Damage Quantification for the Machining of CFRP: An ... - Science Direct

Available online at www.sciencedirect.com

ScienceDirect Procedia Engineering 149 (2016) 2 – 16

International Conference on Manufacturing Engineering and Materials, ICMEM 2016, 6-10 June 2016, Nový Smokovec, Slovakia

Damage quantification for the machining of CFRP: An introduction about characteristic values considering shape and orientation of drilling-induced delamination Fabian Lisseka,*, Jacqueline Tegasa, Michael Kaufelda a

Institute of Production Engineering and Materials Testing, Ulm University of Applied Sciences, Prittwitzstraße 10, 89075 Ulm, Germany

Abstract This study deals with new opportunities to quantify the phenomena of delamination during the drilling of carbon fiber reinforced polymers (CFRP). The focus lies on determining additional characteristic values to describe the shape of the damage dimensions. Therefore, an algorithm was developed enabling an automatic analysis of delamination by generating an involute of the damage contour. This makes it possible to create two different types of characteristics. The statistical distribution of the delamination around the drilling hole is derived from parameters in analogy to the surface metrology. Apart from that different form factors will be introduced, which are calculated on the bases of the delaminated area or the moments of inertia respectively. The suitability of these characteristics is evaluated by a series of drilling tests in different materials with varying feed rates and tool conditions. M21/T800S, which is a quasi-isotropic UD-laminate, and the woven fabric ECG-Carbon 12K were the main materials compared to each other. The tools used for drilling are twist drills (d = 6.0 mm) with a point angle ɐ of 85° and a spiral angle λ of 35° at different states of tool wear. The feed rate was varied from 0.02 mm/rev to 0.1 mm/rev at a cutting speed of 100 m/min.

© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license © 2016 The Authors. Published by Elsevier B.V. (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of [S1]the organizing committee of ICMEM 2016 Peer-review under responsibility of the organizing committee of ICMEM 2016 Keywords: drilling ; CFRP ; damage quantification ; delamination ; form factors ; surface metrology ; M21/T800S

1. Introduction The machining of carbon fiber reinforced polymers (CFRP) is applied in different industrial sections, such as automotive, shipbuilding and aerospace industry. Besides conventional drilling, various manufacturing processes are used for machining of CFRP, such as water jet cutting [1–4] and laser machining [5–7]. But mostly, near-net-shape composite structures are machined in common manufacturing processes with defined cutting edge. This especially involves the drilling for the preparation of fastening elements or the trimming of supernatant material caused by the pressing process. During the machining of CFRP high quality standards are guaranteed by specialized tools and clamping systems. In fact, drilling is probably the most common kind of machining in industry [8]. For instance in the field of aerospace engineering there are as many as 40.000 - 55.000 holes required to be drilled in the production of one single unit of Airbus A350 aircraft [9]. The drilling of CFRP brings some challenges, such as the exorbitant tool wear and problems related to the specific structure of the laminates [10–13]. Regarding to this, delamination is seen as the most critical damage during drilling CFRP [10,12,14–17].

* Corresponding author. Tel.: +49-(0)731-50-28114 E-mail address: [email protected]

1877-7058 © 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license

(http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of ICMEM 2016

doi:10.1016/j.proeng.2016.06.632

3

Fabian Lissek et al. / Procedia Engineering 149 (2016) 2 – 16

Typical delamination phenomena occurring during drilling are the peel-up delamination at the entrance side and the push-out delamination at the exit side of the drilled hole. At the beginning of the drilling process, the cutting edge of the tool abrades the material until it finally starts cutting. During forward movement, the drill can pull some parts of the material upwards along its twist. As some of the top layers are separated from each other, this effect is called peel-up delamination (Fig. 1a). During the whole drilling process, the drill always exerts a thrust force on the layers in front of the chisel edge. With the drill progressing into the laminate material, the number of remaining uncut layers decreases. Shortly before the tool exits the material, the force upon the last layers may exceed the interlaminar bond strength and push-out delamination occurs (Fig. 1b) [18]. (a)

(b)

Fig. 1. Mechanism of drilling-induced delamination in CFRP composite laminates. (a) Peel-up delamination (b) Push-out delamination [19]

The studies concentrate on push-out delamination, because this is considered to have the worst influence on mechanical properties compared to peel-up delamination. Fig. 2a shows a typical damage appearing during drilling a CFRP laminate. On the top layer, predominantly around the drilling hole, delamination and also some overlapping fibers which were not cut accurately, are visible. For a simple description of the damage around the hole, parameters such as the delaminated area Ad and the maximum diameter Dmax including the widest delaminated fiber can be used (Fig. 2b). A short review of the common delamination factors, calculated on the basis of the parameters seen in Fig. 2b, is given in Table 1 and more detailed elsewhere [19,20]. Some researchers described the delamination damage with the help of simple parameters for the delamination size, as radius Rmax or diameter Dmax [21] (Table 1a). Others referred to more complex parameters for instance a combination of the sum of diameter and area ratios. Most commonly the ratio of maximum diameter of the damage zone Dmax to the hole diameter Dnom is applied to quantify the damage (Table 1b) [17,22–28]. In some cases [29], the ratio of the delamination area Ad to the nominal area Anom of the drilled hole was used (Table 1c). Different from that Davim and Rubio et al. [30–32] considered the delamination area Ad as well as the diameter Dmax for their quantification. They combined the parameters for an adjusted and more refined delamination factor Fda (Table 1d).

