Fatigue testing

A variety of coupons can be used but some of the common ones are:

• compact tension coupon (CT). The compact specimen uses the least amount of material for a specimen that is used to measure crack growth.[4] Compact tension specimens typically use pins that are slightly smaller than the holes in the coupon to apply the loads. This method however prevents the accurate application of loads near zero and the coupon is therefore not recommended when negative loads need to be applied.[4]
• Centre Cracked Tension panel (CCT). The centre cracked tension or middle tension specimen is made from a flat sheet or bar containing two holes for attaching to grips .
• Single Edge Notch Tension coupon (SENT).[5] The single edge coupon is an elongated version of the compact tension coupon.

The following instrumentation is commonly used for monitoring coupon tests:

• Strain gauges are used to monitor the applied loading or stress fields around the crack tip. They may be placed beneath the path of the crack or on the back face of a compact tension coupon.[6]
• An extensometer or displacement gauge can be used to measure the crack tip opening displacement at the mouth of a crack. This value can be used to determine the stress intensity factor which will change with the length of the crack. Displacement gauges can also be used to measure the compliance of a coupon and the position during the loading cycle when contact between the opposite crack faces occurs in order to measure crack closure.
• Applied test loads are usually monitored on the test machine with a load cell.
• A travelling optical microscope can be use for measurement of the position of the crack tip.

Certification requires knowing and accounting for the complete load history that has been experienced by a test article. Using test articles that have previously been used for static proof testing have caused problems where overloads have been applied and that can retard the rate of fatigue crack growth.

The test loads are typically recorded using a data acquisition system acquiring data from possibly thousands of inputs from instrumentation installed on the test article, including: strain gages, pressure gauges, load cells, LVDTs, etc.

Fatigue cracks typically initiate from high stress regions such as stress concentrations or material and manufacturing defects. It is important that the test article is representative of all of these features.

Cracks may initiate from the following sources:

• Fretting, typically from high cycle count dynamic loads.
• Mis-drilled holes or incorrectly sized holes for interference fit fasteners.[7]
• Material treatment and defects such as broken inclusions.[8]
• Stress concentrations such as holes and fillets.
• Scratches, impact damage.

A representative block of loading is applied repeatedly until the safe life of the structure has been demonstrated or failures occur which need to be repaired. The size of the sequence is chosen so that the maximum loads which may cause retardation effects are applied sufficiently often, typically at least ten times throughout the test, so that there are no sequence effects.[9]

The loading sequence is generally filtered to eliminate applying small non-fatigue damaging cycles that would take too long to apply. Two types of filtering are typically used:

1. deadband filtering eliminates small cycles that completely fall within a certain range such as +/-3g.
2. rise-fall filtering eliminates small cycles that are less than a certain range such as 1g.

The testing rate of large structures is typically limited to a few Hz and needs to avoid the resonance frequency of the structure.[10]

Fatigue test rig at the Wright-Patterson Air Force Base

All components that are not part of the test article or instrumentation are termed the test rig. The following components are typically found in full scale fatigue tests:

• Whiffletrees. In order to apply the correct loads to various parts of the structure, a mechanism known as a whiffletree is used to distribute the loads from a loading actuator to the test article. Loads applied to a central point are distributed through a series of pin jointed connected beams to produce known loads at the end connections. Each end connection is typically attached to a pad which is bonded onto the structure such as an aircraft wing. Hundreds of pads are usually applied to reproduce the aerodynamic and inertial loads seen on wing. Because the whiffletree consists of tension linkages, they are unable to apply compressive loads and therefore, independent whiffletrees are typically used on the upper and lower sides of wing fatigue tests.
• Hydraulic, electromagnetic or pneumatic actuators are used to apply loads to the structure, either directly or through the use of a whiffletree to distribute the loads. A load cell is placed inline with the actuator and is used by the load controller to control the loads into the actuator. When many actuators are used on a flexible test structure, there may be cross-interaction between the different actuators. The load controller must ensure that spurious loading cycles are not applied to the structure as a result of this interaction.
• Reaction restraints. Many of the loads such as aerodynamic and internal forces are re-acted by internal forces which are not present during a fatigue test. Hence, the loads are reacted out of the structure at non-critical points such as the undercarriage or through restraints on the fuselage.
• Linear variable differential transformer can be used to measure the displacement of critical locations on the structure. Limits on these displacements can be used to signal when a structure has failed and to automatically shut down the test.
• Non-representative structure. Some test structure may be expensive or unavailable and are typically replaced on the test structure with an equivalent structure. Structure that is close to actuator attachment points may see an unrealistic load that makes these areas non-representative.

The following instrumentation is typically used on a fatigue test:

• strain gauges
• accelerometers
• displacement gauges
• load cells
• crack sensor
• structural health monitoring sensors

It is important to install any strain gauges on the test article that are also used for monitoring fleet aircraft. This allows the same damage calculations to be performed on the test article that are used to track the fatigue life of fleet aircraft. This is the primary way of ensuring fleet aircraft do not exceed the safe-life determined from the fatigue test.

Inspections form a component of a fatigue test. It is important to know when a detectable crack occurs in order to determine the certified life of each component in addition to minimising the damage to surrounding structure and to develop repairs that have minimal impact on the certification of the adjacent structure. Non-destructive inspections may be carried out during testing and destructive tests may be used at the end of testing to ensure the structure retains its load carrying capacity.

Test interpretation and certification involves using the results from the fatigue test to justify the safe life and operation of an item.[11] The purpose of certification is to ensure the probability of failure in service is acceptably small. The following factors may need to be considered:

• number of tests
• symmetry of the test structure and the applied loading
• installation and certification of repairs
• scatter factors
• material and manufacturing process variability
• environment
• criticality

Airworthy standards typically require that an aircraft remains safe even with the structure in a degraded state due to the presence of fatigue cracking.[12]