BLD-WELD01-06 · Lesson 4 · 14 clock hours · practice
Multi-Position GTAW on Carbon Steel Plate (2G, 3G, 4G)
Objective. Students will complete groove and fillet welds on carbon steel plate coupons in horizontal, vertical, and overhead positions, meeting AWS D1.1 visual acceptance criteria and passing guided-bend tests.
Multi-Position GTAW on Carbon Steel Plate (2G, 3G, 4G)
Gas Tungsten Arc Welding (GTAW) demands precise hand coordination at all times, but that demand multiplies when you leave the flat position. Gravity acts differently on the weld puddle in every position, and your torch angle, travel speed, and amperage must compensate. This lesson covers the three test positions you will practice and be evaluated on: 2G (horizontal groove), 3G (vertical groove), and 4G (overhead groove).
Why Position Matters
In the flat (1G) position, gravity helps keep the molten puddle in the joint. As soon as you tilt the work or weld overhead, gravity becomes your opponent. The puddle wants to sag, drip, or undercut the top toe of the weld. Controlling heat input and travel speed is how you win that fight. Amperage that works perfectly flat will likely cause sagging or burn-through in overhead because the puddle is less stable without gravity supporting it.
2G — Horizontal Groove Weld
In the 2G position the plate is vertical and the groove runs horizontally. The puddle tends to sag downward toward the lower edge of the joint. Counter this by:
- Tilting the torch slightly upward (roughly 5–15° above horizontal, pointing into the joint) so arc force pushes the puddle back into the groove.
- Keeping travel speed slightly faster than flat so the puddle does not build up and roll over the lower bevel face.
- Using a slight whip or pause-and-advance technique to allow the puddle to freeze between filler additions.
Watch the lower toe constantly — undercut there is a common rejection cause on bend tests.
3G — Vertical Groove Weld (Uphill)
In the 3G position the plate is vertical and you travel upward. Uphill progression gives better fusion and is the standard for structural code work. The puddle tries to run down and away from the arc. Control it by:
- Pointing the torch downward into the puddle at roughly 5–15° below horizontal (slightly uphill-angled) so arc force holds the puddle in place.
- Using a controlled triangular or shelf-building weave: let a small shelf freeze, then advance; this ladders the bead up the joint without sagging.
- Reducing amperage 5–10% compared to flat to keep the puddle stiffer.
- Pausing at each toe to ensure fusion before moving back to center.
If the puddle starts to drip or look glassy and fluid, lift the arc slightly or increase travel speed momentarily.
4G — Overhead Groove Weld
Overhead is the most demanding position. The puddle has no gravity support at all and will fall if heat input is too high. Key adjustments:
- Reduce amperage further — often 10–15% below flat settings — and use the foot pedal to back off instantly if the puddle looks wet or starts to drop.
- Keep the arc length very short. A long arc in GTAW overhead means a wider, hotter, less controllable puddle.
- Use small filler rod additions frequently rather than large additions that overheat the puddle.
- Wear leather sleeves and a full-coverage welding jacket — spatter and dripping hot metal are real hazards here.
- Position yourself so your torch arm is supported against your body or the table edge; fatigue causes arc length drift.
Guided-Bend Testing
After welding, selected coupons are cut and bent using the school's guided-bend jig. Face-bend specimens are bent with the face (crown) in tension; root-bend specimens are bent with the root in tension. AWS D1.1 visual and bend acceptance criteria require that after bending, no single open defect exceed 1/8 in. (3 mm) in any direction, and the sum of all defects on any single face must not exceed 3/8 in. You will record every measurement in your weld log.
When a coupon fails, the failure is almost always traceable to one of these causes: lack of fusion at the root or toes, porosity from contamination or insufficient gas coverage, or incomplete penetration from moving too fast or carrying too long an arc.
Root-Cause Analysis
Every failed bend test is a learning event, not just a mistake. For each failure, write two sentences in your weld log: the first identifies the defect and its probable cause; the second states the technique correction. Example: "Root porosity appeared on the face-bend specimen, likely caused by a gas lens blockage that reduced argon coverage at the root. Corrective action: inspect and clean the gas lens before each session and verify flow rate at the regulator."
Safety Requirements for This Lesson
PPE: Auto-darkening helmet (shade 9–13 for GTAW on steel), leather welding gloves, leather jacket or sleeves (mandatory in overhead), flame-resistant pants, leather boots — no synthetic materials near the arc.
Argon cylinders: Secure cylinders to the cart or wall bracket at all times. Inspect the regulator for damage before opening the valve. Open the cylinder valve slowly and stand to the side.
Grinder safety: When using the angle grinder with a flap disc to clean coupons between passes, wear a face shield over your safety glasses, not a helmet. Keep the flap disc guard in place.
Hot metal: Coupons reach temperatures that cause serious burns long before they glow visibly. Use tongs or pliers to handle all coupons; mark hot pieces with chalk. Never lay a hot coupon on anything flammable.
Ventilation: GTAW on carbon steel with ER70S-2 still produces metal fume. Use local exhaust ventilation at each station and keep your head out of the fume plume.
Key terms10
2G positionA weld position in which the groove joint is oriented horizontally on a vertical plate; the weld axis runs horizontally and the welder travels side to side.
3G positionA weld position in which the groove joint is on a vertical plate and the welder travels upward; also called vertical-up welding.
4G positionA weld position in which the groove joint is on a horizontal plate positioned above the welder's head; the welder works from underneath.
Guided-bend testA mechanical test in which a weld coupon is bent around a mandrel of specified radius in a jig to reveal open defects such as lack of fusion, porosity, or cracks.
Face-bend specimenA bend-test coupon oriented so the weld face (crown) is on the tension side during bending, revealing surface and near-surface defects.