(a)

(b)

Dnom

Fig. 2. Example for push-out delamination in CFRP (magnification: 8x). (a) Push-out delamination in M21/T800S (b) Parameters for calculating delamination factors to quantify the degradation: Ad = damaged area ; Anom = nominal hole area ; Dnom = nominal hole diameter ; Dmax = maximum diameter of delamination

4

Fabian Lissek et al. / Procedia Engineering 149 (2016) 2 – 16 Table 1. Common delamination factors for damage quantification. Delamination factor

Associated formula

Type of delamination assessment

(a)

Delaminated radius

 ୫ୟ୶ െ  ୬୭୫

Delamination size in mm [21]

(b)

Fd

୫ୟ୶ ୬୭୫

Conventional delamination factor [22–28]

(c)

Fa

ୢ ୢ ൅ ୬୭୫ Ǣ ୬୭୫ ୬୭୫

Two-dimensional delamination factor [29]

(d)

Fda

ୢ ൅

ୢ ሺ ୢଶ െ ୢ ሻ ଶ ൬Ɏ ୫ୟ୶ െ ୬୭୫ ൰ Ͷ

Adjusted delamination factor [30–32]

Various studies were done about the effect of delamination damage on the mechanical properties of the laminate during drilling. The results show a significant reduction of fatigue strength and static strength [33,34]. Additionally, it is assumed that high feed forces are the main cause for delamination and it has been proven that the resulting delamination damage increases with higher feed rate [35]. So finally the ambition of most researchers is to find options to reduce delamination or to achieve delamination-free drilling by the determination of optimal machining parameters [19,29]. But in general there are still deficiencies shown in the quantification of damages at the top layers occurring for example during conventional drilling. (a)

(b) Dmax1

Ad2

Ad1

Dmax2 Dnom

Anom

ୟ ൌ

ୢଵ ୢଶ ൌ ୬୭୫ ୬୭୫

ୢ ൌ

୫ୟ୶ଵ ୫ୟ୶ଶ ൌ ୬୭୫ ୬୭୫

Fig. 3. Comparison of different types of delamination with the same delamination factor. (a) Area-based delamination factor Fa (b) Diameter-based delamination factor Fd.

(a)

(b)

Fig. 4. Drilling holes in M21/T800S with different types of delamination but the same delamination factor Fd (magnification: 8x). (a) Single torn out fiber (b) Multiple delamination damages.

5

Fabian Lissek et al. / Procedia Engineering 149 (2016) 2 – 16

The delamination factors actually used to evaluate the delamination of drilled holes in CFRP merely deliver absolute values to describe the extension of the delamination. All these characteristic more or less disregard the shape of the damaged area. In this context Fig. 3a shows a schematic picture of two drilled holes with different types of delamination. The delaminated area on the left side is located concentrically around the drilling hole, while on the right side there is an uneven distribution of the degradation. Both of the marked areas deliver the same value for the delaminated area A d and in consequence the same value for the delamination factor Fa. This point of view can be transferred on Fig. 3b with regard to the delamination factor Fd. The single torn out fiber creates exactly the same value for Fd as the concentric delaminated area. A more practical visualization is shown in Fig. 4a and Fig. 4b comparing a single torn out fiber with multiple delamination damages in autoclaved M21/T800S. It must be assumed that the previously described damage phenomena have differing effects on the mechanical toughness of the workpiece. So the purpose of this paper is to deliver alternatives for damage description by considering the shape and extension of the delaminated area as an addition to conventional delamination factors like Fd and Fa. 2. Quantification technique and delamination characteristics Delamination can be detected by different types of measurement systems. As delamination may occur within the CFRP work pieces as well as at the top layers, the analyzing methods range from acoustic emission for online monitoring [36] to ultrasonic C-scan or X-ray chromatography [27,29,37–39]. Mostly used is the visual inspection of the delaminated surfaces by microscopy, as it occurs in the drilling of CFRP. Normally the microscopic pictures are evaluated with a various number of software tools [10,12,15,16,21,22,29,40] . For this study, there was the need of a simple method to analyze the damages at the top layers of the laminate after drilling. Therefore the dark field microscopy was chosen here. This kind of microscopy is capable to produce highcontrast pictures of the damaged areas, which simplifies the automatic segmentation of the delamination with a self-developed quantification algorithm. 2.1. Segmentation and evaluation of the delamination contour The convenient delamination factors like Fd and Fa are important und very useful for a fast damage quantification. But as seen in Fig. 3 and Fig. 4 they are not very significant with regard to the shape or the distribution of the delamination. For this reason it is necessary to establish further parameters to complement the actual delamination factors with such information.

(b)

(d) Fibre orientation

Peripheral angle α

200 top layer orientation 180

α = 0° Workpiece orientation

(e)

(c)

Delamination factor Fd in %

(a)

160 140 120 100 0

60

120

180

240

300

Peripheral angle α in ° Fig. 5. Description of the quantification process for damage evaluation (magnification 8x): (a)-(c) Automatic segmentation of the delamination (d) Specification of the fiber orientation and the work piece orientation (e) Calculation of the involute and the delamination characteristics with respect to the peripheral angle.

