Tube Weld Acceptance Calculator (ASME BPE MJ-8.4-1)

Checks orbital tube-weld O.D. concavity, convexity and radial misalignment against their ASME BPE Table MJ-8.4-1 limits — concavity and misalignment as fractions of the wall thickness T_w, convexity as an absolute height, with an absolute misalignment limit for the largest tube sizes — and reports each ratio, which discontinuity governs, and a PASS/FAIL.

Method last updated 2026-09-23 (calculation changelog) · fixture-verified on every build — most recently 2026-10-12.

Built and fixture-verified by Matthew Norris, P.E. — active P.E. licensure in Arizona, California, Kansas, Missouri, North Carolina, Texas.

Autogenous orbital welds on hygienic tube are accepted largely on visual and dimensional criteria, and three of those criteria are numeric: how far the weld surface sinks below flush (concavity), how far it crowns above it (convexity), and how far the two tube ends are radially stepped across the joint (misalignment). ASME BPE Table MJ-8.4-1 sets a limit on each, and states them on different bases. This calculator takes the three outside-surface measurements and the limits — each with its basis — that you read from your own copy of that table, forms the ratio of each measurement to its own limit, and reports which of the three governs alongside a PASS/FAIL. Reporting the governing discontinuity is the point: three separate pass/fail checks tell you whether a weld is acceptable, while the ratios tell you which one is close, and a coupon running at 0.95 on convexity is a machine drifting toward a rejection that has not happened yet. Reach for it at weld coupon acceptance, during procedure qualification when parameters are being tuned, and on production weld logs where the trend across a shift matters more than any single joint.

Method

concavity limit = fraction × T_w

convexity limit = absolute height (in) — or fraction × T_w if you set that basis

misalignment limit = fraction × T_w — absolute (in) for the largest tube sizes

ratio = measurement / its own limit · governing = largest ratio · PASS when ≤ 1

Each of the three limits carries its own basis, because Table MJ-8.4-1 does not state them alike, and the calculator takes the basis as an input rather than assuming one. On the outside (non-process-contact) surface, where these three measurements are taken, concavity is limited as a fraction of the nominal wall of the thinner member — in effect a statement about how much section the weld removed, the same proportion on thin-wall and heavy-wall tube. Misalignment is also a fraction of T_w for most tube sizes; for the largest sizes the table switches to an absolute outside-diameter limit, and the card's misalignment basis switches with it. Convexity on the outside surface is an absolute height. The product-contact column reads differently — both concavity and convexity there scale with T_w, and there is no numeric inside-diameter misalignment criterion at all — which is the other reason the basis is an input: enter the column that applies to the surface you measured. Each measurement is divided by its own resolved limit, the largest of the three ratios is reported as the governing discontinuity, and the weld passes only if all three pass. Where a ratio exceeds 1 the calculator reports the disposition MJ-8.4.2 draws: concavity and convexity may be rewelded once by reflow, while misalignment is not on that list, because a reflow cannot move the tube ends. The card opens with the three limit fields blank and computes nothing until you fill them: read each limit and its basis from your own copy of the table for the surface and tube size — the examples below show values as a user would enter them, not defaults.

Limitations — what this calculator is not

• The acceptance limits are inputs, not authority. Each limit and its basis is read from your own copy of Table MJ-8.4-1 — the card pre-fills none of them — and a transcription or basis error produces a confident, wrong verdict.
• These are three of the acceptance criteria, not all of them. Discoloration and heat tint, lack of penetration, incomplete fusion, cracks, porosity, undercut, weld width and the condition of the ID bead are separate criteria assessed visually and borescopically, and a weld can clear all three numbers here and still be rejectable.
• The check is dimensional only and says nothing about metallurgy. Sensitisation, an unfused root, oxide inclusions from an inadequate purge, and sulfur-driven penetration variation are invisible to it — the ferrite and weldability card covers the chemistry side of that question.
• Measurements are only as good as the method that took them. Concavity and convexity assessed by eye against a comparator, by profile gauge, or by a calibrated instrument give different numbers on the same weld, and the acceptance criterion assumes the method the standard specifies.
• ID-side discontinuities are the ones that matter hygienically and are the hardest to measure. This card checks the O.D. entries; the table's product-contact column limits ID concavity and convexity separately and sets no numeric ID misalignment criterion, even though the ID step is the hygienically relevant one. Borescope examination of the ID bead is a separate and non-optional step.
• One weld per run. A production weld log is repeated application, and it is the trend and the worst joint — not the average — that an inspection acts on.
• The misalignment basis follows tube size. For most sizes it scales with T_w; for the largest sizes the table gives an absolute O.D. limit instead, and nothing in the input set flags a basis left on the wrong setting when the tube size changes.
• The O.D. concavity entry tolerates a deeper dip over a limited part of the circumference. The check applies the single fraction you enter, so enter the full-circumference value unless you are dispositioning a local dip against that local allowance.
• Misalignment is excluded altogether for extruded branch outlets and for reducers (Table MJ-8.4-1 Note (9)); the misalignment ratio means nothing on those joints.
• A failing ratio is not automatically a cut-out, and not automatically a reweld. MJ-8.4.2 permits a single reflow attempt for concavity and convexity, among other listed defects, and does not list misalignment — the calculator names which case applies, but the disposition itself is a quality decision, not an arithmetic one.

