8.3 Butt Weld Pipe Fabrication: End Preparation, Alignment, Fit-Up & Cut Lengths

Butt Weld Pipe Fabrication: End Preparation, Alignment, Fit-Up & Cut Lengths

Core Trade Concept: Butt welding is how every load-bearing, radiographable, code pressure joint in a boiler plant is made. Module 34208 is the largest single module in NCCER's Boilermaking curriculum at 40 hours, and it is the largest slice of the Pipe Fabrication content domain. The whole discipline reduces to three things done accurately every time: prepare the end correctly, align the bores, and hold the root opening the procedure calls for.

---

1. End Preparation to ASME B16.25

                        STANDARD BUTT WELD END PREP

                     37.5 deg
                        \  |
                         \ |
                          \|
      ====================  |         <-- pipe wall
      ====================  |
                          /|
                         / |  ROOT FACE (LAND)
                        /  |  1/16 in. +/- 1/32 in.
                     ROOT OPENING (set at fit-up, per WPS,
                     typically 3/32 to 1/8 in.)

• Bevel angle: 37.5° from the pipe axis is the ASME B16.25 standard end preparation for butt-welding components. Two ends together produce the familiar 75° included angle single-V groove.
• Root face (land): approximately 1/16 in., with a tolerance of ±1/32 in. The land's job is to give the root pass something to fuse without burning through (Section 2.2).
• Root opening is not part of the bevel — it is set at fit-up and comes from the WPS, commonly 3/32 in. to 1/8 in. for open-root work.

Preparation methods, in ascending order of precision: oxyfuel bevel with a torch guide, then grinding, then a pneumatic or hydraulic bevelling machine, then a portable clamshell (split-frame) machining lathe for alloy and heavy wall (all developed in Section 2.2). Alloy piping such as P91 is machined, not flame-cut, because flame cutting alters the metallurgy of the prepared surface.

---

2. Alignment: The Hi-Lo Rule

Internal misalignment — hi-lo — is the offset between the two bores at the joint. It matters because a step in the bore is a stress riser, a turbulence source and a radiographic indication.

• ASME B31.1 §127.3 limits internal misalignment to 1/16 in. unless the design specification says otherwise.
• When misalignment exceeds the allowance, the preferred correction is to internally trim the component whose wall extends inward, provided the remaining thickness is not reduced below the minimum design thickness, and the change in contour must not exceed 30° (a 3:1 taper).
• Misalignment is measured with a hi-lo gauge, which reads the internal step directly at several points around the circumference.

Never correct hi-lo by hammering the pipe. Forcing alignment introduces locked-in stress at exactly the joint that will later be radiographed and thermally cycled.

---

3. Fit-Up and Tacking

Alignment clamps, chain clamps, cage clamps and tack welding practice are developed in Section 2.2. Three fabrication-specific points belong here:

1. Minimum four tacks on a pipe joint, equally spaced (12, 3, 6 and 9 o'clock), so the joint cannot close unevenly as the root pass shrinks.
2. Tacks are part of the finished weld. They must be made by a qualified welder with the WPS filler metal, and their ends must be feathered by grinding so the root pass can fuse through them — or they must be removed as the root progresses.
3. Bridge tacks (tacks made on a separate strap across the joint, not in the groove) are used where the WPS prohibits tacks in the groove, and are ground flush and inspected after removal.

Root opening control. As the root pass runs, the joint ahead of the arc closes from weld shrinkage. On long-seam and large-diameter work the fitter sets a slightly wider opening at the far side to compensate. This is why the root gap is checked with a gauge at several clock positions before welding, not once at the top.

---

4. Cut Length: Putting the Numbers Together

The takeout constants and the offset trigonometry are established in Section 7.1. The fabrication step is combining them with the root opening deduction, which is added back because the two ends stand apart:

$L_{\text{cut}} = D_{\text{c-c}} - \sum(\text{fitting takeouts}) - \sum(\text{root openings})$

Worked example — a flanged spool. A shop drawing calls for a 4 in. Schedule 40 spool: weld neck flange – straight pipe – long radius 90° elbow, face of flange to centerline of elbow = 48.000 in. The weld neck flange has a length-through-hub of 4.00 in. (from the B16.5 table for a 4 in. Class 300 flange as detailed on the drawing). The LR 90° elbow takeout is $1.5 \times 4 = 6.00$ in. The WPS root opening is 1/8 in. at each of the two welds.

$L_{\text{cut}} = 48.000 - 4.000 - 6.000 - (2 \times 0.125) = 48.000 - 10.000 - 0.250 = 37.750\ \text{in.}$

The cut length is 37-3/4 in. Note that the flange dimension is subtracted as a length-through-hub, not as a takeout from a centerline — a flange has a face, not a center, and mixing the two conventions is the most common cut-length error on flanged spools.

---

5. Fabrication Sequence and What Governs It

   RECEIVE + VERIFY MATERIAL (MTR, PMI on alloy)
              |
   CUT TO LENGTH  --> square cut, remove the heat number? NO:
              |        transfer heat number to the offcut FIRST
   BEVEL (machine for alloy, torch/grind for carbon)
              |
   FIT + ALIGN (hi-lo check) --> TACK (4 minimum, feathered)
              |
   WELD per WPS (preheat, interpass - Chapter 10)
              |
   NDE  --> PWHT if required  --> NDE again if required
              |
   HYDROSTATIC TEST  -->  clean / blast / coat  -->  SHIP

Heat-number transfer before cutting is a code requirement, not paperwork discipline (Section 13.2). The moment a marked length is cut, traceability on the offcut is broken unless the number was transferred first.

---

6. Realistic Trade Scenario: A Rejected Girth Weld That Was a Fit-Up Problem

Radiography rejects a 6 in. Schedule 80 SA-106 B girth weld for incomplete root penetration around 90 degrees of the circumference, on the 4 o'clock to 7 o'clock arc.

The welder's qualifications and machine settings are fine. The fit-up record explains it:

1. The root opening was gauged only at the 12 o'clock position and read 1/8 in. — correct.
2. At the bottom of the joint, the pipe was resting on a sand bag rather than a roller, and the opening there had closed to under 1/16 in.
3. The root pass ran cleanly through the top of the joint and could not reach the root through the closed section at the bottom.

Correction: cut out and re-prep the joint, re-fit with the pipe properly supported on rollers, and gauge the root opening at all four quadrants before tacking. The site adds a fit-up hold point requiring the fitter to record four gap readings, which is a fifteen-second check that prevented three further rejections in the same outage.