Four digits that read like a spec sheet
A stick electrode's AWS A5.1 number is a compressed data sheet, and learning to unpack it turns the rack at the welding store into a readable catalog. The first two digits after the E are minimum tensile strength of the deposited weld metal in ksi: an E60xx rod deposits at least 60,000 psi, an E70xx at least 70,000. The third digit is position — a 1 means all-position, usable flat, horizontal, vertical, and overhead. The last digit encodes the coating chemistry and the current the rod tolerates, which is where the personality lives: it separates the deep-digging cellulosics from the smooth rutiles and the low-hydrogen rods.
Four classifications cover most of what stick welders actually burn, and they anchor this guide: E6010 and E6011, the cellulosic penetrators; E6013, the smooth-running rutile; and E7018, the low-hydrogen workhorse that carries the 70 ksi deposit. The running example threads through all of it — a 1/8-in E7018 going onto 1/4-in plate — from reading the box to striking an arc at a number both major manufacturers endorse.
The machine decides first: polarity before preference
Before taste enters the picture, the welding machine narrows the menu. E6010's sodium-based cellulosic coating demands the stable arc of DCEP — electrode positive — and will not run properly on alternating current, so on an AC-only buzz box it is simply off the list. E6011 exists precisely for that machine: its potassium-based coating keeps the arc lit through the AC zero crossings while keeping the same deep-penetrating character. E6013 is the least demanding of all, happy on AC, DCEP, or DCEN, and E7018 runs on DCEP or AC. Match the rod to the current the machine can produce first; every other consideration comes second.
Why polarity changes arc behavior at all — what electrode-positive does differently from electrode-negative, and what AC's constant reversals buy — is its own subject, argued properly in the AC-versus-DC welding comparison. Likewise, whether stick is even the right process next to wire-feed and TIG alternatives is the territory of the process comparison. This guide takes the stick machine as given and picks the rod.
Cellulosic, rutile, low-hydrogen: what the coating buys
The last digit's coating families behave like three different tools. The cellulosic pair, E6010 and E6011, drive a forceful arc that punches through rust, paint, and galvanizing better than any other class — the reason they open root passes and rescue dirty repair work. The rutile E6013 trades that aggression for a soft, forgiving arc and an easy puddle, which is why it shows up in training booths and on thin, cosmetic work. Each is the right answer to a different question, and the wrong answer to the other's.
E7018 is the structural default for a different reason: it is a low-hydrogen rod. Its coating is designed to keep hydrogen out of the weld deposit while delivering the 70,000 psi minimum tensile the first two digits promise, and the combination — strength plus clean deposit plus all-position capability — is what makes it the rod specified on structural work. For the 1/4-in plate in the running example, with strength mattering and the steel reasonably clean, the box that comes off the shelf says E7018.
Amperage is not in the standard
Here is the fact that surprises people holding two different amperage charts: AWS A5.1 does not specify welding current. The classification fixes the weld metal's properties and the coating family — the recipe is the manufacturer's, and each maker's flux chemistry, iron powder content, and arc drive differ enough that their published current windows genuinely disagree. Lincoln rates its 5/32-in E7018 at 130–210 A while Hobart's equivalent spans 110–230 A; the same class, different rods. A range printed on a chart is only guaranteed for the specific product it was published for.
The practical defense is consensus: where two manufacturers' published ranges overlap, the window is safe on either brand. For the 1/8-in E7018 of the running example, Lincoln publishes 90–160 A and Hobart 90–150 A, so the overlap is 90–150 A on DCEP. The divergence can be dramatic enough to matter — one Hobart E6013 variant publishes windows like 130–160 A at 5/32-in, far below other makers' figures, because it is tuned as a low-amperage sheet-metal rod. Reading the data sheet for the rod actually in your hand beats any generic chart taped to the machine.
