ASTM F1798

What the test does

A single interconnection is isolated and loaded until it fails or lets go. The rod and its mating component — a pedicle screw, a hook, a cross-connector — sit in a jig built so everything except the joint is rigid, and the load path runs through the connection alone. Load is applied in one of six directions: three forces along the implant axes, three moments about them. Each direction has two variants: a slow ramp to failure, and a sinusoid cycled until the joint fails functionally or reaches a run-out of 2,500,000 cycles.

What it measures, and why it matters

The static tests give the force or moment the joint carries before it yields, slips or comes apart, and the displacement or rotation at which it does. Gripping capacity, the axial case along the rod, matters most directly: lose it and the correction set in theatre goes with it, though nothing has broken. The moment cases matter because a spinal construct sees flexion, extension and torsion far more than pure tension. The fatigue variants report cycles to functional failure — the practical question of whether a connection stays tight for the months before fusion. Because the test medium is a reported variable, a joint characterised dry is not describing the same behaviour as one run under saline.

Specimen

There is no coupon. The specimen is production hardware — a rod section, the mating component and the locking element — assembled with the instruments and tightening procedure a surgeon would use, because how the locking element is tightened strongly influences the result. Each loading direction needs its own group of samples, the fatigue variants several groups at descending amplitudes to locate run-out. The medium is chosen and recorded: dry, saline drip, saline immersion or simulated body fluid. The method does not appear to set a test temperature; the medium, not the temperature, is the environmental variable it requires to be reported.

What the machine must be capable of

Five of the six directions run on a single-axis dynamic frame with the right jig, but the axial torque case is a rotary moment and needs a torsion axis or a biaxial machine.

Nothing in the method fixes a frame rating. Forces are modest and the resolution demands are not: a screw-to-rod joint typically holds from a few hundred newtons to around two kilonewtons, and the moment cases a handful of newton-metres. A dynamic frame of 5 kN or less with a torque axis of roughly 20 to 50 N·m reaches all six directions. The harder requirement is holding a clean, repeatable small-amplitude cycle for two and a half million cycles without drift; an oversized load cell working at the bottom of its range is the usual reason a result cannot be trusted.

Static rates quoted by testing laboratories are of the order of 20 N/s, or about 25 mm/min, for the force cases and about 25 N·m/min, or roughly 25 degrees/min, for the moment cases; fatigue is commonly run at 5 Hz. Those figures reflect laboratory practice rather than rates the method fixes. No public account names the force-verification practice it invokes, so the current edition governs on both counts.

No extensometer is involved; fixture geometry replaces it. Each moment case is defined by a lever arm set by the jig, so a few millimetres of error there scales every reported moment, and the jig must be stiff enough that measured displacement belongs to the joint rather than the rig.

What goes wrong in practice

The most common dispute is over what counts as failure. Interconnections rarely snap; they loosen, and the joint keeps cycling while quietly giving up displacement. Unless a loosening criterion is defined before the run and monitored during it, two laboratories can report very different lives from identical hardware.

Set-screw stripping is next. Over-torquing on assembly, or reusing a locking element already seated once, damages the thread form and shifts the result onto the fastener rather than the connection under test. Amplitude loss follows from it. A joint that has begun to let go absorbs part of every stroke in free movement before it carries anything, so the peak the controller reports is not the peak the interconnection sees, and a run-out logged at 2,500,000 cycles was never earned.

Fretting is the slow one. Under saline the rod and the clamp interface accumulate debris and a dark stain long before any measurable slip, so photographing the interface at intervals is worth more than the load trace.

Related and equivalent standards

F1798 has no true cross-body equivalent. The designation it is most often confused with is ASTM F1717, which fatigues an entire assembled construct in a vertebrectomy model; F1798 does the opposite, stripping the construct away so one interconnection carries everything. ASTM F2193 covers individual components in isolation. All three sit under the same subcommittee and share terminology from ASTM F1582, which is why results are quoted together and misread together — a construct that passes F1717 says nothing about the gripping capacity of its joints.