360

6

Fabian Lissek et al. / Procedia Engineering 149 (2016) 2 – 16

So at this point, the quantification process, which was used in this study, should be presented: After recording a picture of the drilling hole (Fig. 5a) via dark field microscopy, further steps can be applied by a quantification algorithm. Therefore a grey scale comparison is performed to identify all pixel belonging to the delamination contour (Fig. 5b). The algorithm sorts out wrongly marked pixels which can be caused by dust on the work piece surface. After the determination of the contour of the drilling hole the algorithm calculates a spline curve based on the previously marked pixels (Fig. 5c). This curve represents the ordinate values of the delamination, ignoring fibers overlapping into the drilling hole. Additionally the fiber orientation and the work piece orientation have to be defined manually, as they are used for the calculation of form factors (Fig. 5d). Finally the algorithm creates an involute along the peripheral angle α (Fig. 5e) from 0° to 360°. In the following chapters this involute will be designated as delamination profile or delamination curve respectively. In general, the delamination curve forms the basis for all quantification methods in this study. 2.2. Form factors Fq and FIq To describe the shape of a delamination, two types of form factors will be introduced in this chapter. Both of them are using the fiber orientation to calculate their characteristic values. As shown in Fig. 6a the form factor Fq is calculated by the ratio of the areas A1 to A4 which are determined by four quadrants. The quadrants are angled with 45° to the fiber orientation and allow to generate information about the orientation of the delamination: For Fq > 1 the main orientation of the delamination is in fiber direction. Fq ൎ 1 implies a concentric degradation surrounding the drilling hole. And finally, F q < 1 suggests that the delamination is orientated orthogonal to fiber orientation. The latter will occur mainly for woven fabrics, as the fiber orientation of the top layers can be determined in two different ways in this case. The work piece orientation can be used to control the coincidence of the main delamination direction and the top layer orientation. As shown in Fig. 6b the second form factor is calculated by the moments of inertia I z and Iy. For this purpose the fiber orientation is defined as Z-axis whereby the Y-axis is oriented in orthogonal direction to Z. In consequence F Iq is the ratio of Iy and Iz. The values of FIq can be interpreted in the same way as described for Fq. Potential differences of both form factors will be analyzed within the experimental verification of the characteristics.

(a)

Quadrants

(b) A3

Y-axis (orthogonal to fibre orientation)

dA

A1

A2

A4

y

Workpiece orie orientation

Z-axis (fibre orientation)

Form factor Fq ଵ ൅   ଶ ୯ ൌ ଷ ൅  ସ

Z-axis (fibre orientation)

Moments of intertia Iy and Iz 

‫ œ ׬‬ଶ † ୷ ൌ ୅ ଶ ୬୭୫  ୬୭୫

‫ › ׬‬ଶ † ୸ ൌ ୅ ୬୭୫ ଶ୬୭୫

Form factor FIq ୍୯ ൌ

୷ ୸

Fig. 6. Illustration of the calculation methods for the different form factors. (a) Fq: Calculation by an area ratio (b) FIq: Calculation by the moments of inertia.

2.3. Delamination characteristics in analogy with surface metrology In analogy to surface measurement, the most significant parameters from this sector have been taken into account to adapt them for a statistical description of the delamination profiles. In Table 2 the corresponding roughness characteristics are listed as well as the analogies for application in delamination analyses. The characteristics will be described shortly in the following part and have to be considered with respect to Fig. 7 and Fig. 8.

7

Fabian Lissek et al. / Procedia Engineering 149 (2016) 2 – 16 Table 2. Characteristics in surface metrology to be assigned for delamination quantification.

Parameters in surface metrology (DIN EN ISO 4287)

ୟ ൌ

Analogies for application in delamination analysis

ͳ ୪ න ȁሺšሻȁ†š Ž ଴

ୢǡୟୢ ൌ

ͳ ୪  ୯ ൌ ඨ න ଶ ሺšሻ†š Ž ଴

ଷ଺଴ι ͳ න ȁሺȽሻȁ†Ƚ ͵͸Ͳι ଴

ଷ଺଴ι ͳ ୢǡ୰୫ୱ ൌ ඨ න ଶ ሺȽሻ†Ƚ ͵͸Ͳι ଴

 ୱ୩ ൌ

ͳ ͳ ୪୰ ଷ ቈ න  ሺšሻ†š቉ ଷ୯ Ž” ଴

ୢǡୱ୩ ൌ

ଷ଺଴ι ͳ ͳ න ଷ ሺȽሻ†Ƚ቉ ቈ ଷ ͵͸Ͳι ଴ ୢǡ୰୫ୱ

 ୩୳ ൌ

ͳ ͳ ୪୰ ସ ቈ න  ሺšሻ†š቉ ସ୯ Ž” ଴

ୢǡ୩୳ ൌ

ଷ଺଴ι ͳ ͳ න ସ ሺȽሻ†Ƚ቉ ቈ ସ ͵͸Ͳι ଴ ୢǡ୰୫ୱ

Two of the most known characteristics in surface metrology are the average roughness Ra and Rq. Ra is defined as the arithmetic average of all absolute values in the roughness profile, whereas R q is calculated by the root-mean-square of the ordinate values. Transferring both of them to the application of damage quantification by using a delamination profile (cf. Table 2 ; Ra ‫ ؙ‬Fd,ad ; Rq ‫ ؙ‬Fd,rms), they are not influenced strongly by single delaminations, but they can provide information about irregularities in the deviation of the delamination curve. As seen in Fig. 7, for the calculation of the different delamination characteristics the mean line is determined by the arithmetic average of all ordinate values. The difference to this mean line is described as a function Z in dependency of the peripheral angle α. 200

Delamination factor Fd in %

180

Z(α)

160 mean line 140

Fd,ad

Fd,rms

120

100 0

90

180

270

360

Peripheral angle α in ° Fig. 7. Determination of the delamination characteristics Fd,ad and Fd,rms on the basis of the delamination profile.

However, Fd,ad and Fd,rms could be used to build average values, but in consequence they are not affected significantly by single peaks or single torn out fibers respectively. Although Fd,rms should be more sensitive to single peaks for mathematical reasons, there are other characteristic values which are more suitable for damage quantification. The skewness Rsk and the kurtosis Rku are calculated by using Rq powered by three or four respectively. The main attribute of these characteristics is, that they are influenced strongly by peaks and valleys in a roughness profile. In contrast to surface metrology this can be interpreted as an advantage for the damage evaluation. By adapting them to Fd,sk and Fd,ku for delamination quantification this predominantly qualifies them as delamination characteristics in this specific application.

8

Fabian Lissek et al. / Procedia Engineering 149 (2016) 2 – 16

x

x

The skewness Fd,sk describes the asymmetry of the frequency distribution of the delamination curve. As shown in Fig. 8a, for the equivalent Rsk negative values identify surfaces with good load-bearing capability. For the evaluation of a delamination profile this means, that there will be a various number of high plateaus in the delamination curve. Fd,sk = 0 indicates an uniform distribution of the single delaminations and for F d,sk > 0 an increasing amount of narrow peaks in the profile can be assumed. As seen in Fig. 8b, the kurtosis Fd,ku describes the gradient of the frequency distribution of the delamination curve. For normal distributed delamination curves Fd,ku equals 3. Values higher than 3 indicate a high amount of single peaks in the delamination curve, whereas lower values are standing for a more concentric form of delamination.