Root-bend specimenA bend-test coupon oriented so the weld root is on the tension side during bending, revealing incomplete penetration and root-area fusion defects.
Lack of fusion (LOF)A weld defect in which the filler metal fails to fuse completely with the base metal or a previous weld pass; a common cause of bend-test failure.
ER70S-2An AWS-classified GTAW filler rod for carbon and low-alloy steels; the 'ER' indicates electrode/rod, '70' indicates minimum tensile strength in ksi, 'S' indicates solid wire, and '2' identifies the chemical composition.
Arc lengthThe distance between the tungsten electrode tip and the workpiece surface; maintaining a consistent, short arc length is critical in GTAW to control heat input and puddle stability.
Weld logA student-maintained record of each coupon welded, including position, amperage, filler rod, bend-test results, and root-cause analysis entries; required documentation for advancement.
Worked examples4
1. Your face-bend specimen from a 2G coupon shows three open defects after bending. Defect A measures 1/16 in., Defect B measures 3/32 in., and Defect C measures 1/16 in. Does this coupon pass or fail AWS D1.1 acceptance criteria for bend tests?
- Step 1 — Check the single-defect limit: AWS D1.1 requires that no single open defect exceed 1/8 in. (0.125 in.) in any direction. Convert each measurement: 1/16 in. = 0.0625 in.; 3/32 in. = 0.09375 in.; 1/16 in. = 0.0625 in. All three are less than 0.125 in., so no single defect triggers automatic rejection.
- Step 2 — Check the total defect limit: AWS D1.1 also requires that the sum of all open defects on any single face not exceed 3/8 in. (0.375 in.). Add: 0.0625 + 0.09375 + 0.0625 = 0.21875 in.
- Step 3 — Compare the sum to the limit: 0.21875 in. is less than 0.375 in., so the total-defect criterion is also met.
Answer: The coupon passes both AWS D1.1 bend-test criteria: no single defect exceeds 1/8 in. and the total (approximately 7/32 in.) does not exceed 3/8 in.
2. A root-bend specimen shows two open defects: one measuring 1/8 in. exactly and one measuring 1/4 in. Does this coupon pass or fail?
- Step 1 — Check the single-defect limit: 1/8 in. = 0.125 in. This equals the maximum allowable size (not greater than), so it is at the borderline. However, the second defect measures 1/4 in. = 0.25 in., which exceeds 0.125 in. The coupon fails on the single-defect criterion for the 1/4-in. defect alone.
- Step 2 — For reference, check the total: 0.125 + 0.25 = 0.375 in., which equals but does not exceed 3/8 in. Even if the total were acceptable, the single defect of 1/4 in. already causes rejection.
- Step 3 — Record the failure mode in the weld log: the 1/4-in. defect exceeded the 1/8-in. single-defect limit.
Answer: The coupon fails because the 1/4-in. defect exceeds the AWS D1.1 single open-defect limit of 1/8 in.
3. You are welding in the 4G (overhead) position and your flat-position amperage setting is 120 A. Estimating a 12% reduction for overhead, what amperage should you try as a starting point?
- Step 1 — Calculate 12% of 120 A: 0.12 × 120 = 14.4 A.
- Step 2 — Subtract from the flat setting: 120 − 14.4 = 105.6 A.
- Step 3 — Round to a practical dial setting: 105.6 A rounds to approximately 106 A; in practice, set 105 or 106 A and fine-tune with the foot pedal.
Answer: A starting overhead amperage of approximately 105–106 A (a 12% reduction from 120 A flat).
4. After bending, you measure four defects on a face-bend specimen: 1/16 in., 1/16 in., 1/8 in., and 3/16 in. Evaluate the coupon against AWS D1.1.
- Step 1 — Check single-defect limit (must not exceed 1/8 in. = 0.125 in.): The 3/16-in. defect = 0.1875 in., which is greater than 0.125 in. This alone causes failure.
- Step 2 — For completeness, total = 0.0625 + 0.0625 + 0.125 + 0.1875 = 0.4375 in., which also exceeds the 3/8-in. (0.375-in.) total limit — a second failure criterion.
- Step 3 — Record both failure modes: single defect exceeds 1/8 in. and total defects exceed 3/8 in.
Answer: The coupon fails on two counts: the 3/16-in. defect exceeds the 1/8-in. single-defect limit, and the total of all defects (7/16 in.) exceeds the 3/8-in. cumulative limit.
The math in this lesson2
Learners reduce 120 A by 12% for overhead and sum bend-test defect lengths against acceptance limits.
7.RP.3N-Q.3
In class: Rotate through three welding stations (horizontal, vertical-up, overhead) completing one V-groove coupon per position per session, with instructor coaching on torch angle and travel speed adjustments for each position. · Perform guided-bend tests (face and root) on selected coupons using the school's guided-bend jig; students record open defects, measure against AWS D1.1 acceptance criteria, and enter results in their weld log. · Root-cause analysis writing exercise: for each failed bend test, students write a two-sentence cause-and-corrective-action entry in their weld log linking the defect type (e.g., lack of fusion, porosity) to a specific technique error.
Materials: GTAW inverter, foot pedal, argon gas, ER70S-2 filler rod, 3/8-in. and 1/2-in. carbon steel V-groove and T-joint coupons (pre-beveled), guided-bend test jig, weld gauge, angle grinder with flap disc, PPE full set
Check: One passing guided-bend test (face or root) per position with no open defects larger than 1/8 in. in any direction per AWS D1.1; all coupons and results documented in weld log — minimum two of three positions passing required to advance.
Practice check8
Answer each question, then check it. Nothing is recorded; this is your own practice.
Record this lesson on my transcript. Sign in with your key and enrollment email to record lessons (enroll free if you have no key). This page stays open to read and practise.