Expansion Loop / Leg Sizing Calculator (Guided Cantilever)
Produces the free thermal growth ΔL, the required guided-cantilever leg Lgc = √(3EDΔL/kS), and the W and H proportions of the loop, for L bends, Z offsets and U loops. Kellogg guided-cantilever method; layouts outside the ASME B31.3 ¶319.4.1 screen still need formal flexibility analysis.
Method last updated (calculation changelog) · fixture-verified on every build — most recently 2026-09-03.
Built and fixture-verified by Matthew Norris, P.E. — active P.E. licensure in Arizona, California, Kansas, Missouri, North Carolina, Texas.
Sizes expansion loops, Z-offsets, and L-bend legs by the guided-cantilever method — the classic hand calculation for absorbing thermal growth without a full flexibility analysis. Handles metallic pipe (k = 1, the Grinnell/Kellogg rule) and thermoplastics (k = 2, the PPI/Corzan form with working stress at temperature).
.pcp project files and the batch runner —
$2.99 / month. Compare tiers →Method
ΔL = α · |T1 − T0| · L (free growth of the run)
Lgc = √( 3·E·D·ΔL / (k·S) ) (required flexible leg)
Configurations: L change-of-direction uses the full leg; Z offset takes 0.6426·Lgc (the published 4 vs 6.225 coefficient ratio, E/S-invariant); U loop proportions the developed length as W = Lgc/5, H = 2W. With E = 29×10⁶ psi and S = 15,600 psi the metallic form reproduces the published rule W(ft) = 6.225·√(ΔL·D).
| Inputs | ||
|---|---|---|
| config | L bend / Z offset / U loop | — |
| family | Metallic (k=1) or plastic (k=2) | — |
| D | Pipe outside diameter | in |
| α | Expansion coefficient ×10⁻⁶ (user-supplied) | in/in·°F |
| T0 → T1 | Installation and operating temperature | °F |
| L | Anchor-to-anchor run length | ft |
| E, S | Modulus and allowable/working stress (user-supplied) | psi |
| Outputs | ||
| ΔL | Thermal growth to absorb | in |
| Lgc | Guided-cantilever leg | in / ft |
| W, H | Loop width and height (U config) | in |
Limitations — what this calculator is not
- The guided-cantilever model assumes the growth is absorbed by bending of one guided leg — anchors at both ends, guides placed to enforce the geometry, and no intermediate restraints stealing flexibility.
- It is deliberately conservative; where B31.3 ¶319 requires formal analysis (large bore, high temperature, cyclic severity), this sizes the concept and CAESAR II / AutoPIPE proves it.
- Pressure thrust, friction loads on guides, and support spacing inside the loop are separate design steps.
- α, E and S are material data you supply from the governing code tables or manufacturer data — presets in the app are starter values only.
Worked example — fixture-verified
2½ in (2.875 in OD) carbon-steel line, installed at 52 °F, operating at 350 °F, 100 ft anchor-to-anchor. α = 6.7×10⁻⁶ in/in·°F, E = 29×10⁶ psi, S = 15,600 psi — the published flow-sheet example.
| Given | ||
|---|---|---|
| Config / family | L bend, metallic | — |
| OD | 2.875 | in |
| α | 6.7×10⁻⁶ | in/in·°F |
| T0 → T1 | 52 → 350 | °F |
| Run length | 100 | ft |
| E / S | 29×10⁶ / 15,600 | psi |
Step by step
- Growth: ΔL = 6.7×10⁻⁶ · (350−52) · 1,200 in = 6.7×10⁻⁶·298·1,200 = 2.396 in.
- Leg: Lgc = √(3·29×10⁶·2.875·2.396 / 15,600) = √598,000,000/15,600 ≈ √38,415 = 196.0 in = 16.33 ft.
- Cross-check against the published coefficient rule: 6.225·√(ΔL·D) = 6.225·√(2.396·2.875) = 6.225·2.625 = 16.34 ft ✓.