Diameter follows thickness, amps follow diameter
Selecting the rod size is a two-step chain: the plate picks the diameter, and the diameter picks the current range. Common stick diameters run 3/32, 1/8, 5/32, 3/16, and 1/4 in (2.4 to 6.4 mm), and the old shop rule is not to weld metal thinner than the rod diameter — a 1/8-in rod belongs on 1/8-in and thicker steel. Once the diameter is set, the manufacturer's window for that diameter applies, and a rough heuristic lands you near its middle: about one amp per thousandth of an inch of diameter, which puts a 1/8-in (0.125 in) rod near 125 A.
The chain also runs in reverse when things go wrong. If the plate stops absorbing what the rod delivers — the puddle wants to collapse through thin material, or a thick joint refuses to fuse at the top of the range — the fix is a different diameter, not a heroic amperage setting. Step down a size for thin work rather than starving a fat rod, and step up a size rather than overdriving a thin one; position matters too, with both major manufacturers publishing the same adjustment of dropping roughly 15% off the flat-position setting for vertical and overhead work.
Heat input: the number a WPS actually limits
On procedure-controlled work, the amperage knob is only one of three numbers that get audited. Welding procedure specifications limit heat input — the arc energy delivered per length of joint — computed as HI = 60 × V × I ÷ (1000 × S), with V the arc voltage, I the current, and S the travel speed in inches (or millimetres) per minute. AWS D1.1 writes its limits against this gross figure; the European EN 1011 convention multiplies by a process efficiency factor first, 0.8 for stick welding, since an open arc loses some energy to the surroundings. The limits exist because heat input steers the metallurgy: too high slows cooling, coarsens the heat-affected zone, and can soften quenched-and-tempered steels; too low speeds cooling and raises hardness and hydrogen-cracking concerns.
What this means for rod selection is that amperage, voltage, and travel speed form one budget, and cranking current while slowing down can bust a WPS limit even inside the manufacturer's amperage window. The welding heat input calculator does this bookkeeping — gross kJ/in and kJ/mm from volts, amps, and measured travel speed, the EN 1011-1 efficiency factors when a specification wants net figures, and a pass/fail check against the WPS maximum.
One 1/8-in rod on 1/4-in plate, dialed in
Assemble the running example end to end. The job: flat-position welds on 1/4-in mild steel plate, structural enough to want the 70 ksi deposit and a low-hydrogen rod, on a machine that produces DC. The number decodes the choice: E7018 — 70,000 psi minimum tensile, all-position, low-hydrogen coating, DCEP or AC. The plate is twice the thickness of a 1/8-in rod, so 1/8-in is a comfortable diameter. The consensus window for 1/8-in E7018 is 90–150 A on DCEP, and both the middle-of-range logic and the amp-per-thousandth heuristic point to roughly 120–125 A for flat work on plate this thick.
Strike the arc at 125 A and let the rod report back. Dragging and sticking says the setting is cold — for this rod that behavior lives down near and below 100 A — while a violently fluid puddle with the coating glowing and burning back says hot, so back off. If the pass must satisfy a WPS, close the loop with measured numbers: an arc reading, say, 23 V while covering 6 in of joint per minute gives HI = 60 × 23 × 125 ÷ (1000 × 6) = 28.8 kJ/in (about 1.13 kJ/mm) gross — the figure to compare against the procedure's ceiling before the inspector does.
The rod choice in one pass
Recap the selection chain on the example job. Decode: E7018 = 70 ksi deposit, all-position, low-hydrogen, DCEP or AC — compatible with the machine, suited to structural plate. Size: 1/4-in steel comfortably carries a 1/8-in rod, never the reverse. Current: the manufacturer, not AWS, owns the amperage — the Lincoln-Hobart consensus for this rod and diameter is 90–150 A DCEP, entered near 125 A and tuned by arc behavior, minus about 15% if the joint turns vertical or overhead. Budget: volts, amps, and travel speed together set heat input, the number the WPS actually polices. Four digits, one diameter rule, one published window, one formula — that is the entire decision, and every number in it came off a data sheet rather than a guess.