Swaged Needles: Cracking, Splitting, And How To Stop Both

Nov 16, 2026

 

Pain Points: The Defect That Appears After the Part Looks Finished

Transverse splits and longitudinal cracks are the most expensive defects in swaged needle production, because they are the ones that escape. A split that opens during forming is caught and scrapped at negligible cost. A crack that closes under the die and reopens during ultrasonic cleaning, passivation or sterilisation reaches the customer, and the failure is then discovered inside a finished device where the cost of traceability, rework and notification is orders of magnitude higher.

The second cost is diagnostic. A crack at the transition has at least six plausible causes: under-annealed tube, excessive single-pass reduction, worn mandrel, misaligned die halves, contaminated lubricant, or a pre-existing scratch on the incoming tube. Each has a different countermeasure, and shops that respond by tightening every parameter at once lose the information they needed to find which one was responsible. Reject rates fall slowly, nobody knows why, and the same defect returns after the next tool change.

Working Principle: Why a Cold-Formed Taper Tears

Cracking in cold forming is the point at which local strain demand exceeds local ductility. During swaging, the metal at the transition is asked to do two things simultaneously: compress axially as the die closes and draw forward into the taper. The result is a strain gradient along the axis, highest where the taper meets the full tube, and a tensile component around the circumference as material is pulled inward.

Austenitic stainless steel tolerates this well when it is soft, and poorly when it is already work hardened. Since the tube has usually been drawn to its final size, it arrives with accumulated dislocations; the remaining strain capacity is what the anneal restores. Remove that anneal, or apply it unevenly, and the transition is being asked to exceed a budget that has already been spent.

The second mechanism is surface-initiated. A scratch, a die mark or a folded burr acts as a stress raiser, and because the transition is the highest-stress location, a defect anywhere else on the tube is harmless while the same defect at the shoulder becomes the crack origin. This is why the incoming surface condition of fine tube matters more than its nominal dimensions.

The third mechanism is geometric. A sharp shoulder, an undersized blend radius or an oval die cavity concentrates strain in a short arc. Even a well-annealed tube will tear there, which is why tool design and material condition must be evaluated together rather than sequentially.

Classification: The Defect Families and What They Indicate

Longitudinal splits run along the taper axis and typically start at the shoulder. They indicate axial strain beyond ductility, and the usual causes are inadequate pre-swage anneal, too heavy a single pass, or a mandrel that is too large for the bore, which forces the wall to thin rather than thicken.

Transverse cracks cross the taper and are often invisible until the part is bent. They point to a strain gradient problem: too short a taper for the reduction, a sharp blend radius, or a feed rate that changed mid-stroke and left a step in the draw.

Edge chipping at the very tip is a third family and belongs to a different cause set. It appears when the die cavity has wear at the point, when the part is ejected before it has fully relaxed, or when a ground tip was over-polished until the wall at the apex was too thin to support handling.

Delayed cracks, which appear only after cleaning or sterilisation, are the fourth family and the most dangerous. They usually combine a marginal anneal with residual tensile stress from an over-heavy final pass, and they behave like a slow-opening hinge. Because the defect is time-delayed, only an accelerated stress test on first articles will expose it before shipment.

Practical Operation: A Sequence That Isolates the Cause

When a split appears, resist the urge to change several parameters. Start with the material: cut a length from the same lot, bend it to the intended radius, and record the outcome. If the tube cracks on a bare bend, the swage is not the problem and no die adjustment will fix it.

If the material passes, examine the tool. Check die half alignment with a soft wire or a lead imprint, measure the mandrel for wear at the working length rather than at the shank, and confirm the blend radius in the cavity. A mandrel worn by a few hundredths of a millimetre is enough to convert wall thickening into wall thinning at the tip.

Then examine the pass strategy. Reduce a single heavy stroke to two or three lighter passes, with a low-temperature stress relief between them, and compare the crack rate on the same lot. In most shops this change alone removes longitudinal splitting without touching the anneal.

Finish with the process environment. Lubricant that has degraded or absorbed particles increases friction, and friction at the transition is what converts axial push into circumferential tension. Replace lubricant on a schedule, filter it, and record the change so that a later defect can be correlated with it.

Field Experience: Patterns Worth Remembering

First, cracks that appear only after passivation are frequently a residue problem rather than a metallurgical one. Passivation attacks a strained, smeared surface preferentially, opening a fold that was closed but not healed. A light electropolish before passivation, or a controlled stress relief after forming, removes the fold and with it the complaint.

Second, the shoulder is the only place worth polishing. Operators who polish the whole taper to remove marks are thinning the wall precisely where stress is highest. Written limits on diameter change at the transition, rather than on polishing time, keep the well-intentioned repair from creating a fatigue site.

Third, a bend test on every lot is cheaper than a final inspection on every batch. Shops that added goods-in bend testing reported a step change in transition splits, and the test costs a length of tube and a minute of operator time.

Fourth, timing decides what the inspection can find. A crack that is held closed by residual compression will not appear until the part is stressed in the direction that opens it, which usually means the bend test rather than the visual check. For parts that will be sterilised repeatedly, an accelerated stress cycle followed by low-magnification examination of the transition is worth building into the first-article protocol, because it reproduces the sequence that produces delayed failures in the field.

Summary

Cracking in swaged needles is a strain-budget failure, and it has four recognisable families: longitudinal splits, transverse cracks, tip chipping and delayed cracks. Isolating the cause requires testing material first, tooling second, pass strategy third and process environment last. Annealing, blend radius, mandrel condition and lubricant state are the four variables that decide whether the transition forms or tears.

Outlook

Detection is moving earlier. In-line vision at the die exit, eddy-current or dye-penetrant sampling on first articles, and automated bend fixtures that reproduce the clinical radius are replacing the magnifier and the eye. Under ISO 13485 the useful consequence is that crack prevention becomes a documented control rather than an operator skill, and process validation can show a strain window that survives lot-to-lot material variation. Suppliers able to demonstrate that window will be preferred for the thinner-wall work that device designs continue to demand.