MFG-SHIP13-03 · Lesson 8 · 14 clock hours · theory
Weld Distortion Control and Thermal Management in Marine Structures
Objective. Students will predict distortion types (angular, longitudinal, transverse, buckling) produced by welding sequences on a given joint configuration and select proven distortion-control methods — presetting, backstep, skip welding, and clamping — appropriate for a shipyard production scenario.
Why Distortion Happens
Welding deposits heat into a very small zone. The metal directly under the arc expands rapidly, but the surrounding cool metal resists that expansion, so the heated zone is forced to compress (upset) plastically. When the weld cools and tries to contract, it cannot return to its original size because the plastic compression already happened. The result is a net shrinkage force that pulls the base metal out of shape. Every weld produces shrinkage; the welder's job is to plan for it and counteract it.
Shrinkage acts in three directions at once: transverse (across the weld), longitudinal (along the weld), and angular (rotation around the weld line). Understanding which direction dominates for a given joint type is the first step in control.
The Four Main Distortion Types
Transverse shrinkage is the shortening that occurs perpendicular to the weld centerline. A butt weld in a ship hull plate will pull the two plate edges toward each other as it cools.
Longitudinal shrinkage is the shortening along the weld length. On a keel bar, longitudinal shrinkage can bow the bar like a banana. It is smaller in magnitude than transverse shrinkage but significant over long runs.
Angular distortion occurs in fillet and partial-penetration welds — the classic example is a T-joint where the flange tips upward (or downward) because the weld root and face cool at different rates. On deck plates, a row of stiffener fillet welds can leave the plate looking corrugated.
Buckling shows up in thin plates (typically under about 3/16 inch in marine shell plating) when compressive residual stress exceeds the plate's buckling threshold. The panel snaps into a wave pattern. Unlike the other distortion types, buckling is sudden and difficult to correct after the fact.
Distortion-Control Methods
Presetting (Precambering)
The joint is clamped or bent in the opposite direction of the expected distortion before welding begins. When the weld shrinks, it pulls the joint back toward flat. Presetting angle must be estimated from experience or prior test welds on the same material and joint geometry. Shipyards often use preset fixtures called strongbacks — heavy backing bars or frames clamped or tack-welded to the panel — to hold preset position.
Backstep Welding
Instead of running a weld from left to right in one continuous pass, the welder divides the joint into short segments (typically 4–8 inches) and welds each segment in the direction opposite to overall progression. For example, the overall joint is worked left to right, but each individual segment is deposited right to left. This reduces the heat buildup ahead of the arc and distributes residual stresses more evenly, cutting longitudinal shrinkage and angular distortion.
Skip (Intermittent) Welding
Weld segments are deposited at spaced intervals along the joint, then the gaps are filled in later. Skipping spreads heat input over time and limits the shrinkage force that builds up in any one location. Intermittent fillet welds are also specified by design when full-length welds are not needed for strength, which inherently limits distortion.
Clamping and Fixturing
Rigid clamping — to a flat table, to strongbacks, or to previously erected structure — mechanically resists shrinkage movement. Clamping is simple and effective but must be planned carefully: releasing clamps before the weld has fully cooled can allow spring-back; releasing them too late on a highly restrained joint can introduce high residual tensile stress, which affects fatigue life. In shipyard practice, clamping is almost always combined with one of the sequencing methods above.
Weld Sequence and Symmetry
On a ship frame with multiple welds, depositing welds symmetrically about the neutral axis allows opposing shrinkage forces to cancel. Weld from the center of a panel outward rather than from one end, and alternate sides of a double-sided joint, so each pass partially corrects the distortion of the last.
Heat Input and Material Considerations
Higher heat input means a larger heat-affected zone and more shrinkage. Using the lowest current and travel speed that still gives a sound weld reduces distortion. Aluminum is particularly sensitive: it has roughly twice the thermal expansion coefficient of steel and about one-third the modulus of elasticity, so it distorts more easily and requires even more careful heat management and fixturing.
Safety During This Work
Strongbacks and clamps store significant mechanical energy. Inspect all clamps and tack welds before releasing — a strongback that lets go suddenly can become a projectile. Wear full PPE at all times: welding helmet with appropriate lens shade, leather gloves, flame-resistant clothing, and safety glasses under the helmet. Ensure adequate ventilation because correction of distorted panels often involves additional grinding and welding in confined spaces. Follow lockout/tagout procedures whenever repositioning workpieces on powered positioners or roller beds. Never use your hands to test weld temperature — use a contact pyrometer or temperature-indicating sticks.
Applying This in the Shipyard
When you receive a fabrication drawing, read the weld sequence note if one is provided. If none exists, build your plan by identifying: (1) which welds are most likely to cause angular distortion, (2) where thin plate is at risk of buckling, (3) how you can weld symmetrically about the structure's neutral axis. Write the sequence down, number the welds, and check it against the drawing before striking an arc. A few minutes of planning saves hours of flame-straightening and re-inspection later.
Key terms10
Transverse shrinkageDimensional shortening that occurs perpendicular (across) the weld centerline as the weld metal and heat-affected zone cool and contract.
Longitudinal shrinkageDimensional shortening that occurs parallel to (along the length of) the weld, capable of bowing long members such as keel bars.
Angular distortionRotation of base-metal plates or flanges around the weld line, caused by unequal contraction through the thickness of the joint; common in fillet-welded T-joints.
Buckling distortionA wave or snap-through deformation of thin plate panels that occurs when welding-induced compressive residual stress exceeds the plate's buckling resistance.
