You are on page 1of 8

Composites Science and Technology 71 (2011) 17191726

Contents lists available at SciVerse ScienceDirect

Composites Science and Technology


journal homepage: www.elsevier.com/locate/compscitech

Occurrence and propagation of delamination during the machining of carbon bre reinforced plastics (CFRPs) An experimental study
Wolfgang Hintze , Dirk Hartmann, Christoph Schtte
Institute of Production Management and Technology, Hamburg University of Technology, 21073 Hamburg, Germany

a r t i c l e

i n f o

a b s t r a c t
The machining of carbon bre reinforced plastics (CFRPs) is often accompanied by delamination of the top layers of the machined edges. Such damage necessitates time-consuming and costly post-machining and in some cases leads to rejection of components. The work described in this paper systematically investigates the occurrence of delamination of the top layers during the machining of CFRP tape, with the focus being on the process of contour milling. The occurrence and propagation of delamination were studied by milling slots in unidirectional CFRP specimens having different bre orientations and mainly analysing the slot tip. This allowed the key mechanisms to be claried. The results show that delamination is highly dependent on the bre orientation and the tool sharpness. The experiments allow derivation of a novel system for describing the occurrence and propagation of delamination during milling. Furthermore, the principles also apply for drilling. The results allow customisation of the machining procedure to reduce and in some cases totally avoid delamination, leading to a signicant increase in the quality of components. 2011 Elsevier Ltd. All rights reserved.

Article history: Received 15 March 2011 Received in revised form 27 July 2011 Accepted 9 August 2011 Available online 22 August 2011 Keywords: A. Carbon bres A. Laminate B. Delamination B. Machining C. Damage mechanics

1. Introduction During the manufacture of components from carbon bre reinforced plastic (CFRP), it is usually necessary to carry out a postmachining step after curing in order to meet the required tolerances and to manufacture tting and joining surfaces. Classical production processes such as milling and drilling are mainly used for this. Regardless of the production process, damage in the form of delamination can occur during the processing of CFRPs. This characteristic production defect in CFRPs is particularly prevalent in the top layers of the laminate as these are only supported on one side. Top and bottom plies of a multiaxial CFRP tape are even more critical than those of woven fabric, where crossing bres are mutually supported. The bres are cut by the tool in an undened way, deect under the action of the cutting edge and consequently delamination occurs in the form of bre overhang and bre breakout at the cut edges. Such damage must be absolutely avoided, because this requires time-consuming and costly post-machining to rectify and in some cases leads to rejection of components. Although delamination during the drilling of CFRPs has already been the subject of numerous publications [1,2,68,10,13,14,16,

Corresponding author. Tel.: +49 40 42878 3051; fax: +49 40 42878 2295.
E-mail addresses: w.hintze@tu-harburg.de (W. Hintze), d.hartmann@tu-har burg.de (D. Hartmann), christoph.schuette@tu-harburg.de (C. Schtte). 0266-3538/$ - see front matter 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.compscitech.2011.08.002

2123,2527,29,30], little attention has been put on milling processes up until now. Colligan and Ramulu [3,4] characterised typical forms of delamination during contour milling and developed a classication system. They also established that delamination during machining mainly results from poor support of the top laminate layers. Rummenhller [24] and Hohensee [15] identied tool sharpness as a decisive factor for the cutting process. They also observed that bres, in particular in the top layers of the laminate, delaminate. Davim [5] analysed the effect of the cutting parameters on the component quality when milling CFRPs. They came to the conclusion that increasing the feed per tooth leads to increased damage in the form of delamination, although the authors gave no reason for this. The mentioned study highlighted the signicance of bre orientation as an important factor when machining CFRPs. Zhang et. al. [30] investigated damage that occurs when drilling CFRP components. They found that delamination depends on the angle between the cutting direction and bre direction. However, the authors did not show there to be a systematic relationship between this angle and the delaminated regions. For contour milling, Hohensee [15] called this angle the bre cutting angle h (Fig. 1b). In contrast to the bre orientation angle U (Fig. 1a), which is measured between the feed direction and the bre orientation, the bre cutting angle changes continuously during the engagement of the cutting edge (Fig. 1c). The reason for this is the rotary cutting motion of the tool.