In summary, for the skewness of the delamination curve the position of the mean line in proportion to the peak of the relative frequency is significant and for the kurtosis the course of the relative frequency itself is the distinctive feature. (a)

(b)

Fig. 8. Comparison of Rsk and Rku for surface assessment. (a) Frequency distributions for the skewness (b) Frequency distributions for the kurtosis [41].

3. Experimental evaluation 3.1. Experimental setup for drilling The drilling series are carried out using two different CFRP materials: M21/T800S is a multiply laminate and ECG Carbon 12K has been chosen as a woven fabric. For M21/T800S the UD-Prepreg is manufactured by the company Hexcel.The individual plies have a weight per unit area of 268 g/m² and 196 g/m² respectively, each leading to a resin content of 35 % by weight. The CFRP specimens have a quasi-isotropic structure and a symmetrical stacking sequence. The laminates are fabricated in two different ways, which are on the one hand an autoclaving process and on the other hand a heat pressing process. The autoclaved type of material has 24 plies with a total height of 4.6 mm (196 g/m² ; [-45, 90, 45, 0]3s) whereas the thickness of the heat pressed material is 4.0 mm with 16 plies (268 g/m² ; [-45, 90, 45, 0]2s). The plies of ECG-Carbon 12K are manufactured as basket weave, building up a 20-layer laminate with a fiber orientation of 0/90° and 5.2 mm thickness. The drilling tools used for machining are twist drills (d = 6.0 mm) with a point angle ɐ of 85° and a spiral angle λ of 35° at different states of cutting edge rounding, which are rß1 = 6 μm, rß2 = 18 μm and rß3 = 49 μm. The feed rate was varied between f = 0.02 mm/rev and 0.1 mm/rev (cf. Table 3). For each combination of feed rate and tool condition, 5-7 holes were drilled. The quantification of the delamination damage was carried out with the analyzing method described in chapter 2.1. The involute of the delamination profile was used to evaluate the parameters and to characterize the delamination damage (cf. Fig. 5e). Table 3: Experimental setup for the evaluation of the drilling series Work piece material

Cutting edge rounding rβ

Machining parameters

6 μm

Cutting speed vc 100 m/min

M21/T800S heat pressed M21/T800S autoclaved

18 μm 49 μm

ECG Carbon 12K heat pressed

Feed rate f 0.02 mm/rev ; 0.06 mm/rev ; 0.10 mm/rev

Twist drill geometry

Diameter d = 6.0 mm Point angle ɐൌ85° Spiral angle λ = 35°

9

Fabian Lissek et al. / Procedia Engineering 149 (2016) 2 – 16

3.2. Application of form factors Within the machining experiments both of the form factors F q and FIq have been determined and compared respectively. The main idea of the form factors is to complete the common absolute delamination factors with information about the shape of the damage contour. This idea is represented in Fig. 9 which shows the results of the variation of the feed rate from 0.02 mm/rev to 0.1 mm/rev at drilling ECG-Carbon 12k and M21/T800S heat pressed. The figure demonstrates the relationship between the delamination factor Fd and the form factor Fq for different types of delamination. Based on the diagram, different aspects of damage quantification can be pointed out. As reported by literature in great detail the extent of delamination is rising clearly with higher feed rates, especially apparent for the heat pressed multiply laminate M21/T800S. (a) 250

(b)

Delamination factor Fd in %

225

Increasing width of delamination

200

(d) (c)

175 (c) 150

(b) Concentric delamination

125

(d)

(a)

100 0

1

2

3

4

5

6

Form factor Fq M21/T800S heatpr. (0.02 mm/rev)

ECG-Carbon 12k (0.02 mm/rev)

M21/T800S heatpr. (0.06 mm/rev)

ECG-Carbon 12k (0.06 mm/rev)

M21/T800S heatpr. (0.1 mm/rev)

ECG-Carbon 12k (0.1 mm/rev)

Fig. 9. Correlation between the delamination factor Fd and the form factor Fq with respect to different feed rates and materials (vc = 100 m/min): (a) Concentric type of delamination (b)-(d): Increasing Fd and Fq with higher feed rates.

In this context the difference to the delamination factors of the woven fabric ECG-Carbon 12k is remarkable. All drilling series shown in the diagram were made with a cutting edge rounding of 18 μm, whereas F d is much lower for the woven fabric as for M21/T800S. Beside the evaluation of the absolute values of Fd, Fig. 9 allows a clear interpretation of the correlation between the form factor Fq and the shape of the delamination (cf. Fig. 9a-d). The woven fabric ECG-Carbon 12k causes a concentric form of delamination for all three feed rates (Fig. 9a). The form factor Fq, is approximately 1 for this type of degradation. The different kinds of delamination for M21/T800S heat pressed are represented in Fig. 9b to Fig. 9d. Basically the extent of delamination is in fiber direction for Fq > 1 but additionally there is a stepwise increase of the delamination width with higher feed rates. This results in increasing values of Fq as well. In chapter 2.2 two different methods have been presented to describe the form of delamination. Therefore Fig. 9 shows the quadrant-based form factor Fq using area ratios. Furthermore the moments of inertia allow the calculation of the form factor F Iq by building the quotient of Iz and Iy (cf. Fig. 6). The different factors are compared in Fig. 10 by visualizing the ratio of both factors, whereas a value of F q/FIq = 1 implies an exact match of the factors. Most of the values correspond with a maximum deviation of ± 8 %. For that reason Fq can be considered representative for both calculation methods. While form factors with Fq < 1.7 can be associated with concentric types of delamination, which occur for woven fabrics like ECG-Carbon 12k, Fq > 1.7 refers to delaminations in fiber direction such as shown by the multiply laminate M21/T800S (autoclaved and heat pressed).

