Laser processing of continuous carbon fibre reinforced polyphenylene sulphide organic sheets—Correlation of process parameters and reduction in static tensile strength properties

Abstract

In this article, the impact of laser-induced thermal damage on the static strength properties of consolidated continuous carbon fibre reinforced thermoplastics induced by high-power laser cutting is presented. Organic sheets based on a polyphenylene sulphide matrix are machined using a fibre laser providing a maximum output power of 6000 W. In this context, the influence of the applied laser power and the feed rate on the cut quality as well as the resulting tensile strength is discussed. In order to analyse the laser cutting edge through a quantitative evaluation of the damaged areas due to laser impact, optical micrographs are prepared. The results of tensile strength tests are compared with those measured from the specimens that were generated by a conventional processing technique (milling). A linear dependency between a specific part of the heat-influenced zone and the corresponding maximum tensile load is found. A reduced load bearing area, as a consequence of a modified fibre–matrix-structure due to laser impact, is identified as the responsible factor for reduced tensile strengths, especially for low feed rates.

Introduction

Due to their superior specific strength and stiffness, carbon fibre reinforced plastics (CFRPs) are considered to have great potential for weight reduction in the aerospace industry as well as the automotive and energy sectors. Limited automation techniques are one important aspect to be investigated in order to drive down costs in mass production of CFRP-based goods. ^1,2 Although thermoplastic-based composites offer some processing advantages compared with reinforced thermoset-based matrix materials, thermosets are still the predominant form of CFRP. The thermoplastic-based materials are weldable, formable, easy to store and recyclable. ³ Hence, their use increases, and the manufacturing processes for these composites must be further developed.

In contrast to the fabrication of CFRP-parts, laser processing is already common in metal parts fabrication. The major benefits that promote laser use in production are flexibility for three-dimensional machining and its noncontact operation. Hence, tool wear is no issue with lasers, and they offer a process free of forces. Furthermore, high-power lasers allow high feed rates but, as a thermal process, during laser cutting, heat-induced damage appears at the cutting edge of processed CFRP. This heat affected zone (HAZ) of the workpiece is evident for CFRP due to different melting temperatures and the thermal conductivity of the matrix material and the reinforcing carbon fibres. In order to determine the laser-induced influence on the mechanical properties of CFRP workpieces, static tensile strength data must be provided as an important input parameter for the construction steps.

The influence of cutting parameters on the HAZ of CFRP with a thermoset matrix has been investigated for different laser sources. In the study by Riveiro et al. ^4,5 the influence of different pulse shapes of a CO₂ laser on the HAZ and the kerf width was explored, while in the studies by Garcia et al. ^6,7 and Tuersley and Pashby ⁸ pulsed Nd: YAG-lasers were used. The influence of a CO₂ laser on the HAZ and the kerf width with varying laser power and feed rates was investigated. ^9,10 In the study by Jaeschke et al., ¹¹ analysis of the HAZ with varying laser power and feed rate at constant energy per unit length using a continuous wave (cw) disk laser was performed. Efforts to develop models to predict the kerf width and the HAZ in dependence of different laser cutting parameters were studied ^{12 –16} and compared with the results from experiments. Here, the developed models agree with corresponding experiments.

For the thermoplastic-based composites, the majority of investigations have focused on CO₂ laser machining. A correlation between feed rate, material thickness, kerf width and HAZ has been demonstrated ^3,17 using CO₂ lasers. It is shown that the thermal impact on the cutting edge is a function of the fibre orientation. For unidirectional noncrimp fabrics, the highest damage occurs for carbon fibre tows showing a perpendicular orientation relative to the cutting direction. In the study by Tagliaferri and coworkers, ^{18 –20} CO₂ lasers were used and the experiments were compared with models to predict heat-induced damage of the thermoplastic-based composite. It was shown that the quality of the cutting kerf increases and the extent of the HAZ decreases the more similar the thermal material properties of polymer matrix and reinforcing fibres are.³ Therefore, the cutting results for glass- or aramide fibre-reinforced composites are better than for CFRPs.