- Same inputs as a Z-offset: 0.6426·196.0 = 125.9 in; as a U-loop: W = 39.2 in, H = 78.4 in.
| Result COMPUTED | ||
|---|---|---|
| ΔL — thermal growth | 2.396 | in |
| L-bend leg | 196.0 in (16.33 ft) | — |
| Z-offset leg | 125.9 | in |
| U-loop W × H | 39.2 × 78.4 | in |
Five fixture cases lock this module: the L/Z/U trio above, a CPVC (k=2) case, and a cold-contraction case with negative ΔT.
expansion-loop.json — case “PDF sheet L-bend: CS 2.875 OD, 52→350°F, 100 ft (E=29e6, S=15600)” (tolerance 0.005) — in the
calc-core release gate. It re-runs on every commit; a red fixture blocks deployment.
See the validation methodology.Worked example 2 — Z-offset configuration on the same conditions
The same 2.875 in OD carbon-steel line (52→350°F, 100 ft, E=29e6, S=15,600) as the main example, but routed as a Z offset instead of an L bend. The Z configuration uses less absolute leg length because the geometry redistributes the bending: the effective cantilever is shorter when the offset runs in both directions. The trade-off is a longer developed length in the horizontal plane.
| Given | ||
|---|---|---|
| Config / family | Z offset, metallic | — |
| OD | 2.875 | in |
| α | 6.7×10⁻⁶ | in/in·°F |
| T0 → T1 | 52 → 350 | °F |
| Run length | 100 | ft |
| E / S | 29×10⁶ / 15,600 | psi |
Step by step
- Growth is identical to Example 1: ΔL = 6.7×10⁻⁶ · 298 · 1,200 = 2.396 in.
- Lgc (guided cantilever) is also identical: √(3·29×10⁶·2.875·2.396 / 15,600) = 196.0 in.
- For a Z offset the required leg length uses the published coefficient 4/6.225 ratio: 0.6426 · 196.0 = 125.9 in (10.50 ft).
- Compared to the L-bend (196.0 in), the Z saves about 70 in of vertical or axial rise — the trade-off for needing horizontal clearance in both directions.
| Result COMPUTED | ||
|---|---|---|
| ΔL — thermal growth | 2.396 | in |
| Z-offset required leg | 125.9 in (10.50 ft) | — |
The Z offset is often the space-efficient choice when the pipe run has room in the horizontal plane but constrained height. The leg length is roughly 64% of the L-bend value because the offset geometry splits the bending across two legs at an angle, concentrating the deflection differently. Both still use the same thermal growth and the same Kellogg formula; the configuration coefficient is the only difference.
Fixture case “PDF sheet Z-bend: same inputs → leg = (4/6.225)·Lgc = 125.943 in (pub. 4·sqrt(ΔL·D) = 125.98)” (tolerance 0.005) — locked in the same release gate as the example above.
Worked example 3 — U-loop proportions on the same conditions
The same input data again (2.875 in, 52→350°F, 100 ft, E/S = 29e6 / 15,600), but laid out as a U loop — the symmetric expansion loop often drawn in P&IDs where the pipe has clearance both horizontally and vertically. The U configuration proportions the developed cantilever length as width W = L<sub>gc</sub>/5 and height H = 2W, giving a characteristic rectangular profile.
| Given | ||
|---|---|---|
| Config / family | U loop, metallic | — |
| OD | 2.875 | in |
| α | 6.7×10⁻⁶ | in/in·°F |
| T0 → T1 | 52 → 350 | °F |
| Run length | 100 | ft |
| E / S | 29×10⁶ / 15,600 | psi |
Step by step
- Thermal growth ΔL and guided cantilever Lgc are the same as before: 2.396 in and 196.0 in.
- Loop width: W = Lgc / 5 = 196.0 / 5 = 39.2 in.
- Loop height: H = 2·W = 2 · 39.2 = 78.4 in.
- Overall loop footprint: approximately 40 in wide by 78 in tall — a 1:2 aspect ratio typical for U loops, where loop width W and loop height H are set directly by the cantilever length rather than chosen independently. The total developed bending length (W + vertical legs + W) still equals Lgc ≈ 196 in, just reshaped.
| Result COMPUTED | ||
|---|---|---|
| ΔL — thermal growth | 2.396 | in |
| Lgc — cantilever | 196.0 | in |
| Loop width W | 39.2 | in |
| Loop height H | 78.4 | in |
The loop width W = L<sub>gc</sub>/5 and loop height H = 2W formula is the field standard for U-loop proportions — it reproduces the published coefficient rule W(ft) = 6.225·√(ΔL·D) when E = 29×10⁶ psi and S = 15,600 psi. These dimensions are the targets for drafting; the actual installed loop typically gets set by available clearance on the unit and the support locations.
