PSI - Issue 43

Raghu V. Prakash / Procedia Structural Integrity 43 (2023) 190–196 Author name / Structural Integrity Procedia 00 (2022) 000 – 000

194

5

2.2. X-ray Computed Tomography Carbon fiber reinforced polymer composite has been considered for this part of fatigue damage progression study. Quasi-isotropic CFRP laminates typically of 350 mm x 350 mm x 4.5 mm thickness having a stacking sequence of [0/90/±45/0/90/±45] s were prepared by hand lay-up technique using woven carbon fabric of 500 gsm, 0.42 mm thickness with Araldye LY 556 as resin and HY 906 as hardener. Specimens were CNC milled to a finish shape of 250 x 45 x 4.5 mm thickness after the laminate was cured at 80 deg C and a pressure of 1.5 kg/cm 2 for three hours. Fatigue damage in flat specimens could originate all through the gage length; as a consequence, may pose difficulty in tracking fatigue damage in the specimen. To concentrate the focus area, the specimens were impacted with a spherical indenter prior to fatigue testing using a drop weight impactor having a hemispherical indenter of 16 mm diameter. The input impact energy level was pre-set to 51 J. The specimens after impact were subjected to constant amplitude fatigue cycling with a peak stress of 118.5 MPa and a stress ratio of 0.1. The specimen stiffness was estimated based on the load versus actuator displacement. It may be noted that the measurement of stiffness using actuator displacement includes the load train compliance, apart from measuring the local displacement of the specimen at the impact region. However, as the load train compliance is a constant parameter, any changes to specimen stiffness would be reflected in the overall stiffness measurements. To ensure good quality of stiffness measurements, the slope of load versus displacement in the unloading segment over a window of 95-50% of maximum stress was used in stiffness calculations. Figure 9 shows the non-dimensioned stiffness (first cycle stiffness is taken as reference for non-dimensioning) degradation as a function of non dimensioned (w.r.t failure cycle) cycles. Data for un-impacted specimen is also shown in the same figure for comparison of increased stiffness degradation due to drop-impact.

Normalized Stiffness vs No. of Cycles

0 0,2 0,4 0,6 0,8 1 1,2

51J CA UnImpacted

Normalized stiffness, K/Ko

0

0,2

0,4

0,6

0,8

1

Normalized life, N/Nf

Fig. 9 – Graph of non-dimensioned stiffness versus non-dimensioned failure life for 51J impacted CFRP specimen and un-impacted specimen. X-ray Computed Tomography scanning and reconstruction of images was done using GE-Phoenix Tomography system at IIT Madras as well as with North Star Imaging X25 industrial CT X-ray inspection system. Orthographic snap shots of the 3D model from the CT analysis was taken along the Front (F), Top (T) and side (S) directions as shown in Fig. 10 for an un-impacted specimen. Typically 10 slices were extracted along thickness direction (F), while 300 slices for a length of 70 mm at the area of impact (also the area of interest) along the length (T) direction and 100 slices along the breadth (S) direction. The damage area was estimated from the images using ImageJ freeware and damage volume was computed by integrating the damage area across layers for the specimen thickness. It may be noted that image analysis along F direction provides information regarding intra-laminar damage modes, such as matrix cracking, fiber-matrix interfacial de-bonding and fiber breakage; analysis of images along S and T directions provides a picture of inter-laminar damage (delamination) present in terms of length, using which the delamination area can be calculated. Figure 11 presents a typical CT image and its processing using ImageJ software. The damage volume estimation procedure was validated with a CFRP specimen having known