In contrast to the investigations on CO₂ laser machining, few investigations have been published on cutting thermoplastic CFRP with a high power near infrared (NIR) laser operated in cw mode. ^1,21 Until now, the impact of thermal loads on the mechanical properties of laser processed reinforced composites has been hardly discussed. In the study by Howarth and Strong, ²² the mechanical properties of aramide fibre-reinforced thermosets were tested after laser cutting. The mechanical properties of carbon fibre-reinforced thermosets were investigated by Herzog et al. ²³ But the impact of laser cutting on the interlaminar shear strength properties of carbon fibre-reinforced thermoplastics was only studied by Jaeschke et al. ²¹ In this article, corresponding interlaminar shear strength tests reveal a reduction in the maximum shear stress for laser cut specimens compared with conventionally treated samples, provided by milling and abrasive water jet technique. Within the laser working range, an increase in the achievable shear strength for a certain material thickness with a decreasing HAZ was found. For comparably thin laminate arrangements, interlaminar shear strengths of laser-machined specimens converged with those of conventionally treated CFRP samples. Further studies correlating the HAZ with static strength properties are not known in the literature.

Experimentation

Material characterization

The laminate used within the frame of the investigations is a consolidated thermoplastic composite material based on a semicrystalline polyphenylene sulphide (PPS) matrix in conjunction with reinforcing high tenacity fibres. The organic sheets consist of 5-harness satin carbon weaves with an orthotropic fabric design in a 0°/90° fibre orientations [(0,90)]_3s/(0,90), with a thickness of $s$  = 2.17 mm. The fibre content is 50% with a mass per unit area of the fabric of 285 g/m². The material offers outstanding toughness and excellent chemical and solvent resistance. It is inherently flame resistant with low smoke emission All acronyms can be found in Appendix I.

Cutting setup and procedure

The cutting experiments were performed using a multimode fibre laser emitting in the NIR range of the electromagnetic spectrum at a wavelength of $\mathit{\lambda }$  = 1070 nm. The laser was run in cw mode and provided a maximum output power of $P_{\mathrm{L},max}$  = 6000 W. The laser radiation was guided by an optical fibre with a diameter of $d_{\mathrm{o}\mathrm{p}\mathrm{t}}$  = 200 µm to a laser working head. This working head consisted of a fibre coupling device in conjunction with collimating optics in order to parallelize the laser beam. The laser beam then passed adapted focusing optics with a focal length of $f$  = 200 mm. With this configuration, a laser beam focus of $d_{\mathrm{f}}$  = 200 µm was generated and focused on the surface of the workpiece. As process gas, N₂ at a working pressure of $p_{{\mathrm{N}}_2}$  = 0.9 MPa was applied. The cutting process was performed by moving the specimen relative to the laser working head using a high-precision linear axis.

Material testing

Tensile tests according to the standard DIN EN 2561 were performed. During the tests, specimens were prepared to be tested with the force along the warp direction of the composites. For reasons of comparison, reference specimens were prepared by milling. The specified test method evaluates the ultimate tensile strength ( ${\mathit{\sigma }}_{\mathrm{T}}$ ) and the ultimate tensile strength related to the fibre ( ${\mathit{\sigma }}_{\mathrm{F}}$ ). Therewith, the maximum tensile stress of each individual specimen at the moment of the first failure is calculated. First failure occurs, if the specimen fails due to tensile load. In order to emphasize the influence of edge effects due to laser impact, a reduced sample width was used, in contrast to DIN EN 2561. Therefore, specimens at a length of $l_0$  = 200 mm and a width of $w_0$  = 10 mm were prepared. The length of the specimen was parallel to the warp direction of the reinforcing fabrics. Each test result was verified by testing five identical specimens. All tests were performed at standard atmosphere. In order to examine the cutting zone, optical micrographs were prepared.

Results

In order to analyse mechanical properties with respect to laser machining parameters, the static strength properties and the HAZ were quantified and then correlated. Two parameter series were analysed, one with varying feed rate ( $v$ ), the second with varying laser output power (P _L). In the Investigations on the Static Strength Properties section, the impact of both the series on the static strength properties ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{L}}$ is determined, and in the Analysis of Heat Affected Zone section, the mean width of the respective heat-influenced area ${\mathit{\alpha }}_{\mathrm{k}}$ is investigated. With reference to the laser cut samples, additional specimens were prepared by milling.

Investigations on the static strength properties

In order to investigate the influence of the feed rate on the mechanical properties of laser cut CFRP specimens, static tensile strength tests were conducted. During the investigations, the feed rate was varied within a range of $v$  = 1.2–13.1 m/min at a constant laser output power of $P_{\mathrm{L}}$  = 6000 W. It was discovered that the static tensile strength of the laser cut specimens increases linearly with the feed rate (Figure 1).

Figure 1.

Tensile strength as a function of feed rate.