Fixture case “PDF sheet U-loop: same inputs → W = 39.200, H = 78.399 (pub. 39.211 / 78.422)” (tolerance 0.005) — locked in the same release gate as the example above.
Worked example 4 — CPVC plastic pipe, shorter leg from k = 2
A transfer line in plastic piping (Corzan CPVC, 2 in IPS = 2.375 in OD) installed at 70°F, operating at 140°F, 80 ft anchor-to-anchor. Plastic allows a higher k factor (k = 2 in the PPI/Corzan convention, reflecting higher bending stress allowance) and a much lower working stress: E = 423,000 psi (vs ~29e6 for steel), S = 2,000 psi (vs 15,600+ for steel). The expansion coefficient α = 34×10⁻⁶ in/in·°F — more than 5× steel — so even a modest ΔT grows significantly.
| Given | ||
|---|---|---|
| Config / family | L bend, plastic (CPVC, k=2) | — |
| OD (IPS) | 2.375 | in |
| α | 34×10⁻⁶ | in/in·°F |
| T0 → T1 | 70 → 140 | °F |
| Run length | 80 | ft |
| E / S | 423,000 / 2,000 | psi |
Step by step
- Thermal growth: ΔL = 34×10⁻⁶ · (140−70) · 960 in = 34×10⁻⁶ · 70 · 960 = 2.285 in.
- Guided cantilever using the plastic form Lgc = √(3·E·D·ΔL / (k·S)): Lgc = √(3 · 423,000 · 2.375 · 2.285 / (2 · 2,000)) = √(6,886,704 / 4,000) = √1,721.7 = 41.49 in.
- Compare to the steel examples: CPVC grows less (ΔL 2.285 vs 2.396 in over a shorter run), but the k = 2 factor and low modulus compress the required leg to just 41 in — about 21% of the L-bend in Example 1.
| Result COMPUTED | ||
|---|---|---|
| ΔL — thermal growth | 2.285 | in |
| L-bend leg (k=2) | 41.49 | in (3.46 ft) |
| Leg as % of steel example | ~21% | — |
The dramatic short leg arises from two factors: the k = 2 doubling (allowing higher bending stress) and the low modulus of plastic (423,000 psi vs 29e6 psi for steel). A 41 in loop in CPVC is manageable in space-constrained applications — transfer lines, sample circuits, instrumentation taps — where the same growth on steel would need a 196 in leg. The trade-off: plastic allowables are lower and temperature-dependent, so the calculation must use working stress at the operating temperature (2,000 psi at 140°F here), and long-term creep under sustained load can occur if the loop is perpetually near the hot end.
Fixture case “CPVC arm (PPI/Corzan k=2): 2 in IPS D=2.375, α=34, 70→140°F, 80 ft, E=423000, S=2000 → 41.491 in” (tolerance 0.005) — locked in the same release gate as the example above.
Worked example 5 — Cold service contraction with negative ΔT
A stainless-steel line operating in a cold service: installed at 90°F in shop/ambient, operating at 20°F in a freezer or chilled-water circuit. The negative ΔT = −70°F means the line contracts instead of expanding. Thermal growth becomes negative (the pipe shrinks), and the calculation method is identical — the loop now lets the piping pull inward at both anchors rather than push outward. Geometry: 4.5 in OD, 60 ft anchor-to-anchor, Z offset configuration, α = 9.2×10⁻⁶ (stainless), E = 28.3e6, S = 16,700.
| Given | ||
|---|---|---|
| Config / family | Z offset, metallic | — |
| OD | 4.5 | in |
| α | 9.2×10⁻⁶ | in/in·°F |
| T0 → T1 | 90 → 20 | °F |
| ΔT (signed) | −70 | °F |
| Run length | 60 | ft |
| E / S | 28.3×10⁶ / 16,700 | psi |
Step by step
- Thermal movement (contraction): ΔL = 9.2×10⁻⁶ · |−70| · 720 in = 9.2×10⁻⁶ · 70 · 720 = 0.4637 in (negative — inward pull).
- Guided cantilever: Lgc = √(3 · 28.3×10⁶ · 4.5 · 0.4637 / 16,700) = √(177,156,585 / 16,700) = √10,608.2 = 103.0 in.
