Charpy Arrest Energy Calculator (Battelle CVN Requirement)

Returns the full-size Charpy V-notch energy CVN_req = 0.0108·σ_h²·(R·t)^1/3 a pipe must have to arrest a running ductile fracture — the Battelle/AGA correlation behind the arrest-toughness requirement of ASME B31.8 ¶841.1.2 — together with the ratio of the mill-certified CVN you enter to that requirement, and a PASS/FAIL verdict. The available CVN is always your input from the material test report.

Method last updated 2026-07-30 (calculation changelog) · fixture-verified on every build — most recently 2026-07-31.

Most pipe failures are local. A running ductile fracture is not: once a gas transmission line tears axially, the escaping gas keeps pressurising the crack tip from behind, and the fracture can travel for hundreds — historically thousands — of feet in a few seconds before something stops it. The thing that stops it is toughness. This is the one pipeline check where the governing material property is not strength at all: a higher-grade, thinner-walled pipe is harder to arrest, not easier, because the driving force scales with hoop stress squared.

The physics is a race. Ahead of the crack tip, the decompression wave in the gas travels at the speed of sound in the fluid; the crack travels at whatever speed the available energy sustains. If the pipe absorbs energy fast enough that the crack decelerates below the decompression wave, the tip runs into already-depressurised gas and stops. If it does not, the crack outruns its own decompression and keeps being fed. The Battelle two-curve model is the full treatment of that race; the single-equation correlation implemented here is the screening form derived from it, and it is the form the gas pipeline codes reference for conventional line pipe.

What the equation asks for is deceptively simple: the operating hoop stress in ksi, the pipe's outside radius and wall thickness in inches, and the Charpy energy the steel actually delivers. Two of those are geometry you already know. The third — σ_h — is the reason arrest is a design decision rather than a material one: because it enters squared, a line operated at 40 ksi hoop needs roughly 1.8 times the toughness of the same pipe operated at 30 ksi. Derating a line is a legitimate arrest strategy, and sometimes the only affordable one on existing pipe.

The fourth input is the one people get wrong, and it is worth stating plainly up front: the available CVN must be the full-size (10 × 10 mm) equivalent value, at the minimum design temperature, for the pipe body — not a sub-size specimen value read straight off a certificate, not a weld or heat-affected-zone value, and not a room-temperature number. The correlation is calibrated for full-size specimens. Entering a 2/3-size result without converting it understates the toughness you have and can turn a genuinely adequate pipe into a spurious FAIL.

Method

The Battelle/AGA correlation for the minimum full-size Charpy energy required to arrest a propagating ductile fracture in conventional line pipe:

CVN_req = 0.0108 · σ_h² · ( R · t )^1/3

ratio = CVN_available / CVN_req → PASS when ratio ≥ 1

Units are fixed by the coefficient and are not interchangeable: σ_h in ksi (not psi), R the outside radius in inches (not the diameter), t the wall thickness in inches, and the result in ft·lbf. The 0.0108 constant carries the unit system inside it — it is a published empirical coefficient fitted to full-scale burst test data, not a code table value, which is why it is implemented here rather than left as an input.

The shape of the equation is worth reading before the number is. The requirement rises with the square of hoop stress and only with the cube root of the R·t product. That asymmetry is the whole engineering story: operating pressure dominates, and geometry barely moves the answer. Doubling R·t — a much larger, much heavier pipe — raises the toughness requirement by only 26%, while a 25% increase in hoop stress raises it by 56%.

The engine validates that σ_h, R and t are positive and that the available CVN is non-negative, then reports the requirement, the ratio to five decimals, and the verdict. Every run carries a standing warning that the correlation is calibrated for conventional-strength line pipe: rich gas, high grades from X80 upward, and low design temperatures all sit outside the data the fit was built on and need the governing code's correction factors or a full two-curve assessment.