10

Fabian Lissek et al. / Procedia Engineering 149 (2016) 2 – 16

(a)

(b)

7

1,0

Standard deviation σ of Fq and FIq

6

Form factor Fq

5 4 3

ECG-Carbon 12k

M21/T800S

2

0,8

0,6

0,4

0,2

1 0

0,0 0

1

2

3

4

5

6

7

6 μm

Form factor FIq rβ = 6 μm

rβ = 18 μm

18 μm

49 μm

Cutting edge rounding rβ

rβ = 49 μm

Fq M21/T800S heatpr.

FIq M21/T800S heatpr.

Fq M21/T800S autocl.

FIq M21/T800S autocl.

Fq ECG-Carbon 12k

FIq ECG-Carbon 12k

Fig. 10. Comparison of the two form factors Fq and FIq with respect to tool wear and material (a) Verification of the accordance of Fq and FIq (b) Comparison of the standard deviation σ of Fq and FIq.

Fig. 10b demonstrates the standard deviation of both of the form factors for all drilling series at different states of tool wear. In the diagram there is no conspicuous difference between the deviations of Fq and FIq, whereby both factors can be used without disadvantages to each other. But nevertheless there is a limit for the applicability of the form factors which is apparent in Fig. 10a. Although there is a lower extent of delamination at r β = 18 μm than at rβ = 49 μm, the absolute value of the form factors is the highest in all of the investigations. In conclusion as the form factor is used for highly damaged drilling holes, like at rβ = 49 μm, it cannot be associated directly with the delamination width. This fact will be examined in the following part with regard to Fig. 11. (a)

(b)

250

250 225 Delamination factor Fd in %

Delamination factor Fd in %

225 200 Wide delamination in fibre direction

175 150 125

Concentric delamination

200

150 Narrow delamination in fibre direction

125

Narrow delamination in fibre direction

100

Wide delamination in fibre direction

175

Concentric delamination

100 0

1

2

3

4

5

Form factor Fq

6

0

1

2

3

4

5

6

Form factor Fq

M21/T800S heatpr. (6 μm)

ECG-Carbon 12k (6 μm)

M21/T800S autocl. (6 μm)

ECG-Carbon 12k (6 μm)

M21/T800S heatpr. (18 μm)

ECG-Carbon 12k (18 μm)

M21/T800S autocl. (18 μm)

ECG-Carbon 12k (18 μm)

M21/T800S heatpr. (49 μm)

ECG-Carbon 12k (49 μm)

M21/T800S autocl. (49 μm)

ECG-Carbon 12k (49 μm)

Fig. 11. Correlation between the delamination factor Fd and the form factor Fq with respect to different states of tool wear and materials (vc = 100 m/min): (a) ECG Carbon 12K and M21/T800S heat pressed (b) ECG Carbon 12K and M21/T800S autoclaved

Fabian Lissek et al. / Procedia Engineering 149 (2016) 2 – 16

The shown diagrams compare the delamination quantification of all three types of material with respect to the shape of the delamination and the tool wear (rβ = 6 μm; 18μm ; 49 μm). As already seen in Fig. 9 the woven fabric is characterized by a form factor Fq of approximately 1 indicating a concentric shape of degradation around the drilling hole (cf. Fig. 11a und Fig. 11b). While the delamination factors at rβ = 6 μm and rβ = 18 μm can be numbered on the same level, Fd raises from 107 % to 132 % at rβ = 49 μm whereas the form factor Fq remains the same value. In each of the diagrams of Fig. 11 there is one type of the multiply laminate M21/T800S presented. It can be noted that the heat pressed material (Fig. 11a) and the autoclaved material (Fig. 11b) nearly behave in the same way with respect to the quantified delaminations within the applied drilling series. In the diagram both of them show areas with narrow delaminations (F d = 125 %) as well as areas with wide delaminations (Fd = 175 %). But as a whole, the form factors of the autoclaved material are slightly higher with regard to the lower states of cutting edge rounding (rβ = 6 μm and rβ =18 μm). Very noticeable are the areas near Fd = 215 % in both of the diagrams. Although the width of the damaged area is increasing (cf. Fig. 9c), Fq is decreasing compared to rβ =18 μm. The reason for this discrepancy compared to Fig. 9 is, that very large delaminations cause some damaged areas to extent into additional quadrants, although the damage is orientated still in fiber direction. In the presented investigations this effect occurred exclusively for a cutting edge rounding of rβ = 49 μm. So actually these kinds of delaminations are the limiting factor for the applicability of the form factors Fq or FIq respectively. 3.3. Statistic evaluation of the delamination contour The statistical evaluation of the delamination profiles by using the delamination skewness Fd,sk and the delamination kurtosis Fd,ku will be discussed by the means of selected examples of drilling holes in CFRP. For this reason Fig. 12 shows different types of delamination as well as the corresponding delamination curves and the relative frequency distributions. Fig. 12a presents a drilling hole in ECG-Carbon 12k with very good drilling quality. The delamination factor Fd is 105 %, speaking for an almost delamination-free drilling process. More precisely, the quantified damage is caused by slight spalling of the matrix around the drilling hole. The ordinates of the delamination have a standard deviation of ± 0.9 %, which emphasizes the good quality. Regarding the relative frequency distribution of the delamination factors, in first approximation the course of the curve can be compared with a Gaussian distribution. Consequently the delamination skewness F d,sk = 0.49 and the delamination kurtosis Fd,ku = 2.72 confirm this result. Fig. 12b shows a delaminated area in ECG-Carbon 12k, which has a concentric distribution around the drilling hole. With Fd = 130 % the cutting quality is worse than in Fig. 12a. At the beginning and at the end of the delamination profile there are some peaks visible, but overall the delamination factors are distributed evenly around the mean line of the curve. The delamination values increase to 1.34 for Fd,sk or rather 6.26 for Fd,ku. Thereby the kurtosis takes into account the peaks occurring in the delamination profile and the skewness is affected by a slight moving of the maximum of the relative frequency distribution, which is now below the mean line. While Fig. 12a and Fig. 12b were representing a woven fabric, Fig. 12c shows the multiply laminate M21/T800S heat pressed. In this figure a typical degradation with two narrow delaminations is presented, both of them in fiber direction. The delamination factor is much higher with Fd = 189 % and in the delamination profile two distinctive peaks are evident. Both of them can be associated with one of the two torn out fiber bundles. As a consequence of the two maxima, the skewness is calculated with Fd,sk = 2.8 and the kurtosis Fd,ku is 11.18. These values directly correlate with the delamination curve, as the maximum of the relative frequency distribution has moved clearly below the arithmetic average of the delamination ordinates. Additionally the very high value of the kurtosis confirms the high gradient of the frequency distribution (Fd,ku >> 3), which has been described in Fig. 8 previously. Fig. 12d shows the worst drilling hole quality in this comparison, which was machined in M21/T800S autoclaved. The delamination was quantified with a delamination factor of F d = 284 %. The microscopic picture and the delamination profile point out, that it is indispensable to consider the skewness and the kurtosis with respect to the delamination factor Fd. Both of the delamination characteristics (Fd,sk = 1.68 ; Fd,ku = 5.74 ) are comparable with Fig. 12b, but the absolute extent of damage is strongly differing.

