Quick one for the cleared crowd: on defense prints, what’s the biggest root radius you’ve seen allowed on UNJ external threads? I ran a secured job last month that called out 0.150p, but our QA swore an MJ spec once let them go higher — anyone got a concrete example?
UNJ external root radius tops out at 0.180P — “0.150P–0.180P” is the band; MJ is the same per ISO 5855. Quick ref: https://www.amesweb.info/Screws/UNJ-Profile.aspx. If @QA saw higher, it was a special override on the drawing or a nonstandard crest truncation callout.
If that secured run was called out “0.150p,” it was almost certainly straight AS8879; I’ve never seen MJ allow >0.18P unless an OEM note overrides it. We once ran 0.20P per a Boeing mod and the UNJ GO/NO-GO got touchy fast, so we backed off the roll-die radius. @ava73 has the baseline right — QA might be recalling an internal crest limit, not the external root.
Quick example: on a 1/4‑28 UNJF I did last month, keeping the insert nose at 0.006 in held the J blend in the allowed band; swapping to a 0.008 in nose bumped it past the cap even though PD was perfect. If @QA remembers MJ being higher, check for a drawing note like “r per OEM spec” — that’s the only time I’ve seen it exceed the standard.
One thing that bit me: QA gauged the J fillet at the runout and called it oversize — verify the form “≥3P in” on the comparator, then sanity‑check minor dia to confirm the blend’s truly in. When I’m near the ceiling, a modified flank infeed and a no‑DOC spring pass keep the fillet from creeping up; @tommy_j56 have you had luck with that too?
Seen it once on MJ only when the print carried an OEM note — “root fillet may be increased provided minor dia remains within limits”; we cut a 10×1 at about 0.20P and it passed. What spec tag did that job carry (ISO 5855-x or AS8879)? Absent a note, stick to the standard cap and do @ella2190’s “≥3P in” check plus a quick minor-dia verify.