to select a stacking sequence that promotes a non-linear mechan- ical behavior in the presence of progressive damages [34]。 Sec- ondly, the results of some preliminary fatigue tests conducted on specimens with Unidirectional (UD) or quasi-isotropic stacking se- quence did not show any dissipation。 Thirdly, the mechanical behavior of these specimens was almost linear and the type of frac- ture observed was brittle。

In the current experiment, two sets of specimens were used: the first set of specimens is composed of four samples with AWJM holes having diameters of 6 mm, while the second set is composed of four plates with holes machined with a conventional drill bit。  The dimensions of the specimens are 270 mm long, 45 mm wide and 2。1 mm thick。 All holes were situated at the center of the plates。 The machining of the holes was carried out on a numerically controlled machine。 For the CM technique, carbide drills with two lips have been used to machine the holes。 The spindle speed of 2020 rpm and feed rate of 0。1 mm/rev were used。 Two abrasive

sizes of 120 and 220 lm along with a jet pressure of 145 MPa were

used for AWJM。 The standoff distance of 4 mm and water jet inci- dence angle of 90° were kept constant during the drilling process。 Note that all drilled specimens were roughly identical from the mechanical properties viewpoint。

2。2。Drilling quality

In order to investigate the quality of the machined surface and the machined surface texture, a NANOVEA 400 series profilometer and SEM observation were used。 For the profilometer tests, a cut- off and transverse length of 1 and 2 mm respectively along x and y-axis were used。 The average surface roughness (Ra), maximum profile valley depth (Rv) and skewness (Rsk) were measured according to ISO 4287/1 standards using NANOVEA 3D software。

2。3。Mechanical testing

We intended to measure the fatigue at different maximum  stress levels ranging from 17% to 65% of ultimate tensile strength (UTS)。 Therefore, static tensile tests were performed on three sam- ples following the ASTM D 3039 standard to determine the ulti- mate tensile strength (UTS)。 The tests were realized with an Instron model 4206 electromechanical testing machine, equipped with a 150 kN load cell。 The crosshead speed of the testing machine was 1 mm/min for static tests in tension。 The same specimen stacking sequence and dimensions was used for both static and fa- tigue tests。 An extensometer was used to monitor the strain to investigate stress–strain behavior of the specimen in static condi- tion。 A FLIR SC5000 infrared camera with a pixel resolution of   320 × 240 and a temperature sensitivity of <20 mK was used to monitor the specimen surface temperature (Fig。 1)。 The fatigue tests were conducted at room temperature using an MTS 322 tester equipped with hydraulically operated wedge grips。 An extensome- ter was used to monitor the local strain that allowed for the calcu- lation of the stiffness degradation of the specimen during the cyclic loading。 These axial tension–tension fatigue tests were conducted in load control using a constant amplitude sinusoidal waveform,     a loading frequency of 10 Hz and a stress ratio of 0。1, at various maximum applied stress levels。 From the thermographic infrared (IR) temperature plots it was possible to evaluate the damage evo- lution and endurance limit of the specimens [35] (Fig。   1)。

Fatigue tests were conducted for progressive step cyclic loads ranging from 17–65% of UTS, each for 5000 cycles。 An extensome- ter was used to monitor the local strain that allowed for the calcu- lation of the stiffness degradation of the specimen during the cyclic loading。

The IR camera (model FLIR Silver 420) carried out thermal mea- surements with sensitivity, precision, and speed。 The camera   was

M。 Saleem et al。 / Composites: Part A 55 (2013) 169–177