1720

W. Hintze et al. / Composites Science and Technology 71 (2011) 17191726

Fig. 1. Difference between the bre orientation angle and the bre cutting angle.

A major reason for the occurrence of delamination when machining FRPs is tool wear [4,6,9,15]. Due to the increasing cutting edge radius rn, which is the main type of wear in CFRP machining [7,15,18,20], the machining forces increase and dened cutting of the bres becomes more difcult [11]. With regard to the drilling of CFRP fabric, Faraz et al. [7] recently characterised the extend of delamination by the ratio of the delaminated area around the drilled hole and the nominal area of the drilled hole. They found correlations between this parameter and the maximum thrust force for each type of drill, which do not indicate a unique thrust force limit for delamination occurrence as stated in [13] but a strong inuence of the drill geometry. From drilling with twist drills it is well known, that delamination at the exit side of the workpiece is initiated already by the chisel edge [13,14,21,23]. Subsequently, during exit of the major cutting edges delamination propagates until the cutting edge corners have left the workpiece. In edge milling, such history of delamination propagation of the top layers is not considered in the scientic literature up to now [15,17,19,24,28]. These studies mostly deal with damage of the inner laminate plies at the machined surface. This experimental work focuses on delamination of the laminate top layers in edge milling. It highlights for the rst time delamination propagation effects, which are essential for the understanding of frequently observed damage at nished workpiece edges. For that, the motion of the cutting edge along the whole engagement angle has to be considered. The work reported mainly investigates the milling of CFRPs as a function of increasing tool wear. Using new and worn tools, the effect of bre orientation on the delamination is studied. In contrast to previous studies, the interplay between wear and bre orientation on the delamination is systematically investigated and is represented in a novel system for describing delamination. The knowledge that has been gained has high practical relevance as it allows reduction or avoidance of delamination despite progressive tool wear when trimming CFRP components by contour milling. 2. Materials and methods In order to evaluate the delamination, slots were milled in unidirectionally reinforced CFRP specimens (prepreg, HT bre,

65% bre volume fraction, epoxy resin Cycom 977-2). This procedure provided information about the location of delamination and also about the extent of the delamination because the slot end remained in tact. By deliberately aligning the bres at U = 0, 45, 90 and 135 to the feed direction (Fig. 1), the effect of bre orientation was systematically studied. The machining tests were carried out using a milling machine made by Rders (Rders RFM 600) with a spindle power of 10 kW and a maximum speed of 42,000 rpm. For carrying out the tests, a cutting speed (vc) of 800 m/min and a feed (f) of 0.06 mm were chosen. The work pieces were secured with clamping claws. Typical double-edged PCD end mills (diameter = 12 mm, angle of twist = 0) were used. A coolant was not used. Fig. 2 shows the experimental set-up. As tool wear has a major effect on delamination, tools were used in the new state and in a state of dened wear. By trimming the CFRP test specimen over a longer feed path, a dened cutting edge radius of rn = 45 lm and rn = 90 lm could be set at which

Fig. 2. Experimental set-up for slot milling.

W. Hintze et al. / Composites Science and Technology 71 (2011) 17191726

1721

Fig. 3. Delamination as a function of tool wear.

delamination occurs under the selected cutting parameters. The cutting edge radius rn was measured with an optical 3D measuring unit made by Alicona (Alicona Innite Focus). The wear measurements were repeated ve times. The drilling tests were performed using the same CNC-machine, CFRP tape and a twist drill with a tip angle of r = 85.

3. Results and discussion 3.1. Identication of the critical cutting angle range Knowledge from research studies and practical experience indicates that tool wear is a key factor for the occurrence of delamination. Fig. 3 highlights this effect. Using a sharp tool (rn = 9 lm), the bres can be cut cleanly and no delamination occurs. Using a tool having average wear (rn = 45 lm), there are small bre overhangs. Using a tool having high wear, namely rn = 90 lm, very marked delamination in the form of bre overhangs and bre breakouts occurs on the top layers of the cut edges. When slot milling different bre orientations using a very worn tool (rn = 90 lm), as can occur in practice, it was found that delamination only appears in certain regions on the machined edges and/ or in the slot tip (Fig. 4). Fibre overhangs indicate that the delamination not only occurs on the machined edge, but rather beforehand in the region of the slot tip. The shape and length of the delamination vary considerably depending on the milled bre orientation. Furthermore, delamination does not occur in the entire slot tip, but only in the regions that are marked red1 (Fig. 4, bottom). The rest of the machined edge is free of delamination despite the very high tool wear. For a bre orientation of U = 0, there is only delamination in the form of long bre overhangs in the region of the down-milled slot tip. These bre overhangs extend far into the machined slot, but the edges of the component are free of delamination. When milling a bre orientation of U = 45, delamination on the up-milled edge extends far into the machined slot, whilst the down-milled edge shows no delamination. For a bre orientation of U = 90 there is marked delamination on the up-milled edge. The maximum length of the overhanging bres corresponds to half the slot width. The down-milled edge once again shows no delam1 For interpretation of color in Fig. 4, the reader is referred to the web version of this article.