11

12

Fabian Lissek et al. / Procedia Engineering 149 (2016) 2 – 16 Drilling hole Delamination factor Fd in %

a)

Delamination profile 110

105

100 0

ECG-Carbon 12k ; Fd = 105 % Fd,sk = 0.49 ; Fd,ku = 2.72

M21/T800S autoclaved ; Fd = 284 % Fd,sk = 1.68 ; Fd,ku = 5.74

360

0

25

50

75

Frequency distribution in %

120 110 100 0

90

180

270

360

0

25

50

75

Frequency distribution in %

190 160 130 100 0

90

180

270

360

0

Peripheral angle α in °

Delamination factor Fd in %

d)

270

Peripheral angle α in °

Delamination factor Fd in %

M21/T800S heat pressed Fd = 189 % Fd,sk = 2.8 ; Fd,ku = 11.18

180

130

ECG-Carbon 12k ; Fd = 130 % Fd,sk = 1.34 ; Fd,ku = 6.26 c)

90

Peripheral angle α in °

Delamination factor Fd in %

b)

Relative frequency distribution

25

50

75

Frequency distribution in %

300 250 200 150 100 0

90

180

270

Peripheral angle α in °

360

0

25

50

Frequency distribution in %

Fig. 12. Various drilling holes in CFRP comparing the skewness Fd,sk and the kurtosis Fd,ku for damage quantification. Machining parameters: vc = 100 m/min ; f = 0.06 mm/rev ; (a) rβ = 6μm (b) rβ = 49 μm (c) rβ = 18 μm (d) rβ = 49 μm.

75

13

Fabian Lissek et al. / Procedia Engineering 149 (2016) 2 – 16

With regard to the different states of tool wear known from previous discussions, Fig. 13 gives a résumé about the delamination factor Fd and the supporting delamination characteristics Fd,sk and Fd,ku. As the feed force is increasing proportional to the cutting edge rounding, Fd rises with a higher degree of tool wear as well. Comparing the multiply laminates with the woven fabric, the extent of damage is generally higher for M21/T800S. The average values of the skewness and the kurtosis are not significantly affected by higher grades of delamination in ECG-Carbon 12k. This is coincident with the concentric form of the delamination, which is characteristic for this kind of material. Both types of M21/T800S show decreasing values for the skewness and the kurtosis, but in the final analysis the average characteristics are always higher than 0 (Fd,sk) or 3 (Fd,ku) respectively. In conclusion, all delamination profiles are dominated by a various amount of peaks, which are decisive for the relative frequency distribution of the delamination profile. In the course of the drilling experiments a skewness of Fd,sk < 0 could not be determined, which implies that this range of values is not relevant for damage quantification in drilling CFRP. a)

b)

300

c)

30

5

Kurtosis of delamination Fd,ku

Delamination factor Fd in %

200

150

Skewness of delamination Fd,sk

25 250

20

15

10

5

100

0 6

18

0

49

6

Cutting edge rounding rβ in μm

18

6

49

Cutting ede rounding rβ in μm

18

49

Cutting ede rounding rβ in μm

M21/T800S autocl.

M21/T800S autocl.

M21/T800S autocl.

M21/T800S heatpr.

M21/T800S heatpr.

M21/T800S heatpr.

ECG-Carbon 12k

ECG-Carbon 12k

ECG-Carbon 12k

Fig. 13. Comparison of characteristic values for damage quantification with respect to material and tool wear. (a) Delamination factor Fd (b) Delamination kurtosis Fd,ku (c) Delamination skewness Fd,sk.

Delamination factor Fd in %

At first sight, the high values for kurtosis and skewness in Fig. 13 disagree with the characteristic values of a good drilling quality as shown in Fig. 12a. The explanation for this correlation is given in Fig. 14. The drilling holes in M21/T800S with a cutting edge radius of rβ = 6 μm show slight peaks in the delamination profile, which significantly influence the frequency distribution. This speaks for a very high sensitivity of the described delamination characteristics and emphasizes particularly that the asymmetry of the frequency distribution of the delamination curve is the most important factor.