Worked example — fixture-verified

A production coupon on 3/4 in OD × 0.065 in wall tube, measured on the outside surface: concavity 0.004 in, convexity 0.005 in, misalignment 0.007 in. Limits as entered from the project's copy of Table MJ-8.4-1 for this size and surface.

Step by step

1. Allowables: concavity 0.10 × 0.065 = 0.0065 in; convexity 0.015 in as entered; misalignment 0.15 × 0.065 = 0.00975 in.
2. Concavity ratio = 0.004 / 0.0065 = 0.6154.
3. Convexity ratio = 0.005 / 0.015 = 0.3333.
4. Misalignment ratio = 0.007 / 0.00975 = 0.7179.
5. Largest ratio is misalignment → governing, and it is at or below 1: PASS.

An acceptable weld with roughly 28 % margin on its governing discontinuity — and on thin-wall tube the governing one is misalignment, because fifteen percent of a 0.065 in wall is under a hundredth of an inch. Worked example 2 keeps the weld and changes only the fit-up.

Worked example 2 — same weld, 0.010 in of mismatch

Identical weld measurements except that radial mismatch is 0.010 in — a fit-up and clamping problem, not a welding-parameter one, and a mismatch that looks comfortable against a fixed 0.015 in limit.

Step by step

1. Concavity and convexity ratios are unchanged at 0.6154 and 0.3333.
2. Misalignment ratio = 0.010 / 0.00975 = 1.0256.
3. Largest ratio is now misalignment, and it exceeds 1 → governing, FAIL.
4. Disposition note raised: misalignment is not among the MJ-8.4.2 reweld-eligible defects.

Read against a fixed 0.015 in limit — the common misreading of the table — this joint scores 0.67 and passes; against the wall-scaled limit it fails. The weld itself is still acceptable on both profile criteria: the joint failed on how the tubes were presented to the machine, and MJ-8.4.2 offers no reflow for it. Adjusting weld current here would be treating a fit-up defect as a parameter defect. Check tube-end squareness and ovality, confirm both ends are within their dimensional tolerance, re-clamp, and re-make the joint.

Fixture case “same weld, misalignment 0.010 in over the Tw-scaled limit -> fail” (tolerance 0.0001) — locked in the same release gate as the example above.

Sources & citations

• ASME BPE-2026 — Table MJ-8.4-1 and Fig. MJ-8.4-1: O.D. and I.D. acceptance criteria for concavity, convexity and misalignment, with Notes (2) and (7)–(9); limits and their basis are user-supplied inputs.
• ASME BPE-2026 — MJ-8.4.2, rewelding: one reflow attempt, and only for the listed defects.

Per the source & citation policy, allowable-stress and factor table values are inputs — typed from your governing edition, or autofilled on paid tiers from the embedded datasets enumerated there, always cited by standard, edition, table and line. 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

Why do the three limits use different bases?

Because Table MJ-8.4-1 does not state them alike, and a calculator that assumed one basis for all three would be wrong on at least one of them. On the outside surface, where these measurements are taken, concavity is a fraction of the nominal wall T_w of the thinner member — a statement about how much section the weld removed, the same proportion on 0.049 in and 0.083 in tube. Misalignment is also a fraction of T_w for most sizes, per Figure MJ-8.4-1 illustration (b); only the largest tube sizes switch to an absolute outside-diameter limit. Convexity on the outside is an absolute height. The product-contact column reads differently again: inside concavity and inside convexity both scale with T_w, and Note (7) sets no numeric inside misalignment criterion, even though Note (8) recognizes the inside step as the hygienically relevant one. Note (9) takes misalignment off the table entirely for extruded branch outlets and reducers. So the card makes the basis an input for each limit: read the column for the surface you measured and the row for your tube size.