Limitations — what this calculator is not

• Conventional-strength line pipe only. The correlation was fitted to full-scale burst tests on grades through roughly X70 in lean natural gas. Modern high-grade pipe (X80 and above) is known to arrest less readily than this equation predicts, and the calculator will happily return an optimistic requirement for it — the standing warning says so on every run. High-grade projects need the governing code's correction factor or a full Battelle two-curve analysis.
• Lean natural gas assumed. Rich gas, dense-phase gas, CO₂ and other fluids with a saturation plateau decompress far more slowly, holding pressure at the crack tip for longer and raising the toughness needed — sometimes by a large multiple. This single-equation form has no fluid model in it at all: the decompression behaviour that drives the entire arrest problem enters only through the correlation's original calibration.
• CVN_available must be the full-size equivalent at the minimum design temperature, for the pipe body. Sub-size specimen values need conversion before entry, and weld/HAZ toughness is a separate qualification the calculator does not touch. It cannot tell a mis-entered value from a genuine one.
• It is an arrest criterion, not an initiation criterion. Passing here says a fracture that has already started should stop; it says nothing about whether one will start. Initiation is governed by defect size, fracture toughness at the flaw, and the pressure cycle — that is the domain of the metal-loss and defect assessments, not this equation.
• Ductile fracture only. Brittle (cleavage) fracture arrest is a different problem governed by transition temperature and the ductile-to-brittle behaviour of the steel, typically addressed by a drop-weight tear test (DWTT) shear-area requirement rather than a Charpy energy. Both requirements normally apply; only one of them is calculated here.
• Mechanical crack arrestors — integral, wire-wrapped, composite or heavy-wall inserts — are outside the scope. Where pipe-body toughness alone cannot arrest, they are the usual answer, and their spacing and design are not evaluated here.
• No code verdict. A PASS is a statement that the entered CVN clears the correlation's requirement at the entered hoop stress. Which correlation, which correction factors, and what test regime the governing code and jurisdiction actually require for your project is your determination, and it varies by edition and by fluid.

Quick reference — what actually moves the requirement

The two exponents in the equation are wildly asymmetric, and reading them is more useful than any single result. Multipliers below are the equation's own sensitivity, not test data:

The practical consequence: when a line fails this screen, look at the operating hoop stress first and the pipe geometry last. Changing steel is usually the most expensive fix and the smallest lever.

Worked example — fixture-verified

An NPS 12 gas transmission line: 12.75 in OD (6.375 in outside radius) with a 0.375 in wall, operating at a hoop stress of 40 ksi. The mill certificates for the heat report a full-size Charpy energy of 30 ft·lbf at the minimum design temperature.

Step by step

1. Geometry term: R·t = 6.375 · 0.375 = 2.390625 in².
2. Cube root of the geometry term: (2.390625)^1/3 = 1.33712.
3. Stress term: σ_h² = 40² = 1,600 ksi², and 0.0108 · 1,600 = 17.28.
4. Required arrest energy: CVN_req = 17.28 · 1.33712 = 23.105 ft·lbf.
5. Compare against what the pipe has: ratio = 30 / 23.105 = 1.2984.
6. The available toughness exceeds the requirement by about 30%, so the verdict is PASS — a fracture in this pipe at this operating stress is predicted to arrest.

A comfortable pass, but read the margin in the right currency. The 30% headroom is in energy, and energy scales with the square of hoop stress — so it is consumed by roughly a 14% increase in operating stress, not a 30% one. An uprate from 40 to 46 ksi hoop on this same pipe would put it under the requirement. That asymmetry is why arrest margin should be checked against the pressure the line might be uprated to, not only the pressure it runs at today.

Worked example 2 — the same pipe, a leaner heat of steel

Identical pipe, identical operating stress — the only change is the material certificate. This heat reports 20 ft·lbf full-size Charpy energy rather than 30. Nothing about the geometry or the operation moved; only the steel did.

Step by step

1. The requirement is a property of the pipe and the operating stress, not of the steel's toughness — it is unchanged at 23.105 ft·lbf.
2. Ratio: 20 / 23.105 = 0.8656 — the pipe delivers about 87% of the energy the correlation asks for.
3. Verdict: FAIL. A running ductile fracture in this pipe at 40 ksi hoop is not predicted to arrest on pipe-body toughness alone.
4. The shortfall is 3.1 ft·lbf. Recovering it by derating means dropping the hoop stress by roughly 7% — to about 37.2 ksi — because the requirement falls with the square of σ_h.

The useful part of this result is how small the gap is and how many ways there are to close it. Three are ordinary: specify a tougher heat for new pipe, derate the segment, or fit mechanical crack arrestors at an appropriate spacing. One is not ordinary and is worth naming because it happens — re-reading the certificate and discovering the reported value was a sub-size specimen, which converts upward to a full-size equivalent and may clear the requirement outright. Check the specimen size before you spend money on the other three.

Fixture case “20 ft·lb available — fails” (tolerance 0.001) — locked in the same release gate as the example above.