M21/T800S heat pressed Fd = 123 % Fd,sk = 4.35 ; Fd,ku = 26.55

130 Highly asymmetric with Fd,sk = 4.35

120 110 100 0

90

180

270

Peripheral angle α in °

360

0

25

50

Frequency distribution in %

Fig. 14. Example for the high sensibility of the kurtosis and the skewness in the application of delamination quantification. Machining parameters: vc = 100 m/min ; f = 0.06 mm/rev ; rβ = 6 μm

75

14

Fabian Lissek et al. / Procedia Engineering 149 (2016) 2 – 16

4. Conclusions The presented studies could prove a good applicability of the complementary characteristics for describing the shape of the delamination as well as the statistical distribution of the delamination profile of various drilling holes. In particular the diagram corresponding to Fig. 15 is suitable to evaluate the shape of the delaminated area very well. On the one hand the common absolute delamination factors like Fd and Fa can be integrated and on the other hand the shape of delamination can be described by the form factors Fq or FIq respectively. Hence, concentric delaminations can be determined and additionaly the preferred direction of delamination can be identified with respect to the fiber orientation. Furthermore the form factors are correlating with the delamination width as far as parts of the delaminated areas in fiber direction are not extending to additional quadrants. Delamination orthogonal to the fiber orientation is included in the form factors although these kind of delaminations occur mainly for woven fabrics.

Delamination factor in %

a) delamination orthogonal to fibre direction

delamination in fibre direction

200 (a) concentric delamination

Occurring rarely for woven fabrics

(b) narrow delamination

150

b)

(c) wide delamination

c) 100 1

0,1

10

lg (Form factors Fq or FIq)

Fig. 15. Summary of the quantification of delamination using form factors.

Table 4. Summary of the quantification of delamination using the delamination kurtosis and the delamination skewness. Combination of delamination characteristics

Type of delamination

Fd ~ 100 % ; Fd,sk ~ 0 ; Fd,ku ~3

High quality drilling hole (cf. Fig. 12a)

Slight increase of Fd, moderate increase of Fd,sk, Fd,ku

Concentric delamination (cf. Fig. 12b)

Slight increase of Fd, very high increase for Fd,sk, Fd,ku

High quality drilling hole with very small peaks (cf. Fig. 14)

High Fd, moderate increase for Fd,sk, Fd,ku

Large scale delamination (cf. Fig. 12c)

High Fd, high increase for Fd,sk, Fd,ku

Single torn out fiber bundles(cf. Fig. 12d)

The description of the delamination profile by analogy with the surface metrology can be used especially to describe irregularities in the delamination profile. The skewness and the kurtosis provide information about the symmetry and the gradient of the relative frequency distribution of the delamination ordinates. Although it is a statistic evaluation method both of the characteristics are capable to give an idea about the form of the delamination. Therefore two assumptions have to be done: Firstly the main direction of the delamination is in fiber orientation and secondly the values have to be considered with respect to a delamination factor. The characteristic values Fd,sk and Fd,ku are highly sensible for single peaks in the delamination profile, which was shown especially in Fig. 14. In the course of the investigation different types of delamination could be associated with a specific combination of delamination characteristic. These correlations are summed up in Table 4. In conclusion, delamination can be quantified by different priorities with strongly differing types of characteristic values, whereas the interpretation of the relative frequency distribution is more difficult in comparison to the evaluation of the form factors presented in this study.

Fabian Lissek et al. / Procedia Engineering 149 (2016) 2 – 16

Acknowledgements The presented studies were performed within the framework of the project “SPANTEC-light” as a part of the research association „Zentren für angewandte Forschung an Fachhochschulen“ (ZAFH), founded by the German federal state BadenWuerttemberg. The project and its associates, the Universities of Applied Sciences in Aalen, Mannheim and Ulm, focus on the investigation of the quantitative correlations between material and application properties at machining CFRP. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35]