- For Z offset: 0.6426 · 103.0 ≈ 66.2 in, matching the fixture-locked 66.181 in (5.52 ft) required leg — notably shorter than the steel L-bend examples because the cold contraction (0.46 in) is less than the hot expansion (2.4 in) from the main example.
- Engineering note: the loop layout is identical to warm service — both anchors get the same flexibility. The difference: in cold service the anchors pull together instead of spreading apart; equipment nozzle loading is reduced (inward pull vs outward push), which is often the preferred mode for vessels and pump casings.
| Result COMPUTED | ||
|---|---|---|
| ΔT (signed) | −70 | °F |
| ΔL — thermal contraction | 0.4637 | in |
| Z-offset leg required | 66.181 | in (5.52 ft) |
Fixture case 4 closes the validation set for this module — the same Z-offset coefficient was already exercised in fixture case 1 above, just with growth instead of contraction. Contraction and expansion are mathematically and mechanically symmetric in the guided-cantilever method — the formula applies identically, and the fixture-lock includes this case to verify sign handling. In practice, cold service loops often need less absolute leg length because the temperature swing is smaller, and the inward anchor pull can be preferable to outward push on rigid equipment. The Kellogg method and the loop-sizing proportions (W = L<sub>gc</sub>/5 for U loops, 0.6426·L<sub>gc</sub> for Z offsets) apply in both directions without modification.
Fixture case “Contraction Z: SS 4.5 OD, α=9.2, 90→20°F (ΔT=−70), 60 ft, E=28.3e6, S=16700 → ΔL=0.4637, leg=66.181” (tolerance 0.005) — locked in the same release gate as the example above.
Sources & citations
- M.W. Kellogg, Design of Piping Systems — guided cantilever method (public-domain engineering method).
- Published coefficient rule W = 6.225·√(ΔL·D) (L/U) and 4·√(ΔL·D) (Z) — reproduced exactly by the general form at E = 29e6, S = 15,600 psi.
- PPI / Corzan CPVC practice — plastic form L = √(3·E·D·ΔL/(2·S)) with working stress at max temperature.
Per the source & citation policy, allowable-stress and factor table values are user-supplied. Where a page does reproduce specific ASME data (the B16.5 ratings, the quick-reference tables), it states the source table and conditions inline.
FAQ
When is a hand-sized loop enough, and when do I need pipe stress analysis?
B31.3 ¶319.4.1 exempts systems that duplicate a successful layout or pass the simplified screen; the guided cantilever is the accepted conservative sizing for those. Outside the exemption — large bore, > 350–400 °F, sensitive equipment nozzles, cyclic service — size the loop here, then verify in a flexibility program. The screen's own limits are also the honest checklist for staying inside it: the guided-cantilever picture assumes the legs are free to flex — a guide, a branch tie-in, or equipment placed on a loop leg silently voids the sizing — and it assumes the anchors take the reactions the loop generates, which this card reports precisely so someone verifies them. The loop that fails in service is rarely undersized; it is the correctly-sized loop whose legs got braced, tapped or shortened after the calculation was filed. If the routing changes, the sizing reruns — thirty seconds here against a field reroute is no contest.
Why does plastic pipe use k = 2?
The thermoplastic industry form (PPI, Corzan) allows the leg to be sized against 2× the working stress because the short-duration bending stress during excursions is checked against a different basis than the long-term hydrostatic rating. The low modulus of plastics keeps the legs short anyway — the CPVC fixture case needs only a 41.5 in leg. Two cautions carry the plastic case honestly. The material inputs shift with temperature far faster than steel's — the modulus at the operating temperature, not the room-temperature catalog value, belongs in the calculation, and manufacturers publish the derated curves for exactly this use. And solvent-cemented or fused joints near the loop corners live in the highest-strain region, so the fabrication quality of four joints decides whether the theoretical flexibility exists in the installed system. The short legs plastics permit make it tempting to skip the check entirely; the low allowable stress that comes with them is why that temptation loses.
Related calculators & tools
- Pipe Thermal Growth Calculator (in / 100 ft) — Supplies the growth the loop must absorb
- Displacement Stress Range Check (ASME B31.3 ¶319.4.4) — Verify SE once ¶319 requires formal analysis
- Stress Intensification & Flexibility Factor Calculator (B31.3 Appendix D) — Elbows in the loop set flexibility factor k
- Piping ISO Drawing (browser) — Sketch the loop as a dimensioned iso with BOM and cut list