ination. At a bre orientation of U = 135 both machined edges show delamination in the form of short bre overhangs, whilst the slot tip is free of delamination. It can be concluded from these experiments that top layer delamination does not primarily depend on the bre orientation angle but rather on the bre cutting angle h. Considering the bre orientation of U = 90 in Fig. 4 it becomes clear that delamination occurs in a critical bre cutting angle range of 90 6 h < 180. At the same time, there are regions where there is absolutely no delamination. This nding also holds for the other bre orientation angles that were considered. At U = 0 delamination also appears in the slot tip in the down-milled region for bre cutting angles of 90 6 h 6 180. At U = 135 delamination is present on both edges of the component. This involves cutting angles of 135 6 h 6 180 for the up-milled edge and cutting angles of 90 6 h 6 135 for the down-milled edge. As a result, delamination when milling CFRPs only occurs within a critical bre cutting angle range of 90 6 h 6 180. Only at a bre orientation of U = 45 is delamination also found outside the critical range on the up-milled edge. 3.2. Occurrence and propagation of delamination In order to investigate the mechanisms responsible for the occurrence and propagation of delamination, up-milling was undertaken at an orientation of U = 45 with variable working engagement using a very worn tool (rn = 90 lm) (Fig. 5). When the working engagement ae corresponds to the tool diameter (ae/ d = 1), delamination occurs on the up-milled edge despite the fact that this region lies outside the critical bre cutting angle range. If ae is chosen such that the bres are only cut at a bre cutting angle h < 90 (namely ae/d 6 0.13), then no delamination occurs. Increasing the working engagement ae beyond a bre cutting angle of h = 90 leads to delamination with bre overhangs which depend on the working engagement ae. Important for the occurrence and propagation of delamination here is the chronological order of the bre cutting angle h the cutting edge passed through. Fig. 6 shows, for a selected bre, the change in the bre cutting angle with increasing feed path lf. With the exception of a bre orientation of U = 0, the cutting edge contacts the bre for the rst time at a bre cutting angle of 0 and 180, providing the working engagement corresponds to the tool diameter (ae = d). Based on rst contact at h = 180, the delamination only propagates during the further feed motion if the subsequent cutting an-

1722

W. Hintze et al. / Composites Science and Technology 71 (2011) 17191726

Fig. 4. Delamination when milling using a worn tool (rn = 90 lm).

Fig. 5. Delamination at different working engagement ae for U = 45.

gle lies in the critical range, namely the critical cutting angle range is passed through in reverse (h = 180 to h = 90). Fig. 7 claries this situation using the example of a bre orientation of U = 45. The critical cutting angle range is travelled from the rst contact point at h = 180 backwards to h = 90. Delamination can only occur on the edge machined by up-milling, because the bres there are continuously cut at a critical angle h. Noteworthy

here is that the delamination due to the tool feed motion also propagates beyond the critical cutting angle range in the direction of angles smaller than h = 90. When the working engagement ae corresponds to the tool diameter (ae = d), then the length of the bre overhangs in the region of the slot tip corresponds to the distance of the up-milled edge to the point at which the cutting edge cuts the bres for the rst time

W. Hintze et al. / Composites Science and Technology 71 (2011) 17191726

1723

Fig. 6. Change in the bre cutting angle as a function of the feed path lf.

Fig. 7. Delamination when slot milling at a bre orientation of U = 45.

1724

W. Hintze et al. / Composites Science and Technology 71 (2011) 17191726

Fig. 8. Delamination when slot milling at a bre orientation of U = 135.

