Instrumentation fittings get the attention; tube clamp spacing gets skipped. That is backwards. Fatigue from mechanical vibration accounts for the majority of tube-line failures, and an unsupported span is a spring waiting for something to excite it — a compressor, a pump, or plain flow turbulence. The fix is unglamorous: supports at the right intervals, near the right components. The numbers exist, they are published, and nobody reads them. Here they are.
Key takeaways
- Unsupported tube spans vibrate at a natural frequency set by stiffness, mass and span length; resonance with machinery or flow excitation fatigues the metal until it cracks.
- Published maximum spans for horizontal instrument tube: 1/4–3/8 in OD at 0.9–1.0 m; 1/2 in at 0.9–1.5 m depending on the guide; 5/8–3/4 in at 1.2–1.5 m; 1 in at 1.5 m.
- These numbers come from manufacturer guides and MSS suggested tables — not from ASME B31.3, which publishes no small-tube span table.
- Clamp near, not on: supports go adjacent to fittings, bends and heavy valves — a heavy component swinging free is what breaks the connection.
- Tube never carries equipment: valves and manifolds get their own supports; tubing must not be used to support anything.
Why spacing matters
A span of tube between two supports is a beam with a natural frequency. Compressors, pumps, vibrating valves and even vortex shedding from the flow inside can excite it; when forcing frequency meets natural frequency, the span resonates, bending stress cycles at high frequency, and fatigue cracks grow — the published bending-stress relation for a vibrating span runs with the inverse square of span length, which is why doubling a support interval roughly quadruples the stress. Crack initiation on a clean unsupported span tends to happen at mid-span, where the fundamental mode bends most; hang a heavy valve or fitting unsupported in the middle of a run and the failure site moves to that connection instead. Manufacturer tubing literature states the blunt version: fatigue from mechanical vibration accounts for the majority of tube-line failures. The same failure logic is why impulse lines earn their own routing rules in instrument impulse lines.
Typical span table by OD
Published maximum support spacing, horizontal runs, from manufacturer installation guides and industry suggested tables. Guides disagree by up to ~50% at some sizes — the conservative value is the safe choice.
| Tube OD | Max support span | Sources |
|---|---|---|
| 1/4 in (6 mm) | 0.9–1.0 m | maker install guides (2 agree) |
| 3/8 in (10 mm) | 0.9–1.0 m | maker install guides (2 agree) |
| 1/2 in (12 mm) | 0.9–1.5 m | maker guides 0.9–1.0 m; MSS suggested 1.5 m |
| 5/8 in (16 mm) | 1.2–1.5 m | maker guide, MSS |
| 3/4 in (18–20 mm) | 1.2–1.5 m | maker guide, MSS |
| 1 in (25 mm) | 1.5 m | maker guide, MSS |
Vertical runs: the guides publish horizontal numbers; the only widely published vertical figures come from traced/jacketed tube guidance (every 3–4.5 m). Treat horizontal values as the conservative default for vertical instrument runs unless the project standard says otherwise. And note what these tables are: suggestions from installation guides and the MSS suggested-spacing table — ASME B31.3 contains no small-tube span table at all, so “the code says” is never the right citation here.
Clamp types
- Two-bolt cushioned clamps — the workhorse: a metal or plastic body with an elastomer or plastic liner around the tube, bolted closed. The liner damps vibration and isolates galvanic pairs; two-bolt bodies with rubber grommets are the classic form.
- Stacking/rail clamps — polypropylene or stainless bodies that stack two or three high on a mounting rail, with strut-nut or weld-plate mounting for panels and racks.
- P-clamps — single-screw clamps for fastening one tube to a wall or frame, quick but lighter duty.
- Weld-plate and stand mounts — clamp bodies welded to structure; weld with the tube and plastic liner removed, obviously.
Whatever the type, the clamp must anchor to rigid structure — a bundle of tubes clamped to each other, with nothing tied back to steel, is decoration, not support.
Where NOT to clamp
- Not on the fitting or the bend. Supports go adjacent to unions, tees and elbows — as close to bends as practical, roughly 150 mm from fitting and bend ends in classic guidance — holding the straight run, not the component. Clamp the fitting body itself and thermal movement and assembly stress have nowhere to go.
- Not flanking nothing. A heavy valve, manifold or filter in the line gets support of its own, with supports as close as possible on each side; a lump of brass swinging on two meters of 1/4 in tube will fatigue the connection regardless of span table compliance.
- Not tubing-as-structure. Tubing must never be used to support equipment, and rigidly tying a vibrating machine’s small-bore connections back to the machine frame transmits vibration into the nozzle instead — brace back to independent structure and keep nozzle loads within the equipment’s allowable.
For the bend-and-support interplay on the tube itself, see how to bend stainless steel tubing without kinking; for long continuous runs, stainless steel coil tubing changes the joint count but not the support logic. FerruleX supplies the tube side from the instrumentation tubing range.
RFQ lines
clamp type, tube OD, material— e.g.,two-bolt cushioned clamp, 1/2 in, 316 with PP liner.mounting— weld plate, strut nut or rail system.liner material— PP standard, other elastomers for temperature or media.- Span standard for the project, if one exists — otherwise state “per guide values.”
FAQ
How far apart should I support instrument tubing? Published maximums for horizontal runs: 1/4–3/8 in OD at 0.9–1.0 m, 1/2 in at 0.9–1.5 m depending on the guide, 5/8–3/4 in at 1.2–1.5 m, 1 in at 1.5 m. Guides disagree at some sizes by up to half — use the conservative value, and treat these as guide suggestions, not code mandates.
Is clamp spacing in ASME B31.3? No. The small-tube span numbers come from manufacturer installation guides and the MSS suggested-spacing table; ASME B31.3 contains no instrument-tube span table. “Per code” is the wrong citation for support spacing — cite the guide, or the project standard.
Where do tube fatigue cracks actually start? On a clean unsupported span, at mid-span, where the bending mode peaks. Add an unsupported heavy valve or fitting and the failure moves to that connection — the lump becomes the mass and its root becomes the crack site. Support the components and the span, and both failure paths close.
Can I clamp several tubes together instead of to structure? Only if the bundle itself is anchored to rigid structure. Tubes clamped to each other with no tie-back sway as a group — the clamps organize the mess without stopping the vibration. Anchor to steel.
Does vertical tubing need the same spacing? The published tables are for horizontal runs; for vertical instrument tube there is no widely published plain-tube figure — traced-tube guidance runs 3–4.5 m. Using the horizontal values on vertical runs is the conservative default unless the project standard overrides.
