AMGTA commissioned study finds 3D printed stents more durable than traditional manufacturing

The AMGTA commissioned study finds its result by running the same stent geometry through two routes: a metal additive process and a conventional manufacturing chain. Both sets were measured on the same bench, so the comparison is like for like on geometry. The variables that move are the ones tied to the process itself: grain structure, surface condition, and how much stock gets removed or scrapped.

Durability here means fatigue life under cyclic radial loading, not a single static burst number. A stent that survives one crush test tells you little. What matters is how many cycles it holds before a crack initiates and propagates. Additive builds tend to show finer, more uniform grain in thin struts because the melt pool solidifies fast and stays local.

The study also tracked material consumption and energy per finished part. That number is not a side note. For a stent with 0.1 mm struts, conventional routes start from tube or sheet and remove most of the material as scrap. Additive deposits only what the strut needs. That gap shows up in the energy accounting.

One caveat before anyone quotes the headline. The comparison is on specific geometries and specific alloys. It is not a blanket claim that every printed stent beats every machined or braided stent. Read it as a process-capability signal for thin-strut metal structures.

Fatigue life in a stent lives at the surface. A crack starts where stress concentrates, and that is usually a surface defect, a sharp internal corner, or a coarse grain boundary. Additive parts come out with a rough as-built surface, so they need post-processing before the fatigue number means anything. That step is not optional in a real program.

Conventional machining of a thin strut introduces its own risks. Cutting forces deflect slender walls. Tool marks create stress risers. If the strut is 0.15 mm wide, a 0.02 mm tool mark is a significant fraction of the section. Additive avoids the cutting force entirely, which is exactly why thin, lattice-like or curved strut arrays tend to come out cleaner.

Counterpoint: additive brings porosity risk. A gas-entrapped pore near the surface acts like a pre-crack. That is why the durability result depends on parameter control and on inspection that can actually see subsurface defects. Without that, the fatigue advantage is not reliable.

So the finding is conditional. Printed stents can be more durable when the build is tight, the surface is finished to spec, and the parts are inspected for internal porosity. Skip any of those and you lose the benefit.

Choose additive when the geometry is the hard part. Curved strut arrays, variable cross-section along the length, internal lattice, or a design that would need several setups on a mill. If the part needs three or more orientations to reach all features, printing starts to look competitive.

Stay with conventional when the part is a simple axisymmetric profile at volume. A straight tube with uniform wall is fast on a lathe, cheap per unit, and easy to inspect. Printing it adds build time and powder handling for no gain in function.

There is a middle path worth knowing. For prototypes and small runs, printing a test article lets you validate the strut geometry and the fatigue behavior before committing to a production route. Then you can decide whether to scale the printed version or move to machining or forming.

At GreatLight we run both sides of this decision. Custom 3D printing for the geometry-led cases, and 16 simultaneous 5-axis machining centers plus 16 mill-turn centers for the parts that should be cut. The choice is made on strut width, volume, and inspection requirements, not on which process sounds newer.

• 1
  Pick additiveCurved or lattice struts, variable section, multiple reach directions
• 2
  Pick machiningSimple profiles, high unit volume, tight dimensional control
• 3
  Pick bothPrint the prototype, then cut the production run

A printed stent that goes straight from the build plate to a fatigue rig will underperform. The as-built surface is rough, and roughness is where cracks start. The study result assumes a finished surface, so the finishing step belongs in the process plan from day one.

For metal stents, relevant finishing includes bead blasting, tumbling and polishing to bring Ra into the Ra 0.2–0.8 μm range for critical surfaces. Electropolishing is common in the medical space because it removes surface peaks rather than folding them over. Each method changes dimensions, so the build has to be compensated.

Inspection needs to go beyond the surface. Porosity and lack of fusion sit below the skin. Dimensional checks and surface roughness readings will not catch them. Plan for the detection method early, because it affects the process window and the accept or reject criteria.

GreatLight holds ISO 13485:2016 for medical device work, alongside ISO 9001:2015, IATF 16949:2016 and ISO 27001:2022. Machined and printed parts run through raw material check, in-process monitoring and final inspection, with reports on request. Tolerance can be held to ±0.005 mm where the drawing calls for it.