Fig. 9. Systematic scheme for describing the occurrence of delamination in milling.

below a bre cutting angle of h = 180 (see Fig. 7). This correlation holds for all bre orientations U. This effect is generally not evident on trimmed component edges, because some of the overhanging bres are broken off or shortened due to subsequent multiple contact of the cutting edge. At a bre orientation U = 135, delamination occurs on both the up-milled and down-milled edges of the component because the bre cutting angle assumes the critical value of h = 135 on both component edges (see Fig. 6). Fig. 8 shows that the critical cutting

angle range on the two component edges is equally large (h = 135 180, and h = 90135) and hence the bre overhangs on the two component edges are equally long. With further increasing bre orientation angle (U > 135) the length of the bre overhangs increases on the down-milled edge and decreases on the up-milled edge, until at U = 180 delamination only occurs in the down-milled region. Propagation of the delamination beyond a bre cutting angle of h = 0/h = 180 into the non-critical range was not observed in the experiments, and

W. Hintze et al. / Composites Science and Technology 71 (2011) 17191726 Table 1 Active force in planning of UD-CRFP-laminate, vc = 10 m/min, f = 0.03 mm, ap = 4 mm, cf = 0, af = 12, PCD [11]. Cutting edge radius rn Fibre cutting angle h 0 8 lm 70 lm 152 N 357 N 65 112 N 173 N 90 247 N 350 N 135 169 N 538 N

1725

consequently no delamination took place on the down-milled edge for bre orientations of 0 6 U 6 90. Independent of the working engagement ae and the bre orientation angle U delamination of the top layers occurred as soon as the cutting edge contacted the bres for the rst time within the critical cutting angle range. The length of the bre overhang corresponds to the distance of the milled edge to the rst contact point of the bres and cutting edge in the critical cutting angle range.

edge is in the 0 6 h 6 90 range and down-milling must be undertaken. The ndings reported in this study are based on experimental observations. They correspond to former force measurements conducted in planning of this unidirectional laminate [11]. These re2 0.5 sults clearly indicate the increase of the active force (F 2 c +F f ) with the cutting edge radius rn and a higher active force level in the so called critical bre cutting angle range, see Table 1. Of course, the stress exerted on the workpiece surface during machining must be considered in order to assess delamination rather than only the active force. This is realised by calculating the stress distribution and taking into account the static failure modes for orthotropic materials. The basic concept is reported in [12]. 3.4. Delamination when drilling The proposed system for describing delamination when milling can also be used to describe the situation when drilling CFRP tape with spiral drills (point angle of the drill r  180). Even when the feed per tooth fz is perpendicular to the laminate surface there is a radial feed component fr = fz tan(r/2) which at a point angle of about a half of r = 90 corresponds to the feed per tooth fz and due to the conical geometry of the drill bit is tangential to the laminate surface. Accordingly, there is a signicant radial force component that acts in addition to the axial feed force and cutting force and corresponds to the passive force. Such passive forces act on each of both major cutting edges. They are the reason why similar delamination processes are observed when drilling as are when milling. Because of their opposite direction the passive forces are balanced and therefore difcult to measure during drilling. Fig. 10 shows the tool exit side of a CFRP specimen (unidirectional tape). As in milling, the individual bres are rst of all cut at a cutting angle of h = 0 or h = 180 (Fig. 10b). Subsequently, non-critical cutting angles of 0 < h 6 90 are travelled through in two opposite quadrants of the hole and no delamination results. In the other two quadrants, critical cutting angles of 90 6 h < 180 are travelled through and delamination occurs (see Fig. 10a). To explain the situation, consider the bre marked by a full black line in Fig. 10b). This bre is cut in either a cutting angle range of 0 < h 6 70 or in a range of 180 > h P 110. In contrast,

3.3. Systematic scheme for describing the occurrence of delamination The knowledge we have acquired allows a systematic scheme to be drawn up to describe the occurrence of delamination. Fig. 9 shows this scheme for typical bre orientations U. Region A refers to the critical bre cutting angle range in which delamination occurs. Region B is where the described propagation of the delamination occurs. Region C is the cutting angle range where there is no propagation of delamination. On the one hand this concerns the up-milled side of the U = 0 orientation. When the bres are orientated at 0 (i.e. parallel) to the feed direction, each individual bre is cut at the same cutting angle h during the whole machining process, and no propagation can occur. On the other hand, no delamination occurs in the slot tip at a bre orientation of U = 135. This is due to the fact that the bres there are exclusively cut in the non-critical range 0 6 h 6 90 and only with increasing feed path near the milled edge in the critical cutting angle range. Different to the situation at U = 45, propagation whereby the bres are rst cut in the critical range is hence not possible. For delamination-free machining, the process must be carried out such that the bre cutting angle h on the relevant component

Fig. 10. Delamination at the drill exit when drilling CFRP.