Lau W, Yue T, Lee T, Lee W. Un-conventional machining of composite materials. Journal of Materials Processing Technology, 1995, 48, 199–205. Mathew J, Goswami G, Ramakrishnan N., Naik N. Parametric studies on pulsed Nd:YAG laser cutting of carbon fiber reinforced plastic composites. Journal of Materials Processing Technology, 1999, 89–90, 198–203. K.T. Voisey, S. Fouquet, D. Roy, T.W. Clyne. Fiber swelling during laser drilling of carbon fiber composites. Optics and Lasers in Engineering, 2006, 44, 1185–1197. Herzog D, Jaeschke P, Meier O, Haferkamp H. Investigations on the thermal effect caused by laser cutting with respect to static strength of CFRP. International Journal of Machine Tools and Manufacture, 2008, 48, 1464–1473. M.A. Azmir, A.K. Ahsan. A study of abrasive water jet machining process on glass/epoxy composite laminate. Journal of Materials Processing Technology, 2009, 209, 6168–6173. Shanmugam D, Nguyen T, Wang J. A study of delamination on graphite/epoxy composites in abrasive waterjet machining. Composites Part A: Applied Science and Manufacturing, 2008, 39, 923–929. Lemma E, Chen L, Siores E, Wang J. Study of cutting fiber-reinforced composites by using abrasive water-jet with cutting head oscillation. Composite Structures, 2002, 57, 297–303. Tsao C, Hocheng H. The effect of chisel length and associated pilot hole on delamination when drilling composite materials. International Journal of Machine Tools and Manufacture, 2003, 43, 1087–1092. Koller P. Robotergeführter Sensor checkt beim Airbus jede einzelne Niete. Elektronik Praxis, 2014. H. Hocheng, H.Y. Puw. On drilling characteristics of fiber-reinforced thermoset and thermoplastics. International Journal of Machine Tools and Manufacture, 1992, 32, 583–592. König W, Wulf C, Graß P, Willerscheid H. Machining of Fiber Reinforced Plastics. CIRP Annals - Manufacturing Technology, 1985, 34, 537–548. E.Ugo Enemuoh, A.Sherif El-Gizawy, A Chukwujekwu Okafor. An approach for development of damage-free drilling of carbon fiber reinforced thermosets. International Journal of Machine Tools and Manufacture, 2001, 41, 1795–1814. J Mathew, N Ramakrishnan, N.K Naik. Trepanning on unidirectional composites: delamination studies. Composites Part A: Applied Science and Manufacturing, 1999, 30, 951–959. P. Ramesh Babu, B. Pradhan. Effect of damage levels and curing stresses on delamination growth behaviour emanating from circular holes in laminated 5FRP6 composites. Composites Part A: Applied Science and Manufacturing, 2007, 38, 2412–2421. Davim J, Reis P. Study of delamination in drilling carbon fiber reinforced plastics (CFRP) using design experiments. Composite Structures, 2003, 59, 481–487. Capello E. Workpiece damping and its effect on delamination damage in drilling thin composite laminates. Journal of Materials Processing Technology, 2004, 148, 186–195. Chen W. Some experimental investigations in the drilling of carbon fiber-reinforced plastic (CFRP) composite laminates. Albuquerque V de, Tavares J, Durão L. Evaluation of Delamination Damages on Composite Plates using Techniques of Image Processing and Analysis and a Backpropagation Artificial Neural Network, 2008. Liu D, Tang Y, Cong W. A review of mechanical drilling for composite laminates. Composite Structures, 2012, 94, 1265–1279. Davim J, Machinability of fiber-reinforced plastics: Delamination in composite materials: measurement, assessment and prediction; p. 139-162, De Gruyter, Berlin/Boston, 2015. El-Sonbaty I, Khashaba U, Machaly T. Factors affecting the machinability of GFR/epoxy composites. Composite Structures, 2004, 63, 329–338. Davim J, Reis P. Drilling carbon fiber reinforced plastics manufactured by autoclave—experimental and statistical study. Materials & Design, 2003, 24, 315–324. J.Paulo Davim, Pedro Reis, C.Conceição António. Experimental study of drilling glass fiber reinforced plastics (GFRP) manufactured by hand lay-up. Composites Science and Technology, 2004, 64, 289–297. J. Paulo Davim, Pedro Reis, C. Conceição António. Drilling fiber reinforced plastics (FRPs) manufactured by hand lay-up: influence of matrix (Viapal 5VUP6 9731 and 5ATLAC6 382-05). Journal of Materials Processing Technology, 2004, 155–156, 1828–1833. Gaitonde V, Karnik S, Rubio J, Correia A, Abrão A, Davim J. Analysis of parametric influence on delamination in high-speed drilling of carbon fiber reinforced plastic composites. Journal of Materials Processing Technology, 2008, 203, 431–438. S. Arul, L. Vijayaraghavan, S.K. Malhotra, R. Krishnamurthy. The effect of vibratory drilling on hole quality in polymeric composites. International Journal of Machine Tools and Manufacture, 2006, 46, 252–259. Tsao C, Hocheng H. Effect of tool wear on delamination in drilling composite materials. International Journal of Mechanical Sciences, 2007, 49, 983– 988. Sanjay Rawat, Helmi Attia. Wear mechanisms and tool life management of WC–Co drills during dry high speed drilling of woven carbon fiber composites. Wear, 2009, 267, 1022–1030. Faraz A, Biermann D, Weinert K. Cutting edge rounding. International Journal of Machine Tools and Manufacture, 2009, 49, 1185–1196. DAVIM J, RUBIO J, ABRAO A. A novel approach based on digital image analysis to evaluate the delamination factor after drilling composite laminates. Composites Science and Technology, 2007, 67, 1939–1945. Tsao C, Kuo K, Hsu I. Evaluation of a novel approach to a delamination factor after drilling composite laminates using a core–saw drill. Int J Adv Manuf Technol, 2012, 59, 617–622. Nagarajan V, Selwin Rajadurai J, Annil kumar T. A digital image analysis to evaluate delamination factor for wind turbine composite laminate blade. Composites Part B: Engineering, 2012, 43, 3153–3159. Khashaba U, El-Sonbaty I, Selmy A, Megahed A. Machinability analysis in drilling woven GFR/epoxy composites. Composites Part A: Applied Science and Manufacturing, 2010, 41, 1130–1137. Erik Persson, Ingvar Eriksson, Leif Zackrisson. Effects of hole machining defects on strength and fatigue life of composite laminates. Composites Part A: Applied Science and Manufacturing, 1997, 28, 141–151. Campos Rubio J, Abrao A, Faria P, Correia A, Davim J. Effects of high speed in the drilling of glass fiber reinforced plastic. International Journal of Machine Tools and Manufacture, 2008, 48, 715–720.

15

16

Fabian Lissek et al. / Procedia Engineering 149 (2016) 2 – 16 [36] [37] [38] [39] [40] [41]

X. J. Cai, S. Qin, Q. L. An, H. Z. Zhang, S. Han, M. Chen. Experimental Analysis on Delamination Damage by Acoustic Emission in High Speed Drilling of Carbon Fiber Reinforced Plastics. Key Engineering Materials, 2014, Vols. 589-590. Tsao C, Hocheng H. Computerized tomography and C-Scan for measuring delamination in the drilling of composite materials using various drills. International Journal of Machine Tools and Manufacture, 2005, 45, 1282–1287. Silva D, Pamies Teixeira J, Machado C. Methodology analysis for evaluation of drilling-induced damage in composites. Int J Adv Manuf Technol, 2014, 71, 1919–1928. Hounsfield G. Computerized Transverse Axial Scanning (Tomography). British Journal of Radiology, 1973. Tsao C, Hocheng H. Taguchi analysis of delamination associated with various drill bits in drilling of composite material. International Journal of Machine Tools and Manufacture, 2004, 44, 1085–1090. Mahr GmbH Göttingen. Perthometer Oberflächen-Kenngrößen, 1999.