Every thread page in this database prints its flank angle in a badge, and the thread index groups the 10 systems by exactly that number: 164 of the 210 designations are cut at 60°, 20 at 55°, and the remaining 26 follow the 30° trapezoidal and 29° Acme profiles.
That grouping is not wrong — the flank angle is the first thing a thread gauge checks and usually the only angle on a drawing. It is simply not the number that decides whether two threads will run together. Two families can share a 60° flank and differ by 30% in how deep the profile is cut, and the 55° family is not at either end of that range.
Depth is the part the drawing leaves out
A thread is a triangle wound around a cylinder. The flank angle sets the directions of its two sides; it says nothing about how far that triangle is allowed to stand out. That is the thread height, and each standard decides it separately by choosing how much of the sharp triangle to truncate at the crest and at the root.
So depth on its own cannot be compared across sizes: a 4 in pipe thread is deeper than an M3 because it is bigger, not because its profile is different. Divide depth by pitch and the size cancels out, leaving a pure shape factor. Every row in the database has both numbers, so the ratio can be computed rather than quoted:
| Thread | Flank (°) | Sizes | Depth ÷ pitch | Minor Ø, internal vs external |
|---|---|---|---|---|
| ISO metric | 60 | 82 | 0.613 | different |
| UNC | 60 | 25 | 0.613 | different |
| UNF | 60 | 23 | 0.613 | different |
| UNEF | 60 | 9 | 0.613 | different |
| BSP | 55 | 11 | 0.640 | same |
| BSPT | 55 | 9 | 0.640 | same |
| NPT | 60 | 14 | 0.800 | different |
| NPS | 60 | 11 | 0.800 | different |
Three values for eight families. The four 60° screw-thread families share 0.613. The 60° pipe families sit at 0.800 — 30% deeper than the screw threads of their own angle. And the 55° British families come in at 0.640: 4% above the 60° screw threads, and 20% below the 60° pipe threads.
Read the angle column alone and the 55° profile looks like one end of the ordering. It is in the middle, with a 60° family on either side.
The reason the pipe threads are deeper is not the angle but the joint: a pipe thread has to interfere along the flanks to be a joint at all, so it is cut closer to the sharp triangle than a fastener thread, which only has to clear its root and carry load on the flanks. The NPT pages print their profile as "0.8P thread height", the BSP pages as "Whitworth basic profile" — the name is doing the work the angle cannot.
One minor diameter, or two
The last column of the table is the quiet one. In 20 of the 184 V-profile rows the internal and external minor diameters hold the same number, and those 20 rows are exactly the two British families. The other 164 print two, with the external always the smaller.
That is not an editorial choice. The Whitworth profile rounds the crest and the root by the same amount, so a screw and the nut it fits are cut to the same basic minor diameter and one number describes both. The Unified and pipe-60° profiles truncate the external root flat and deeper than the internal one, so the screw bottom and the nut bottom are two different diameters. One of them is what callipers reach inside a nut; the other is what a thread micrometer measures on a bolt.
Nor is the split confined to the V profiles. Of the 26 power-thread rows, the 10 Acme rows carry one number and the 16 trapezoidal rows carry two: a profile that shares the clearance between the two members collapses the two columns, one that puts it all on the external thread keeps them apart.
A drawing that quotes a single "minor diameter" on a 60° thread is therefore ambiguous; on a BSP or an Acme thread it is not.
The taper is why a drill size is missing
The tap-drill column is empty on 23 of the 210 rows, and those 23 are exactly the tapered ones: all 14 NPT and all 9 BSPT. Every parallel row has a number, including NPS — which is the same thread as NPT in every column except taper and drill.
That sameness is exact. All 11 size labels the two families share agree on tpi, major diameter, pitch diameter, minor diameter and depth. The only differences between the two pages are the word "taper" and the drill size that taper makes unquotable, because a tapered hole is a cone: there is no single diameter to drill to before the tap follows the taper in. On the 1/2-14 NPT page the field is written as not applicable with the reason attached, rather than left blank or filled with the parallel-thread number.
1/2 is not 1/2
Both families publish 9 trade sizes, from 1/8 to 2. Line them up and the difference is not where most people expect it:
| Trade size | NPT TPI | BSP TPI | NPT major (mm) | BSP major (mm) | Difference (mm) |
|---|---|---|---|---|---|
| 1/8 | 27 | 28 | 10.287 | 9.728 | 0.559 |
| 1/4 | 18 | 19 | 13.716 | 13.157 | 0.559 |
| 3/8 | 18 | 19 | 17.145 | 16.662 | 0.483 |
| 1/2 | 14 | 14 | 21.336 | 20.955 | 0.381 |
| 3/4 | 14 | 14 | 26.670 | 26.441 | 0.229 |
| 1 | 11.5 | 11 | 33.401 | 33.249 | 0.152 |
| 1-1/4 | 11.5 | 11 | 42.164 | 41.910 | 0.254 |
| 1-1/2 | 11.5 | 11 | 48.260 | 47.803 | 0.457 |
| 2 | 11.5 | 11 | 60.325 | 59.614 | 0.711 |
Only 2 of the nine share a pitch — 1/2 and 3/4, both at 14 threads per inch — and neither of those shares the flank angle. Every other pair differs in pitch as well, so the count on its own rules them out: 1 in is 11.5 threads per inch against 11.
The closest pair is not at the small end. At 1 in the two major diameters are 0.152 mm apart — 0.46% of the BSP diameter: small enough for the thread to start, large enough that it cannot run down, and the 60° against 55° mismatch guarantees the flanks never sit flat on each other.
At 1/2 the two are 0.381 mm apart on diameter with an identical pitch, which is exactly the combination that gets a fitting started by hand and then cross-threaded. The 1/2-14 BSP page and its NPT twin are two pages for a reason.
What the angle still tells you
- It separates the families at a glance. 60° against 55° is the difference a thread gauge shows immediately, and it rules out every cross-family substitution before you measure anything.
- It does not set the depth. The same 60° covers 0.613 and 0.800, a 30% spread, and the 55° family falls between them.
- It does not predict interchangeability inside an angle. The 60° families are not interchangeable with each other either: metric and Unified share a profile but not a diameter series, and NPT and NPS share a diameter series but not a taper.
Where the constant stops working
The ratio is a constant for V profiles because the height is a fixed fraction of the pitch. It is not a constant for the power threads, where the depth is set by the clearance class rather than the angle: across the 16 trapezoidal rows it runs from 0.550 to 0.625, and across the 10 Acme rows from 0.540 to 0.660. Not one of them repeats a value often enough to be called a profile constant, and a table that quoted a single number for TR would be wrong for most of its own rows.
So the rule this page ends on is narrow and it is the one that matters at the bench: for a V thread, depth divided by pitch identifies the profile; for a power thread, it identifies nothing and the clearance class has to be quoted alongside.
What this page does not tell you
- Pressure. No pressure rating is in this database for any of these threads, and the wall thickness under the thread is what sets it.
- Which sealant. Not a column here, and the right answer depends on the fluid, the temperature and whether the joint will be opened again.
- Tolerance class. The rows carry basic dimensions. A 1/2 NPT and a 1/2 NPS page agreeing on tpi, major, pitch, minor and depth says nothing about whether a given pair of parts will assemble.
- Material. The thread is a shape; the pipe it is cut into is not.
The two families compared here are indexed at NPT and BSP, the parallel versions at NPS and BSPT, and the ISO metric series covers the rest of the fastener threads. The thread index has the rest.