1726

W. Hintze et al. / Composites Science and Technology 71 (2011) 17191726 [3] Colligan K, Ramulu M. Delamination in surface plies of graphite/epoxy caused by the edge trimming process. Process manufacturing of composite materials, PED-vol. 49/MD-vol. 27. New York, NY: ASME; 1997. p. 11325. [4] Colligan K, Ramulu M. The effect of edge trimming on composite surface plies. Manuf Rev 1992;5(4):27483. [5] Davim JP. Damage and dimensional precision in milling. J Mater Process Technol 2005;16:1607. [6] Duro LMP, Magalhes AG, Tavares JMRS, Torres Marques A. Delamination analysis after carbon/epoxy plate drilling. Instituto Superior de Engenharia do Porto; 2006. [7] Faraz A, Biermann D, Weinert K. Cutting edge rounding: an innovative tool wear criterion in drilling CFRP composite laminates. Int J Mach Tools Manuf 2009;49:118596. [8] Ghidossi P. Contribution ltude de lEffet des Conditions dUsinage dprouvettes en Composites Matrice polymre sur leur Rponse Mcanique, Thse Docteur, Centre de Chlons en Champagne; 2003. [9] Ghidossi P, Mansori ME, Pierron F. Edge machining effects on the failure of polymer matrix composite coupons. Compos Part A 2004;35:98999. Usinage et la Tenue en Matage [10] Gohorianu G. Interaction entre les Defauts d Assemblages Boulonnes en Carbone/Epoxy, Thse Docteur, Toulouse; 2008. d [11] Hintze W, Clausen R, Hartmann D, Kindler J, Santos S, Schwerdt M, et al. Precision of machined CFRP the challenge of dimensional accuracy. In: Proceedings of the 1st international workshop on aircraft system technologies (AST 2007), March 2930, 2007, Hamburg, Germany Shaker Verlag Aachen; 2007, pp. 36174. [12] Hintze W, Dose F, Hartmann D, Schtte C. Assessment of surface integrity of machined CFRP Methods and Results. In: Proceedings of the 5th international conference Supply on the wings, Airtec; 2010, 02 04.11.2010, Frankfurt (Main). [13] Hocheng H, Tsao CC. Comprehensive analysis of delamination in drilling of composite materials with various drill bits. J Mater Process Technol 2003;140:3359. [14] Hocheng H, Tsao CC. The path towards delamination-free drilling of composite materials. J Mater Process Technol 2005;167:25164. [15] Hohensee V. Umribearbeitung faserverstrkter Kunststoffe durch Frsen und Laserschneiden, Dissertation, Hannover, Univ.; 1992. [16] Iliescu, D. Approches experimentale et numerique de lUsinage a sec des composites carbone/epoxy, Thse Docteur, Bordeaux; 2008. [17] Iliescu D, Gehin D, Nouari M, Girot F. Damage modes of the aeronautic multidirectional carbon/epoxy composite T300/914 in machining. Int J Mater Prod Technol 2008;32(2/3). [18] Iliescu D, Fernandez A, Gutierrez-Orrantia ME, Lopez de Lacalle LN, Girot F. Modeling and tool wear in routing of CFRP. In: International conference on advances in materials and processing technologies (APMT 2010), vol. 1315. AIP conference proceedings; 2011. p. 164550. [19] Kaneeda T. CFRP cutting mechanism. Trans North American Manuf Res Instit SME 1991;19:21621. [20] Kindler J. Werkstckqualitt und Standzeitoptimierung von Zerspanwerkzeugen bei der Umrissbearbeitung von kohlenstofffaserverstrkten Kunststoffen, Dissertation, Hamburg, TUHH; 2009. [21] Lachaud F, Piquet R, Collombet F, Surcin L. Drilling of composite structures. Compos Struct 2001;52:5116. [22] DiPaolo G, Kapoor SG, DeVor RE. An experimental investigation of the crack growth phenomenon for drilling of ber-reinforced composite materials. J Eng Indus 1996;118. ASME. [23] Rahme P, Landon Y, Lagarrigue P, Lachaud F, Piquet R, Marguet B, et al. Chisel edge effect on delamination when drilling thick composite materials with a twist drill. SAE Int J Aerospace 2008;1(1). [24] Rummenhller S. Werkstofforientierte Prozessauslegung fr das Frsen Kohlenstoff-faserverstrkter Kunststoffe, Dissertation, Aachen, TH; 1996. [25] Tsao CC, Chen W-C. Prediction of the location of delamination in the drilling of composite laminates. J Mater Process Technol 1997;70:1859. [26] Tsao CC, Hocheng H. Effect of tool wear on delamination in drilling composite materials. Int J Mech Sci 2007;49:9838. [27] Vijayaraghavan A, Dornfeld D, Dharan CKH. Quantifying edge defects in drilled FRP composites, Laboratory for manufacturing and sustainability, consortium on deburring and edge nishing, Berkeley: University of California; 2006. [28] Wang DH, Ramulu M, Arola D. Orthogonal cutting mechanisms of graphite/ epoxy composite. Part I: unidirectional laminate. Int J Mach Tools Manuf 1995;35(12):162338. [29] Won MS, Dharan CKH. Drilling of aramid and carbon ber polymer composites. J Manuf Sci Eng 2002;124. ASME. [30] Zhang H, Chen W, Chen D, Zhang L. Assessment of the exit defects in carbon bre-reinforced plastic plates caused by drilling. Key Eng Mater 2001;196:4352.

