Coverage angle per film

Two independent things limit how much of a girth weld a single exposure can record. The first is simply how much film is there: a cassette of diagnostic length L laid on a pipe of outside diameter D covers an arc of 360 x L / (pi x D) degrees. The second is beam obliquity: away from the perpendicular the beam passes through 1/cos(phi) times the nominal wall, and once that ratio exceeds what the density band and IQI will tolerate the film is no longer diagnostic even though it is still exposed.

Which one bites depends on the technique. On a panoramic crawler shot the beam is perpendicular everywhere, obliquity is not a limit at all, and film length alone decides how many cassettes go round the joint - typically two or three overlapping strips on a large-diameter line. On an external shot the obliquity limit is far more restrictive than film length, so the arc is set by the thickness ratio and the film is longer than the diagnostic area it carries.

Diagnostic length is not physical film length. Both ends carry identification, location markers and the overlap that proves adjacent films joined without a gap, and none of that counts as evaluated area. Deduct the overlap and identification allowance before dividing into the circumference, and always round the film count up - a joint short by half a film is a joint not radiographed.

Location markers must be placed so that the evaluated coverage can be demonstrated on the film itself, which is what turns this arithmetic into an auditable record rather than an estimate.

Circumference C = pi x OD
Diagnostic length L = film length - overlap allowance
Arc per film = 360 x L / C
Films to wrap the joint = ceil(C / L)
Obliquity arc = 2 x acos(1/k)