With a feed rate of $v$  = 1.2 m/min, the specimens achieved a static tensile strength of ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{L}}$  = 468 MPa, while the maximum feed rate of $v$  = 13.1 m/min led to a tensile strength of ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{L}}$  = 585 MPa. The results show an increase in the strength by 25%, when the feed rate is increased from slowest to fastest. However, laser-treated specimens do not yet reach the static tensile strength of ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{M}}$  = 640 MPa of milled specimens.

Varying the laser output power from $P_{\mathrm{L}}$  = 550 to 6000 W, with a constant feed rate of $v$  = 1.2 m/min, the static tensile strength of the prepared specimens remained at a level of ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{L}}$  = 480 MPa. Consequently, it can be stated that in contrast to the feed rate, the laser output power has no major impact on the static tensile strength of laser cut specimens (Figure 2).

Figure 2.

Tensile strength as a function of laser power.

Analysis of HAZ

Analyzing the cross sections of laser cut PPS-based CFRP laminates, three different HAZs can be defined (Figure 3). In these regions, the heat-induced damage appears in different forms. In the cutting kerf (w _k), the reinforcing carbon fibres and the PPS matrix are vaporized. Close to the cutting kerf, the section nominated as A ₁ is affected by temperatures exceeding the level, where the matrix is supposed to be vaporized and the fibre tows are charred. In the adjacent area A ₂, it appears that the matrix damage predominantly occurs due to the temperatures exceeding the PPS decomposition temperature of $T_{\mathrm{V}}$  = 370°C. In the final area A ₃, interlaminar damage is primarily visible, arising from structural modifications within the polymer matrix between the rovings. In particular, porosities within the PPS matrix depots between the carbon fibre rovings are observed in this section. Analogue to the inclusion of pores occurring in imperfect thermoforming processes of organic sheets due to insufficient moulding pressure, the authors conclude that the melting temperature for PPS of $T_{\mathrm{M}}$  = 285°C is reached in the respective areas. Therefore, it can be stated that the interface between the HAZ and the unmodified bulk material is attributed to the melting and resolidification of the polymeric PPS matrix.

Figure 3.

Identification of different regions of the heat-influenced area due to laser impact.

A quantitative analysis of the HAZ was conducted by taking cross section images of the HAZ and measuring each heat-influenced area ( $A_{\mathrm{k}}$ ). Due to the heat-induced burr at the cutting kerf, the heat-influenced area ( $A_{\mathrm{k}}$ ) was measured within the borders of the original laminate thickness s = 2.17 mm, (Figure 3). With this data, a mean width a _k of the respective heat-influenced area was calculated and a mean width ${\mathit{\alpha }}_{\mathrm{k}}$ beginning at the cutting edge can be derived from the following equation

${\mathit{\alpha }}_{\mathrm{k}}=\frac{1}{s}\cdot \sum _{i=1}^k{A}_i$

1

The mean kerf width $w_{\mathrm{k}}$ was also computed by dividing the kerf area $W_{\mathrm{k}}$ by the laminate thickness $s$  = 2.17 mm. For each combination of working parameters, micrographs of three laser cuts were prepared. For increasing feed rates $v$ at a constant laser output power of $P_{\mathrm{L}}$  = 6 kW, each of the affected sections and, therefore, the mean width ${\mathit{\alpha }}_{\mathrm{k}}$ decreases linearly (Figure 4). In fact, each mean width ${\mathit{\alpha }}_{\mathrm{k}}$ was nearly halved as the feed rate increased from the minimum rate of $v$  = 1.2 m/min to the maximum rate of $v$  = 13.1 m/min. Also, the kerf width $w_{\mathrm{k}}$ decreases with increasing feed rates $v$ until it reaches the width of the laser focus of $d_{\mathrm{f}}$  = 0.2 mm at a feed rate of $v$  = 7.2 m/min.

Figure 4.

Mean width of heat-influenced area and kerf width as a function of the feed rate.

Altering the laser output power $P_{\mathrm{L}}$ at a constant feed rate of $v$  = 1.2 m/min, the mean width of the respective heat-influenced area ${\mathit{\alpha }}_{\mathrm{k}}$ rose only slightly (Figure 5). Accordingly, it can be stated that the different mean widths ${\mathit{\alpha }}_{\mathrm{k}}$ remain constant for changing laser output powers $P_{\mathrm{L}}$ . In contrast to the mean width ${\mathit{\alpha }}_{\mathrm{k}}$ , the kerf width $w_{\mathrm{k}}$ increases linearly with the laser output power $P_{\mathrm{L}}$ .