the bre marked by a dashed black line, which runs through the centre of the hole, has a constant cutting angle of h = 90. Propagation of the delamination into the non-critical cutting angle range, as occurs when milling at a bre orientation of U = 45, is therefore not possible. Hence, the depicted delamination pattern in Fig. 10a, which is well known in drilling of CFRP tape at the exit side but is not sufciently explained by former models [13,14], matches the empirical ndings from the milling experiments summarised in Section 3.3. 4. Conclusions Delamination is a major issue when machining bre reinforced plastics and avoiding this would bring considerable time and cost savings. This paper presented a systematic scheme for describing the delamination that occurs when milling CFRPs. Systematic prediction of delamination allows machining procedures to be adapted in order to avoid delamination when contour milling. It was demonstrated that two mechanisms are of key importance for describing delamination: occurrence and propagation. The results can be summarised as follows: 1. The occurrence of delamination and bre overhangs during the machining of CFRPs generally depends on the condition of the tool (tool wear) and the bre cutting angle h on the top laminate layers. 2. Even when using a very worn tool there is a preferred bre cutting angle range which results in no delamination and no bre overhangs (0 < h < 90). Furthermore, there is a critical bre cutting angle range (90 6 h < 180) in which delamination and bre overhangs occur. 3. Delamination occurs where bres are initially cut in the critical cutting angle range. When the working engagement equals the tool diameter, the bres are always initially cut at h = 180. 4. Delamination can propagate from the critical cutting angle range to the component edge, provided the bres are initially cut at a cutting angle of 90 6 h < 180 and at the component edge with a cutting angle of 0 < h < 90. 5. When milling below the critical cutting angle range in which delamination occurs (90 6 h < 180), the length of the bre overhang equals the distance of the component edge to the point at which the cutting edge cuts a bre at a critical cutting angle for the rst time. 6. Empirical ndings of milling and drilling experiments match with regard to the critical and uncritical ranges of the bre cutting angle thereby indicating their overall basic validity. The results of these experiments have high practical relevance. The ndings allow milling processes to be congured such that delamination is largely avoided, meaning considerable time and cost savings for the manufacture of composite components. References
[1] Abrao AM, Faria PE, Rubio JC, Campos JC, Reis P, Davim JPaulo. Drilling of ber reinforced plastics: a review. J Mater Process Technol 2007;186:17. [2] Bhatnagar N, Singh I, Nayak D. Damage investigation in drilling of glass ber reinforced plastic composite laminates. Mater Manuf Processes 2004;19(6):9951007.