Why does the tool report a governing discontinuity instead of three verdicts?

Because three independent pass/fail results tell you only whether today's coupon was acceptable, while the ratios tell you where the process is heading. A weld log where the governing ratio is convexity and it has crept from 0.6 to 0.9 over a shift is a machine drifting — electrode wear, gas flow, a heat change — and it is a rejection that has not happened yet. Naming the governing discontinuity also directs the corrective action to the right place: convexity and concavity respond to welding parameters, misalignment responds to fit-up and tube tolerance, and treating one as the other wastes a shift. The verdict is still all-or-nothing; the governing ratio is what makes it actionable. It is also what a sample-weld schedule is built to catch: MJ-8.4 recommends a coupon at the start of every shift, whenever the purge-gas bottle is changed and whenever the orbital weld head itself is swapped — exactly the points where a drift begins. Logging the governing ratio at each of those checkpoints turns a shift's scattered PASS verdicts into a trend pointed at the failure before it arrives.

The weld passed all three numbers. Is it acceptable?

Not on this evidence alone. These are three numeric criteria among a larger acceptance set that is mostly visual: discoloration and heat tint level, lack of penetration, incomplete fusion, cracks, porosity, undercut, weld width and uniformity, and the condition of the ID bead. Hygienically the ID surface is the one that matters and it is not characterised by OD measurements at all — a weld with excellent OD geometry can have an oxidised, rough or unfused root if the purge was inadequate. Table MJ-8.4-1 even leaves weld bead width open on the process contact side, with no numeric limit at all provided full penetration is achieved; width becomes a measured criterion, checked for straightness and a minimum against the maximum bead, only once the ID can't be reached and the OD bead is inspected as its stand-in. Borescope examination of the ID is a separate and non-optional acceptance step, and the coupon regime exists precisely because the production weld's inside cannot always be seen.

What causes convexity to creep up during a shift?

Usually the arc losing the force that depresses the pool, or the internal purge pushing it outward. Electrode wear and tip geometry change are the most common: as the tip degrades the arc broadens, arc force and penetration drop, the pool is depressed less, and the bead sits proud. Excessive backing-gas pressure does the same from the inside, lifting the pool outward while hollowing the ID, and a heat change to material at a different point in its sulfur range shifts the convection direction. The opposite drift — arc energy going up, or a fixture running hotter as the shift progresses so successive welds start from a higher baseline temperature — pushes the profile the other way, toward drop-through and concavity. That is exactly why watching which of the two governs is diagnostic: rising convexity with stable misalignment points at the weld head or the gas, not at fit-up, and not at too much heat.

Can a rejected weld simply be rewelded?

Within limits set by MJ-8.4.2, and those limits are narrower than a simple retry: the clause names only four reweldable defects — incomplete penetration, incomplete fusion, an unconsumed tack weld that can still be inspected on the process contact side, and convexity or concavity. Misalignment is not on that list. Reflow rewelding fuses more of the same joint; it cannot move a tube end that was out of square or clamped off-center, so a joint rejected on misalignment is not reweld-eligible at all — the fit-up has to be corrected and the joint recut, not reflowed. Repeated fusion of the same joint also degrades the metallurgy and the ID surface each time it is attempted, which is why the standard bounds it rather than leaving it open-ended. The calculator's contribution is naming the governing discontinuity so the disposition starts from the right cause: a convexity or concavity rejection may pass after a parameter or electrode change and a reweld attempt, while a misalignment rejection goes straight to requalifying the fit-up.

Related calculators & tools

• Ferrite / Weldability Calculator (WRC-1992, ASME BPE) — Chemistry behind erratic penetration and cracking
• Ra Surface-Finish Calculator (ASME BPE SF Designations) — Weld finish is checked against its own Ra population
• Dead-Leg L/d Ratio Calculator (ASME BPE) — Fitting geometry the weld joins, on the same line
• ASME BPE Fitting Dimensions Lookup (Part DT-4.1) — The tangent length the orbital head clamps on, per fitting
• ASME BPE hygienic design guide — Joining and examination in the hygienic design case