Presetting (precambering)Intentionally deflecting or clamping a joint in the direction opposite to expected distortion before welding, so that weld shrinkage returns the part toward the desired geometry.
Backstep weldingA weld sequence technique in which individual short segments are deposited in the direction opposite to the overall weld progression, reducing heat buildup and distributing residual stress.
StrongbackA rigid backing bar, beam, or frame clamped or tack-welded to a weldment to mechanically restrain distortion during and after welding.
Heat-affected zone (HAZ)The region of base metal adjacent to the weld fusion line that is not melted but whose microstructure and properties are altered by welding heat; the size of the HAZ grows with heat input.
Residual stressStress remaining in a weldment after welding and cooling are complete, caused by uneven thermal expansion and contraction; can be tensile or compressive depending on location.
Intermittent (skip) weldingA distortion-control technique in which weld segments are deposited at intervals along a joint, spreading heat input over time and reducing cumulative shrinkage force.
Worked examples4
1. A T-joint coupon is 12 inches long. After a single continuous fillet weld, the flange measures 3° of angular distortion. After a backstep sequence (four 3-inch segments, welded right-to-left within each segment), the angular distortion measures 1.2°. What is the percentage reduction in angular distortion achieved by the backstep sequence?
- Step 1 — Identify the two measured values: continuous weld distortion = 3.0°; backstep weld distortion = 1.2°.
- Step 2 — Calculate the reduction in distortion: 3.0° − 1.2° = 1.8°.
- Step 3 — Divide the reduction by the original (baseline) value and multiply by 100 to get a percentage: (1.8 ÷ 3.0) × 100 = 60%.
Answer: The backstep sequence reduced angular distortion by 60%, from 3.0° to 1.2°.
2. A welder must deposit ten fillet welds on a ship frame. She decides to skip-weld by laying every other weld first, then returning to fill the gaps. If each weld takes 4 minutes to deposit and she waits 3 minutes between passes to allow heat to dissipate, how long will the complete sequence take (welding time plus wait time only)?
- Step 1 — Count the welds: 10 total. In skip-welding, she deposits 5 welds in the first pass (odd-numbered positions) and 5 in the second pass (even-numbered positions).
- Step 2 — Calculate welding time: 10 welds × 4 minutes each = 40 minutes total welding time.
- Step 3 — Count the wait periods: she waits once between the first set of 5 welds and the second set of 5 welds = 1 wait period of 3 minutes. (Waits between individual welds within a set are not specified in this problem, so count only the between-pass wait.)
- Step 4 — Add welding time and wait time: 40 + 3 = 43 minutes.
Answer: The complete sequence takes 43 minutes (40 minutes welding + 3 minutes waiting between passes).
3. A fabricator presets a deck plate assembly 4 mm upward at mid-span before welding the stiffeners. After welding and cooling, the panel springs back 3.5 mm toward flat. What residual (remaining) distortion does the panel have, and in which direction?
- Step 1 — The preset introduces a 4 mm upward bow (opposite to expected distortion).
- Step 2 — Weld shrinkage pulls the panel downward (in the direction of expected distortion) by 3.5 mm.
- Step 3 — Net position = preset amount − spring-back = 4 mm upward − 3.5 mm downward = 0.5 mm upward.
- Step 4 — The panel is still 0.5 mm proud (above flat) at mid-span, which is well within most shipyard flatness tolerances.
Answer: The panel retains 0.5 mm of upward bow — a near-flat result, confirming that the 4 mm preset was a good estimate for this joint.
4. A continuous butt weld 24 inches long is expected to produce 0.08 inches of transverse shrinkage. The welder switches to a backstep sequence that historically reduces transverse shrinkage by 35% for this joint type. What transverse shrinkage should she expect with the backstep sequence?
- Step 1 — Identify the baseline shrinkage: 0.08 inches.
- Step 2 — Calculate the reduction amount: 0.08 × 0.35 = 0.028 inches.
- Step 3 — Subtract the reduction from the baseline: 0.08 − 0.028 = 0.052 inches.
Answer: With the backstep sequence, expected transverse shrinkage is approximately 0.052 inches.
The math in this lesson2
Learners compute net camber after a 4 mm preset springs back 3.5 mm.
7.NS.17.RP.3
In class: Case-study analysis: students examine three photographs of distorted ship panels (one angular distortion on a deck plate, one buckling on a thin shell plate, one longitudinal shrinkage on a keel bar), identify the distortion type, and write a one-paragraph root-cause explanation using thermal contraction principles. · Welding sequence lab: in pairs, students weld an identical 12-inch T-joint twice — once with a single continuous weld and once using a backstep sequence — measure angular distortion of each result with a digital angle gauge, and record the quantitative difference in their lab log. · Group problem-solving: teams receive a simplified ship frame drawing with ten welds and must write a weld sequence (numbered weld order, direction arrows) that minimizes distortion, then present their sequence rationale to the class for critique and instructor scoring.
Materials: Digital angle gauges, 12-inch carbon steel T-joint coupons (pre-tacked), SMAW or FCAW machines, distortion case-study photograph set (printed), ship frame drawing with weld callouts, straightedges, clamps and strongbacks, Lincoln Electric Welding Handbook or equivalent reference chapter on distortion, lab log notebooks
Check: Students submit a lab log entry with measured angular distortion values from both weld sequences and a written comparison; they also score 70% or higher on a 10-question written distortion-control quiz administered at the end of the lesson.
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.