Mechanical Coupon and Fatigue Testing¶

Structural verification test — instantiates Functional Porosity Design

Destructively loads sampled coupons — monotonic and cyclic — to measure what the porous skeleton can actually bear and how long it survives, exposing how sharply pore-borne defects cut fatigue life.

Every void removed for function is load path removed for free, and no model of a porous solid is trusted until something has been pulled until it breaks. Mechanical Coupon and Fatigue Testing is that reckoning: it takes representative specimens of the porous structure and loads them to measure strength, stiffness, toughness, creep, and — the part that matters most for pores — fatigue life under repeated cycling. Its distinctive contribution is not the single-pull strength number but the durability one: pores are stress concentrators, so a porous part can pass a static test and still fail in service after a few million cycles from a crack that started at a pore wall. The method exists to find that out on a coupon rather than in the field, and to turn "we think the skeleton holds" into measured margins with a known scatter.

Example¶

An aerospace supplier has an additively-manufactured titanium bracket whose weight was cut with an internal lattice. It passes its static proof load easily. Before it flies, the team must know it survives a service life of vibration, so they build a batch of coupons carrying the same lattice and run constant-amplitude axial fatigue per a standard method (ASTM E466), assembling an S–N curve from specimens cycled at descending stress levels. Two findings change the design. First, the fatigue strength sits well below what the static number would suggest — cracks nucleate at the rough, un-melted-powder surfaces inside the struts, exactly where a pore wall meets a stress riser. Second, the lives scatter widely: nominally identical coupons fail across a broad band, because each one's worst internal defect is different and fatigue is governed by the worst, not the average. The team fits the scatter with Weibull statistics, sets an allowable stress at a low survival percentile rather than the mean, and feeds the defect sensitivity back to the build team as a surface-finish requirement.

How it works¶

What distinguishes the method is that it interrogates the skeleton, destructively and statistically:

Sample representatively. Cut or build coupons that carry the real pore architecture and, ideally, the real as-built defects — witness specimens from the same run.
Load across regimes. Monotonic tests for strength, stiffness, and toughness; cyclic tests for fatigue life; hold tests for creep and collapse — whichever the service demands.
Push to fracture and read where it started. Failure location and fractography reveal whether pores, walls, or defects are the initiation site.
Quantify the scatter. Repeat enough specimens to characterize the spread, then set allowables at a survival percentile rather than the mean.

Tuning parameters¶

Load regime and R-ratio — static vs cyclic, and the min/max stress ratio of the cycle; targets the failure mode the service actually sees.
Coupon vs component fidelity — simple standard coupons are cheap and comparable, but a witness carrying real defects and real geometry is more honest; the trade is cost against representativeness.
Sample size — how many specimens; more narrows the confidence on the scatter and lets you quote a lower percentile safely.
Environment — temperature, corrosion, or fluid exposure applied during the test to match service.
Run-out / cycle ceiling — how many cycles counts as "survived"; sets how far into high-cycle fatigue the data reaches.

When it helps, and when it misleads¶

Its strength is that it is ground truth: it measures what the porous skeleton actually does under load, catches the fatigue and creep failures no static check and no idealized model will show, and produces the scatter needed to set a defensible allowable. For any load-bearing porous part, it is the gate the design must clear.

It misleads when the coupon is not the part. Fatigue is defect-driven, so results obey a size effect — a large component has far more stressed volume in which to hide a life-limiting defect than a small coupon, and coupon data can be dangerously optimistic for the full structure.^[1] Testing only the nominal architecture misses the as-built variability that actually fails parts; too few specimens hides the scatter that matters; and the classic misuse is to run a thin test to ratify a design already committed rather than to probe it, quietly choosing the load case the part will pass. The discipline is to test witness specimens that carry real defects, size the sample to the scatter, and set allowables from the distribution's tail rather than its mean.

How it implements the components¶

Mechanical testing fills the structural-verification components — turning a claimed skeleton into a measured one:

load_bearing_skeleton_map — it measures where and how much the solid skeleton actually carries load, and where it fails first.
process_variability_and_defect_band — coupon-to-coupon scatter and fractography quantify how far the as-built defect population moves strength and life.
functional_property_envelope — it certifies whether the structural corner of the required envelope (strength, stiffness, fatigue life) is actually met.

It measures rather than makes: it does not set the property envelope's targets (that is Topology Optimization for Void Placement) or create the voids (the fabrication siblings); pore-size and surface metrology is Multi-Method Porometry, and transport and storage testing is Transport, Storage, and Breakthrough Testing.

Instantiates: Functional Porosity Design — the structural gate that proves the void architecture kept the strength it had to preserve.
Consumes: the as-built structure produced by the fabrication mechanisms — Additive Lattice or Gyroid Fabrication, Gas Foaming or Blowing, Graded-Density Manufacturing, Particle Packing and Sintering Control.
Sibling mechanisms: Multi-Method Porometry · Topology Optimization for Void Placement · Transport, Storage, and Breakthrough Testing · Porosity Statistical Process Control

Notes¶

A coupon is a witness, not the part. Because fatigue life is set by the single worst defect in the stressed volume, data from small specimens can overstate the durability of a large component that has far more volume to hide one — the size effect must be applied before coupon allowables are trusted structurally. Pair the scatter this test measures with the production monitoring of Porosity Statistical Process Control: the test tells you how sensitive life is to defects; only ongoing process control keeps the defect band inside what was actually tested.

References¶

[1] Brittle- and defect-driven strength follows a size effect captured by Weibull statistics: the larger the stressed volume, the higher the probability it contains a critical flaw, so measured strength falls with specimen size. Coupon allowables must therefore be scaled to the component's stressed volume rather than applied directly, or they flatter the part. ↩