ASTM F606/F606M

What the test does

Most of the work is done on the fastener as manufactured, not on a coupon cut from it. A bolt is engaged by its threads in a hardened fixture or a nut and drawn along its axis to destruction; in the wedge version a hardened block of specified angle sits under the head, loading it out of square with the shank. The proof-load test stops short of failure: the fastener is measured for overall length, loaded once to a specified load, released and measured again. Nuts are run onto a hardened mandrel or cone and loaded until the proof load is reached or the threads strip. Hardness and rotational-capacity tests are covered too, the latter on a bolt tension calibrator.

What it measures, and why it matters

Proof load is a pass or fail on elastic behaviour: a fastener that returns to its original length has not yielded at the load its joint will hold it at — the basis of preload design. Wedge tension gives breaking strength while testing the head-to-shank junction: a bolt that passes in straight tension but breaks under the head on the wedge has a fault an axial pull cannot see — a bad fillet, a quench crack, the wrong heat treatment. Nut and cone proof load confirm a nut can develop its bolt's strength, so an overloaded joint fails in the bolt rather than by stripping. Hardness screens heat treatment, and a fastener hardened beyond its grade's range is generally regarded as more susceptible to hydrogen embrittlement even when its tensile figures pass.

Specimen

The specimen is normally the product itself, tested full size and as supplied, coating included, since it affects both strength and friction. There is no gauge length in the usual sense: for proof loading the overall length is measured before and after, and the difference is what matters. Machined specimens, used where the product is too large to pull full size, follow the geometry and gauge lengths of Test Methods E8/E8M. Lot sampling is a matter for the product specification, and no atmospheric conditioning is called for.

What the machine must be capable of

Capacity follows the product, not the method: an M12 property class 8.8 bolt fails near 65 kN and a 1 1/2 in. ASTM A490 bolt approaches 1 000 kN, so fastener laboratories run frames from roughly 300 kN for small sizes to 1 500 kN for large structural product. Practice E4 governs the force verification.

The method's own rate of testing is not quoted here, because no public source gives a numeric crosshead or stress rate for F606/F606M. The nearest published limits — 25 mm/min, and 10 mm/min where yield is determined — belong to the ISO counterpart, not to F606.

Extensometry is unsettled too: Practice E83 sits among the referenced documents, but which class F606/F606M demands is not publicly stated. Machined-specimen yield work inherits E8/E8M's Class B-2 in practice, while proof loading by length measurement needs a length gauge, not an extensometer.

Fixturing decides whether a result means anything. Threaded holders, nut and stud fixtures and hardened wedge blocks are all needed, and the wedge angle is not a free choice: 10 degrees is usual, reduced for heat-treated bolts threaded close to the underside of the head, and reduced again for embrittlement testing. The angle for a given diameter is tabulated, and only the current table settles which one applies. Use the standard angle on a bolt that calls for a reduced one and it fails under the head for reasons that have nothing to do with the product. Testing is in ambient air; no temperature or humidity band is set.

What goes wrong in practice

The failure that wastes most time is the fixture giving way instead of the fastener. Holder threads wear or soften with use, engagement is too short for the grade, and the assembly strips before the bolt breaks; the load recorded is then a property of the fixture.

Galling hits coated and stainless product hardest. A seized thread ruins a rotational-capacity result outright, and in tension it can carry load in a way that flatters the fastener.

Proof loading has a trap of its own: the permanent set looked for is very small, so measurement resolution and repeatable seating points dominate. Published sources are not consistent about the resolution required, so the laboratory's own procedure has to settle it.

Related and equivalent standards

ISO 898-1 is the nearest counterpart for externally threaded steel fasteners, but the two systems are organised differently. F606/F606M is a methods document and sets no acceptance values of its own; the numbers a fastener must meet live in the product specification that invokes it. Citing it as though it imposed requirements is a common error.

Despite the torque applied in rotational capacity, it is neither a torsion nor a fatigue method: fastener endurance and behaviour under repeated tightening are outside its scope, and hardness and machined-specimen tension are read across from E10, E18 and E8/E8M.