. |
The coarse-grained zone is the enemy. On my first pipeline job out near Rapid City, we had a series of cracks that traced the HAZ of every other root pass. The root cause? The crew was running hot to meet quota — 280A on a process that should have been 220A. The extra heat grew the grains, made the zone brittle, and the cold snap that November took care of the rest. Cracks ran the length of the seam like a zipper.
I measure the HAZ visually by the discoloration it leaves on stainless steel. Blue means 600–650°C. Straw means 800–850°C. Purple means you're in the coarse-grained range already. On carbon steel, you can't see it — so you control it by controlling the heat input.
When I look at a finished weld, I'm reading it like a topographic map. The bead profile — the height, width, and cap shape — is the elevation data. From the profile, I can reconstruct the thermal history of the joint.
| Profile Shape | Thermal Cause | Diagnosis |
|---|---|---|
| Wide, flat, low reinforcement | High heat input, slow travel | Good penetration but watch for excessive HAZ growth |
| Narrow, tall, peaked cap | Low heat input, fast travel | Watch for lack of fusion at the toe |
| Uniform width, smooth ripples | Stable arc, consistent speed | Clean bead. No action needed. |
| Irregular ripples, width varies | Unstable travel speed | Operator inconsistency. Re-run if structural. |
| Convex, bulging cap with undercut at toe | Excessive current, improper angle | Undercut is a stress riser. Redo. |
| Concave cap (scooped) | Too fast or weave too wide | Reduced throat dimension. Strength compromised. |
The ripple spacing on a GMAW bead is a metronome of your travel speed. At 250A on ER70S-6, I expect ripples every 3–4 millimeters. If they're 6 mm apart, I'm too slow and the heat is building up. If they're 1.5 mm apart, I'm rushing and the puddle won't have time to form properly. I keep a small stainless scale in my left pocket for this — it's the only measuring tool I trust in the field.
On carbon steel, the HAZ is invisible to the eye. On stainless, the oxide layer changes color with temperature, and the colors form a natural heat map right on the weld:
| Color | Approx. Peak Temp | Condition |
|---|---|---|
| Bright silver (no color) | <400°C | Excellent. No sensitization risk. |
| Straw / light yellow | 600–650°C | Acceptable. Minimal chromium carbide formation. |
| Blue | 650–750°C | Watch. Some sensitization possible. Pickling recommended. |
| Purple | 750–850°C | Concerning. Significant sensitization. Clean or re-run. |
| Gray / rainbow to black | >900°C | Failure zone. Chromium depleted at grain boundaries. Corrosion will find this. |
I once cut open a food-grade tank that looked perfect from the outside — silver bead, clean ripples. Cut it open and the root pass was purple-black on the inside. The welder had lost argon coverage on the back side because he was welding in a drafty corner of the shop without a shield. The tank held pressure but it was already dying. Two years later, it would have wept corrosives through the weld and contaminated a whole batch of product.
Welding on the prairie adds a layer most shop welders never face. Wind doesn't just cool the bead — it strips shielding gas from the puddle. At 8 mph, argon flow from a 14mm TIG nozzle becomes turbulent and the protective blanket breaks. I've written a calculator for this (Wind Factor Calculator), but the visual signs are what matter in the field:
My remedy is always the same: build a baffle. Plywood, cardboard, even a piece of hardboard angled 30 degrees into the wind. It's not pretty, but it drops the local wind speed enough to save the weld. I've carried a 2×3 foot piece of ⅛-inch plywood on my truck since 2016. It's held more jobs than any certificate.
There are four conditions on the prairie that make welding a waste of time and a danger to the joint. I call them the Four Killers. If any one is present, you pack up:
| Killer | Threshold | Why |
|---|---|---|
| Wind | >10 mph (unshielded) | Shielding gas displaced. Nitrogen porosity follows. |
| Humidity on base metal | Visible moisture or condensation | Hydrogen pickup. Delayed cracking in the HAZ, sometimes 24h later. |
| Temperature | Base metal <-18°C (0°F) | Too fast solidification. Hydrogen has no time to escape. Cracking. |
| Rain / snow | Any precipitation on the weld zone | Immediate quench. Hydrogen cracking. Do not argue with rain. |
People will tell you to push through. They'll say "it'll hold" or "I've done it worse." I've seen those welds fail. The hydro testing always reveals what pride hides. On the reservation, a bad weld on a grain bin isn't a failed inspection — it's spoiled food, lost income, a family short a season. We weld right because it matters.
Thermal data from ASM Handbook Vol. 1 (Q131172). Color scale adapted from AWS A5.4/A5.4M. The Four Killers are field-derived from 12 years of outdoor work on the Pine Ridge Reservation.
Related work in town: