ASME BPE Fitting Dimensions Lookup (Part DT-4.1)

Looks up ASME BPE-2026 Part DT-4.1 nominal fitting dimensions — center-to-end A and bend radius R for 90° and 45° elbows, center-to-end A and run end-to-end B for tees and crosses, overall length A and maximum internal taper α for reducers, ferrule A/B/C, and cap height — for 1/2 in through 6 in tube, and derives the straight weld tangent A − R·tan(θ/2) and the eccentric-reducer centerline offset.

Method last updated 2026-09-01 (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.

Every hygienic spool is a chain of fitting take-outs. The cut length of a straight, the center-to-center of two elbows, the room an instrument tee leaves for a diaphragm valve, whether an orbital weld head can clamp on the end of an elbow at all — all of it comes back to the nominal dimensions ASME BPE tabulates in Part DT-4.1 for automatic-weld and hygienic-clamp fittings. This calculator reads those tables from embedded reference data: select a fitting type and a tube size and it returns the governing table and every dimension that table carries for the row — center-to-end A and centerline radius R for weld elbows, center-to-clamp-face A for clamp elbows, A and B for tees and crosses, run and branch center-to-end for instrument tees, overall length and maximum internal taper for reducers, A/B/C for weld-end ferrules and minimum cap height for solid end caps. Two derived quantities come with the lookup: the straight tangent length on a weld elbow, which is what the weld head actually clamps on, and the centerline offset of an eccentric reducer. Reach for it at isometric take-off, when checking a fabricator's cut list, when dispositioning a received fitting against the drawing, and whenever a B16.9 habit — elbow center-to-end equals radius — is about to produce a spool that is short by two tangents.

Method

weld elbow tangent = A − R·tan(θ/2) (θ = 90° → A − R; θ = 45° → A − 0.4142·R)

eccentric reducer centerline offset = (OD_large − OD_small) / 2 (BPE tube OD = nominal size)

The lookup is a direct table read, and it is deliberately not more than that: the fitting type selects the DT-4.1 table, the size selects the row, and every dimension the row carries is returned with the table id so the number can be traced to its source line. Equal fittings — elbows, tees, crosses, ferrules, caps — are tabulated by a single tube size from 1/2 in to 6 in; reducers and instrument tees are tabulated as run × branch pairs, entered exactly as the table designates them. The elbow tangent is derived from the geometry of a bend: a centerline arc of radius R turning through θ reaches the tangent point R·tan(θ/2) short of the centerline intersection, so whatever remains of the center-to-end dimension beyond that is straight tube. For a 90° elbow that is simply A − R; for a 45° elbow the bend takes only 0.4142·R of the length, which is why the 45° A is so much shorter than the 90° A at the same radius while the tangent lands in the same neighbourhood. The two tables carry independent A values rather than a shared formula, so the 45° and 90° tangents agree to a few hundredths of an inch at most sizes and differ by about 0.13 in at 1 in nominal — comparable, not identical. The eccentric offset follows from BPE tubing being an OD-based product: a 2 × 1 reducer joins a 2.000 in OD to a 1.000 in OD, so with both tubes tangent on one side the centerlines sit 0.500 in apart. The card reports COMPUTED rather than PASS/FAIL — there is no acceptance criterion in a dimension lookup — and an unlisted size returns a failure that names the sizes the table does carry.

Limitations — what this calculator is not

• Nominal dimensions only. DT-4.1 carries tolerance columns alongside every dimension and they are not reproduced here — dispositioning a received fitting means reading the tolerance from the table, not comparing the measurement against this nominal.
• Coverage is the US Customary tube range, 1/2 in through 6 in. Metric hygienic fittings to DIN or ISO tube sizes are a different table set and are not in this lookup; neither are sizes above 6 in.
• This is the dimensional table and nothing else. Tube outside diameter and wall come from Part DT-3, material from Part MM, surface finish from Part SF and weld acceptance from Part MJ — each has its own card on this line, and none of them is implied by a dimension match.
• ASME B16.9 butt-weld fittings are not comparable and not interchangeable. A B16.9 long-radius elbow is dimensioned center-to-end with A equal to its 1.5 × NPS radius and no straight tangent; a BPE weld elbow carries a tangent beyond the bend so the orbital head can clamp, and BPE tube OD equals the nominal size where pipe OD does not. A 2 in B16.9 LR elbow and a 2 in BPE elbow are different products with different take-outs.
• Reducer and instrument-tee sizes are run × branch as tabulated — instrument tees carry a branch larger than the run, so a 1 X 1-1/2 is a 1 in run with a 1-1/2 in branch. Entered the other way round the row does not exist and the lookup says so.
• Reducer lengths are the BPE-2026 short-pattern values. Catalogs derived from earlier editions list different overall lengths for some sizes, so a spool detailed against an old catalog will not close against fittings supplied to the current table.
• Concentric and eccentric reducers share one table in BPE-2026 — identical overall length and taper. The offset reported for the eccentric pattern is geometric, from the two tube ODs; the table itself does not tabulate it.
• Ferrule dimensions B and C are reported as tabulated. The clamp, gasket and mating-ferrule stack-up at a joint is not computed, and the length a clamp joint adds to a spool depends on the gasket and clamp selected.
• Solid end cap height is a minimum, not a nominal. The derived tangent and offset are rounded to three and four decimals respectively; the tabulated values are returned exactly as the table gives them.

Worked example — fixture-verified

A 2 in hygienic transfer line turns two 90° corners in the same plane. The isometric shows the two elbow centerline intersections 24 in apart. The fabricator needs the cut length of the straight between the elbows and the length of tube the orbital head will clamp on at each elbow end.

Step by step

1. Table DT-4.1.1-1, 2 in row: center-to-end A = 4.75 in, centerline radius R = 3.0 in.
2. Weld tangent = A − R·tan(45°) = 4.75 − 3.0 = 1.75 in of straight tube beyond the bend at each weld end.
3. Straight between the elbows = 24 − 2 × 4.75 = 14.5 in, before any fit-up allowance the WPS calls for.

That 1.75 in tangent is the whole reason the elbow is 4.75 in long rather than 3.0: it is the straight the weld head collet grips. Detail the spool against the B16.9 reflex — center-to-end equals radius — and the straight comes out 3.5 in too long. The three examples below take the same table through a tee, an eccentric reducer and a 45° elbow.

Worked example 2 — a 1 in weld tee, and where the dead-leg measurement starts

A 1 in equal tee, weld ends, is set into a run to carry a sample valve. The take-off needs the run take-out; the hygienic review needs to know how far the branch weld end already sits from the run before a valve is added.

Step by step

1. Table DT-4.1.2-1, 1 in row: center-to-end A = 2.125 in, run end-to-end B = 4.25 in — B is 2A, so the run take-out is the whole fitting.
2. The branch weld end sits 2.125 in from the run centerline. On 1 in × 0.065 in wall tube (ID 0.870 in) the run's inside wall is 0.435 in from the centerline, so the branch cavity is already 2.125 − 0.435 = 1.69 in deep at the weld end.
3. Against the 0.870 in branch ID that is L/d = 1.69 / 0.870 = 1.94 before any valve body is added.

An equal tee spends almost the whole SD-3.1.2.2 dead-leg allowance on its own geometry, which is exactly why instrument tees exist: the DT-4.1.2-10 row for a 1 in run with a 1-1/2 in branch gives a branch center-to-end of 1.125 in on a larger bore. Run the Dead Leg calculator with the valve seat location for the verdict — this page only tells you where the measurement starts.

Fixture case “tee weld 1 in: A 2.125 center-to-end, B 4.25 run end-to-end = 2A (DT-4.1.2-1)” (tolerance 1e-9) — locked in the same release gate as the example above.

Worked example 3 — 2 × 1 eccentric reducer: length, taper and centerline offset

A horizontal 2 in header steps down to 1 in with an eccentric reducer so the invert stays continuous and the line drains. The isometric needs the overall length and the centerline drop across the reducer.

Step by step

1. Table DT-4.1.3-1, 2 × 1 row: overall length A = 3.375 in, internal taper α ≤ 30°. The concentric pattern shares the row.
2. BPE tube OD equals nominal: 2.000 in and 1.000 in. With the flat side on the bottom, offset = (2.000 − 1.000) / 2 = 0.500 in — the 1 in centerline sits half an inch below the 2 in centerline.

Flat side down keeps the bottom of the line continuous through the size change, which is the drainability case; flat side up is the pump-suction habit from non-hygienic practice and leaves a pocket on the invert. The offset is what the downstream elevation on the isometric has to absorb, and on a sloped line it also changes where the GSD check is measured from.

Fixture case “eccentric reducer 2 X 1: A 3.375, taper 30 deg max, centreline offset (2.0 - 1.0)/2 = 0.5 (DT-4.1.3-1)” (tolerance 1e-9) — locked in the same release gate as the example above.

Worked example 4 — 6 in 45° elbow: why the tangent is the same and A is not

A 6 in line makes a 45° offset. The fabricator has the 90° elbow dimensions in hand and wants to know why the 45° fitting is nearly half the length at the same bend radius.

Step by step

1. Table DT-4.1.1-4, 6 in row: center-to-end A = 6.25 in, centerline radius R = 9.0 in.
2. The bend reaches its tangent point R·tan(22.5°) = 9.0 × 0.4142 = 3.728 in short of the centerline intersection.
3. Weld tangent = 6.25 − 3.728 = 2.522 in. The 6 in 90° elbow (A 11.5, R 9.0) carries 2.5 in — close to the same straight.

The fitting family is built on one radius per size and a comparable tangent for each angle; the angle mostly changes how much of the centerline the arc consumes. That is the sanity check to run on any 45° dimension that looks wrong on a drawing: subtract R·tan 22.5° and the remainder should land near the 90° elbow's tangent for the same size — within a few hundredths of an inch at most sizes, and as much as 0.13 in at 1 in nominal, because the two tables carry independent A values rather than a shared formula. A remainder that is off by half an inch is a transcription error; one that is off by a tenth is the table.

Fixture case “45 deg weld elbow 6 in: A 6.25, R 9.0, tangent A - R*tan(22.5 deg) = 2.522 (DT-4.1.1-4)” (tolerance 0.000001) — locked in the same release gate as the example above.

Additional verified cases in this fixture

input:  {"fittingType":"ferrule-weld","size":"6"}
expect: {"listed":1,"table":"DT-4.1.4-1","A":3,"B":1.5,"C":0.75}
tol:    1e-9

input:  {"fittingType":"instrument-tee-weld","size":"1 X 1-1/2"}
expect: {"listed":1,"table":"DT-4.1.2-10","A":2.5,"B":1.125}
tol:    1e-9

input:  {"fittingType":"reducer-conc-weld","size":"6 X 3"}
expect: {"listed":1,"table":"DT-4.1.3-1","A":7.25,"taper":44}
tol:    1e-9

input:  {"fittingType":"solid-end-cap-clamp","size":"4"}
expect: {"listed":1,"table":"DT-4.1.5-2","A":0.312}
tol:    1e-9

input:  {"fittingType":"elbow-90-weld","size":"8"}
expect: {"listed":0,"table":"DT-4.1.1-1"}
tol:    1e-9

input:  {"fittingType":"reducer-ecc-weld","size":" 2 x  1 "}
expect: {"listed":1,"table":"DT-4.1.3-1","A":3.375,"taper":30,"eccentricOffset":0.5}
tol:    1e-9

Sources & citations

• ASME BPE-2026 — Part DT-4.1 fitting-dimension tables: DT-4.1.1-1 / -3 / -4 / -6 (90° and 45° elbows, weld and clamp ends), DT-4.1.2-1 (tees, crosses), DT-4.1.2-10 (instrument tees), DT-4.1.3-1 (concentric and eccentric reducers), DT-4.1.4-1 (weld-end ferrules), DT-4.1.5-2 (solid end caps); centerline radius R per Part DT-3.

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

What is dimension A on a BPE elbow — center-to-end or face-to-tangent?

Center-to-end: from the intersection of the two centerlines to the weld end of the fitting, the same convention the tee table uses. It is not the tangent length. The tangent — the straight tube between the end of the bend and the weld end — is what remains after the arc's share R·tan(θ/2) is taken out, which for a 90° elbow is A − R. On a 2 in 90° weld elbow that is 4.75 − 3.0 = 1.75 in; on the 45° elbow of the same size the arc takes only 0.4142 × 3.0 = 1.24 in of the 3.0 in center-to-end, leaving a tangent of about 1.76 in — close to the 90° elbow's 1.75 in, which is the internal consistency check that confirms A is center-to-end in both tables. The agreement is not exact at every size (the two tables carry independent A values, and the 1 in row differs by about 0.13 in), but it holds to within a few hundredths at most sizes. The reason the tangent exists at all is the orbital weld head: an enclosed head clamps a collet on each side of the joint and needs straight tube to grip, so a hygienic elbow cannot end at the bend the way a B16.9 elbow does.

Why is a BPE elbow so much longer than a B16.9 elbow of the same size?

Two reasons, and they compound. The tangent is the first: a B16.9 long-radius 90° elbow is dimensioned center-to-end with A equal to its radius of 1.5 × NPS, so the fitting ends exactly where the bend ends; the BPE elbow carries the straight tangent beyond that for the weld head. The second is the product itself — BPE tubing is OD-based, so a 2 in tube is 2.000 in OD where a 2 in pipe is 2.375 in OD, and the fittings are dimensioned around the tube, not the pipe. A 2 in B16.9 LR elbow has a 3.0 in center-to-end; the 2 in BPE weld elbow has 4.75 in. Detailing a hygienic spool from B16.9 habit under-lengths every elbow by the tangent and mis-sizes the tube, and the two fittings will not weld to one another in any case. Treat them as different product lines that happen to share a size name.

Do concentric and eccentric reducers have the same length?

In BPE-2026, yes — the two patterns share Table DT-4.1.3-1, so a 2 × 1 reducer is 3.375 in overall with an internal taper of 30° maximum whichever pattern you order, and a 6 × 3 is 7.25 in at 44°. The difference is where the centerlines sit. The eccentric pattern keeps one side of the two tubes tangent, so the centerlines are offset by half the OD difference: 0.500 in on a 2 × 1, 1.500 in on a 6 × 3. In a horizontal hygienic line the eccentric reducer goes flat side down so the invert stays continuous and the line drains toward the low point; a concentric reducer on a horizontal line leaves a step on the bottom that holds liquid after CIP. The pump-suction convention of flat side up — to avoid trapping air at the impeller — is the opposite orientation and belongs to a different problem.

Which sizes and units does the lookup cover?

The US Customary hygienic tube range, 1/2 in through 6 in, with every dimension in inches exactly as DT-4.1 tabulates it. BPE tube is designated by its outside diameter — a 1-1/2 in tube is 1.500 in OD — which is why the eccentric offset can be derived from the size names alone. Reducers and instrument tees are tabulated as run × branch pairs and must be entered with the same designation the table uses; a reducer size that is not a tabulated pair, or an instrument tee entered branch × run, is reported as not listed along with the sizes the table does carry. Metric hygienic tube to DIN 11850 or ISO 1127 sizes uses a different fitting-dimension system and is not covered.

Does this give tolerances, wall thickness or the tube OD?

No. It returns the nominal dimensions from DT-4.1 and nothing from the tolerance columns beside them, so it is a take-off and design tool rather than an inspection acceptance tool — a received fitting is dispositioned against the tolerance in the table, not against the nominal here. Tube OD and wall come from Part DT-3, and the wall matters for the dead-leg calculation because it sets the branch inside diameter; the 0.065 in wall used in the tee example above is the common gauge for 1 in hygienic tube, not a table value from this lookup. Surface finish, material chemistry and weld acceptance each have their own card on this line and are independent of a dimension match.

How do I turn these dimensions into a cut list?

Work in centerlines. Every fitting is placed by its centerline intersection — the point the isometric dimensions to — and each fitting's A is the distance from that point to its weld end. A straight between two fittings is the centerline distance minus the two A values that face it: 24 in between two 2 in 90° elbows leaves 24 − 2 × 4.75 = 14.5 in of tube. For a tee in a run, B is the whole run take-out, so the straights either side lose A each. Reducers consume their overall length A along the run, and on the eccentric pattern the far-side centerline moves by the offset, which the next dimension has to be taken from. Autogenous orbital butt welds are normally fit up with no root gap, but any allowance is the welding procedure's call, not the table's — add it after the geometry, per joint, and note it on the cut list so the fabricator and the inspector are reading the same number.

Related calculators & tools

• Dead-Leg L/d Ratio Calculator (ASME BPE) — Branch center-to-end sets the L the dead-leg ratio starts from
• Tube Weld Acceptance Calculator (ASME BPE MJ-8.4-1) — What the weld at each fitting end has to meet
• Hygienic Line Slope / GSD Drainability Calculator (ASME BPE) — Eccentric reducer orientation keeps the invert draining
• CIP Flow & Velocity Calculator (5 ft/s Rule, ASME BPE) — The tube size you looked up is the ID the CIP velocity is checked at
• ASME BPE hygienic design guide — Where fitting geometry sits in the hygienic design case