Pipeline crawler positioning

Pipeline radiography puts the source on a self-propelled crawler inside the line, stopped in the plane of the girth weld and fired panoramically through the wall onto film wrapped round the outside. It is the fastest production radiography there is - one exposure per joint, complete coverage - but it depends on the source stopping accurately in the weld plane.

If the source stops short by e, the beam meets the weld at an angle beta = atan(e / (ID/2)) instead of perpendicular. The wall path stretches to wall / cos(beta), and the weld image is displaced and distorted. Requiring the path ratio to stay within k gives the stopping tolerance e_max = (ID/2) x tan(acos(1/k)). On a 300 mm bore at a 1.1 ratio that is about 69 mm - generous, which is why magnetic weld-crown detection or an external marker is sufficient in practice. On small bore it tightens quickly, because the tolerance scales with the radius.

The panoramic geometry also fixes the unsharpness with no room to negotiate. The source sits on the axis, so SOD = ID/2 and OFD = wall, giving Ug = F x wall / (ID/2). There is nothing to adjust: if the source is too big for the bore, the shot fails the ASME V T-274.2 limit and a smaller source or a different technique is required.

Production rate is the other half of the job. Cycle time per joint is travel plus exposure plus set-up and film handling, and on a long line the travel term between joints dominates the arithmetic that decides how many joints a shift will produce.

e_max = (ID/2) x tan(acos(1/k))
beta = atan(e / (ID/2))
Ug = F x wall / (ID/2)
travel time = joint spacing / crawler speed
cycle = travel + exposure + set-up