Drilling of carbon fibre reinforced laminates – a study

The distinguishing characteristics of carbon fibre reinforced laminates, like low weight, high strength or stiffness, had resulted in an increase of their use during the last decades. Although parts are normally produced to “near-net” shape, machining operations like drilling are still needed. In result of composites non-homogeneity, this operation can lead to delamination, considered the most serious kind of damage as it can reduce the load carrying capacity of the joint. A proper choice of tool and cutting parameters can reduce delamination substantially. In this work the results obtained with five different tool geometries are compared. Conclusions show that the choice of adequate drill geometry can reduce thrust forces, thus delamination damage. Introduction The use of composite laminates, like carbon fibre reinforced polymers, in complex structures has increased significantly for the last decades. Reasons for this can be found in some unique properties like low weight, high strength and stiffness. Nevertheless, there are still some issues when considering the use of composite laminates. Some of these issues are cost-related, but considerations about machining also lead to some difficulties and lack of acceptance for the implementation of these materials. One of the main machining operations needed for parts assembly in structures is drilling, when it is necessary to join different parts. Generally, drilling can be carried out using conventional machinery, with adaptations. However, this operation is likely to cause several damages in the laminates, namely in the region around the drilled hole, being delamination the most serious as it causes a loss of mechanical and fatigue strength around the drilled hole area [1]. Normally, two kinds of delamination are identified: peel-up delamination and push-out delamination. Other damages are likely to occur, like fibre torn-out or thermal degradation of the matrix [2]. The main mechanism responsible for delamination is the indentation effect caused by the quasistationary drill chisel edge, acting over the uncut plies of the laminate. These plies tend to be pushed away from the plate, causing the separation of two adjacent plies of the laminate [3]. If the thrust force exerted by the drill exceeds the interlaminar fracture toughness of the plies, delamination takes place [4]. The referred delamination is known as “push-out“ delamination. It can be found at the drill exit side of the laminate, and is difficult to avoid. Several approaches had been presented in the last years [3, 5-9]. Generally, it is accepted that delamination can be reduced with an adequate combination of feed rate, cutting speed and drill geometry [5, 10-14]. Apart from that, another type can also be identified when drilling laminate composites, the “peel-up“ delamination. This delamination is a consequence of the drill entrance in the upper plies of the laminate. As the drill moves forward it tends to pull the abraded material along the flute and the material spirals up [4]. Normally, the adoption of low feed rates can avoid this delamination. Another possible option to consider in order to reduce “push-out“ delamination, is the pilot hole drilling. The use of a pilot hole enables a thrust force reduction by dividing the operation in two, thus reducing the indentation effect of the final drill [15, 16]. In this work, laminates with a cross-ply stacking sequence and 4 mm thickness were produced using prepreg carbon/epoxy plies. Experimental drilling tests were carried out on these plates and thrust force was monitored. Five different drill geometries are used for comparison: two twist drills with different point angles (85o and 120o), a Brad drill, a Dagger drill and a special step drill. During drilling, thrust forces were monitored. After drilling, hole wall roughness was measured and “push-out“ delamination extent was determined using enhanced radiography combined with an algorithm of image analyis and processing. This work is concerned with the reduction of this type of delamination. Conclusions from experimental work demonstrate the influence of drill geometry and drilling parameters feed rate or cutting speed in delamination occurrence. Experimental work The experimental work was divided in three steps: drilling of the laminate plates for thrust force monitoring, hole wall roughness measurement, delamination evaluation by enhanced radiography and the use of a numerical criterion, like the Delamination Factor [17], for results comparison. In order to accomplish this work, a batch of plates using prepreg CC160 ET 443 with a cross-ply stacking sequence and 24 layers were produced. The plate was then cured under a pressure of 300 kPa and a temperature of 130 oC for one hour, followed by air cooling. Final plate thickness was 4 mm. The plates were cut in test coupons of 165 * 96 mm for drilling experiments. Drilling operation was carried out in a 3.7 kW DENFORD Triac Centre CNC machine. A total of five different drill geometries, all in K20 tungsten carbide, were used: a twist drill with a point angle of 120o, a twist drill with a point angle of 85o, a Brad drill, a Dagger drill and a special step drill, Fig. 1. All the holes had a diameter of 6 mm a) b) c) d) Figure 1 – Drills used: a) twist (120o); b) Brad; c) Dagger; d) special step. Twist drill is a standard drill commonly used. Two point angles – 85o and 120o – are compared in this work. Brad drill has a specific point geometry causing the fibre tensioning prior to cut thus enabling a “clean cut” of the fibres. In consequence, machined surfaces are smoother. Dagger drill has a small point angle of 30o, reducing the indentation effect but need more space available at the exit side of the plate. The special step drill has the intention of performing pilot and final hole in one operation only, dividing the thrust force and consequently, delamination risk. The helix profile was maintained, in order to compare the cutting performance of this drill with twist or Brad drills. During drilling, axial thrust forces were monitored with a Kistler 9257B dynamometer associated to an amplifier and a computer for data collection. No sacrificial plates were used, Fig. 2a), b). Cutting parameters were selected according to author’s previous experience, published papers from other authors and fabricant recommendations [6, 12, 18]. As it has been already demonstrated the major importance of feed rate when compared with spindle speed in thrust forces development [12], cutting speed was always equal to 53 m/min, corresponding to a spindle speed of 2800 rpm and three levels of feed rate were used – 0.02 (low), 0.06 (medium) and 0.12 mm/rev (high). a) b) Figure 2 – a) Experimental setup adopted; b) aspect of a drilled plate. After drilling, hole wall roughness was measured with a Hommelwerke profilometer, with a cutoff length – λc – of 0.25 mm and an evaluation length – lm – of 1.25 mm. In this work, the roughness parameter considered was Rmax, corresponding to the maximum peak-to-valley dimension obtained from the five sampling lengths within the evaluation length. Three measurements were made for each hole. Finally, plates were inspected by enhanced radiography. With this purpose, plates were prior immersed in di-iodomethane for contrast for one and a half hour. The acquired radiographies were scanned for delamination around the hole measurement and Delamination Factor results. The Delamination Factor (Fd) [17] is a ratio between the maximum delaminated diameter (Dmax) and hole nominal diameter (D),