Sources & citations

• ASME B31.8, Gas Transmission and Distribution Piping Systems — ¶841.1.2, fracture control and arrest toughness requirements for steel pipe.
• W. A. Maxey et al., Battelle Memorial Institute / American Gas Association — full-scale burst test programme and the ductile fracture arrest correlation, including the two-curve model this screening form derives from.
• API 5L, Line Pipe — source of the supplementary Charpy impact requirements and the pipe body test regime; the certified CVN value is user-supplied from the material test report.
• ASTM E23, Notched Bar Impact Testing of Metallic Materials — full-size (10 × 10 mm) specimen basis the 0.0108 coefficient is calibrated to; sub-size results require conversion before use.
• The 0.0108 coefficient is a published empirical constant of the correlation, not a code table value.

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 does so under ASME authorization and states the source table and conditions inline.

FAQ

Is this the same as the DWTT shear area requirement?

No, and they are usually both required. The Charpy arrest energy addresses ductile fracture propagation — how much energy the steel absorbs while tearing. The drop-weight tear test shear-area requirement addresses whether the pipe fractures in a ductile mode at all at the minimum design temperature, rather than cleaving in a brittle one. A pipe can have generous Charpy energy measured above its transition temperature and still fail a DWTT requirement at design temperature; the two screens catch different failure modes. This calculator does the first only.

Why does higher-grade pipe make arrest harder rather than easier?

Because grade buys strength, and strength gets spent on operating at a higher hoop stress, which is the term that enters squared. A line built in X70 rather than X52 and operated at the same design factor runs at a proportionally higher σ_h, so its arrest requirement rises by the square of that ratio while the toughness the mill delivers does not automatically follow. It is compounded by a second, separate effect the equation does not model: at grades from X80 upward the correlation itself becomes non-conservative, over-predicting arrest. That is a known limitation of the single-equation form, and it is the reason high-grade projects use the two-curve model or a code correction factor rather than this screen.

What CVN value should I enter from my material test report?

The full-size (10 × 10 mm) equivalent energy for the pipe body, at or below the minimum design temperature, and normally the minimum or the average of the specified test set as your specification requires — not the best single result. Two traps are common. First, sub-size specimens: thin-wall pipe is often tested with 2/3 or 1/2-size bars, and those values must be converted to a full-size equivalent before entry, otherwise you understate the toughness you have. Second, temperature: a room-temperature number on a line whose minimum design temperature is well below that is not the relevant value, because Charpy energy falls as temperature drops through the transition range.

The line fails. What are my options in practice?

Four, roughly in order of how often they are used. Derate the segment — the requirement falls with σ_h squared, so a modest pressure reduction goes a long way, and on pipe already in the ground it is often the only option. Specify higher toughness on new pipe, which is straightforward at order time and impossible afterwards. Fit mechanical crack arrestors at a spacing that limits how far a fracture can run before it reaches one; that is standard practice on large-diameter, high-pressure gas lines and is a design exercise this calculator does not perform. Or re-examine the inputs: a sub-size specimen entered without conversion, or a hoop stress computed at test rather than operating pressure, is a cheaper thing to find than any of the other three.

Does this apply to liquid pipelines?

Not usefully. Running ductile fracture is a compressible-fluid problem: it needs stored energy in the fluid column to keep pressurising the crack tip as the tear advances. A liquid line depressurises essentially immediately at a rupture, so a fracture has nothing feeding it and arrests within a short distance regardless of toughness. That is why arrest toughness requirements live in the gas pipeline codes. Liquid lines still have toughness requirements — for initiation resistance, for low-temperature service, and for weldability — but they are not set by this correlation.

Where does the hoop stress input come from?

From the pressure you are qualifying and the pipe geometry, by the Barlow form σ_h = P·D/(2·t), converted to ksi. Use the operating stress at the pressure the arrest case is being argued at — normally MAOP — rather than a test pressure or a design-factor-derated allowable. If MAOP has not been established yet, or if the segment is limited by something other than the pipe, settle that first with the MAOP calculator: qualifying arrest against a hoop stress the line will never see is a wasted calculation, and qualifying against one it will exceed is a dangerous one.

Related calculators

• Pipeline MAOP Calculator — Maximum Allowable Operating Pressure (ASME B31.8 / B31.4) — Sets the operating hoop stress the arrest requirement scales from
• ASME B31.4 / B31.8 / B31.11 Pipeline Wall Thickness Calculator — Wall and grade behind the hoop stress and the R·t term
• Pipeline Hydrostatic Test Pressure Calculator (ASME B31.4 / B31.8) — The other pipe-qualification check on the same line