Figure 5.

Mean width of heat-influenced area and kerf width as a function of laser power.

Correlation of static strength properties and the HAZ

In this chapter, the static tensile strength ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{L}}$ and the mean widths ${\mathit{\alpha }}_{\mathrm{k}}$ of the different heat-affected sections are correlated. Their tensile properties are compared with a milling process, as a reference emitting only low thermal loads. For this reason, the thermal impact of the laser radiation on the static tensile properties is investigated.

By changing laser output power $P_{\mathrm{L}}$ at a constant feed rate $v$ , it was found that the static tensile strength ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{L}}$ and the heat-influenced area ${\mathit{\alpha }}_{\mathrm{k}}$ remain constant as illustrated by plotting the static tensile strength ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{L}}$ against the three defined mean widths ${\mathit{\alpha }}_{\mathrm{k}}$ (Figure 6). The static tensile strength of the milled specimens of ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{M}}$  = 640.3 MPa was not reached.

Figure 6.

Tensile strength as a function of mean width of the heat-influenced area at varying laser power.

In contrast, varying the feed rate $v$ led to a linear relationship for each mean width ${\mathit{\alpha }}_{\mathrm{k}}$ illustrated by plotting the different mean widths of the heat-influenced area ${\mathit{\alpha }}_{\mathrm{k}}$ against the static tensile strength ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{L}}$ (Figure 7). With a decreasing width ${\mathit{\alpha }}_{\mathrm{k}}$ , the static tensile strength ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{L}}$ of the tested specimens increases.

Figure 7.

Static tensile strength as a function of mean width of the heat-influenced area for varying feed rates.

Supposing that a section ${\mathit{\alpha }}_{\mathrm{k}}$ has no contribution to the static tensile strength, the extrapolation ${\mathit{\alpha }}_{\mathrm{k}}$  = 0 would provide the static tensile strength of the bulk material. Following this assumption using a simple linear regression to fit the parameters, the static tensile strength ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{L}}$ ( ${\mathit{\alpha }}_{\mathrm{k}}$  = 0) can be extrapolated for a zero mean width of the heat-influenced area ( ${\mathit{\alpha }}_{\mathrm{k}}$  = 0). The resulting tensile strengths of ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{L}}$ ( ${\mathit{\alpha }}_1$  = 0) = 627.8 MPa and of ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{L}}$ ( ${\mathit{\alpha }}_2$  = 0) = 633.8 MPa are similar to the tensile strength of milled specimens with ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{M}}$  = 640.3 MPa. In contrast for ${\mathit{\alpha }}_3$  = 0 mm, a significantly higher value of ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{L}}$ ( ${\mathit{\alpha }}_3$  = 0) = 683.8 MPa was calculated. This indicates that the area ${\mathit{\alpha }}_3$ contributes to the tensile strength. Hence, a further optimization of the static tensile strength ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{L}}$ of laser cut workpieces can be accomplished by a further reduction in the mean width of the heat-influenced area ${\mathit{\alpha }}_{\mathrm{k}}$ . In accordance to Analysis of Heat Affected Zone section, the width ${\mathit{\alpha }}_{\mathrm{k}}$ can be reduced by increasing the feed rate $v$ .

As shown in Figure 7, an increasing width of the heat-influenced area ${\mathit{\alpha }}_{\mathrm{k}}$ leads to a reduced tensile strength ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{L}}$ . The different regions of the heat-influenced area $A_{\mathrm{k}}$ of laser cut specimens are suspected to be unable to partly or fully transmit tensile loads $F$ , due to a shortened or missing fibre–matrix adhesion. Consequently, the load transmitting cross section areas of the tested laser-treated laminates are reduced by the heat-influenced areas $A_{\mathrm{k}}$ on each cutting edge of the specimen. In order to display the impact of this reduced load bearing cross section area on the tensile load $F$ , milled specimens with a reduced width were prepared and tested.

In Figure 8, the maximum tensile load $F$ is plotted against the width reduction $\mathit{\chi }$ of laser cut and milled specimens. In case of laser-treated specimens, the width reduction refers to the laser-induced thermal damage that appears on both the cutting edges of the laminates being unable to transmit tensile loads, as given in equation (2)

Figure 8.

Maximum tensile load as a function of width reduction with varying feed rates.

$\mathit{\chi }=2\cdot {\mathit{\alpha }}_{\mathrm{k}}$

2

Additionally, the tensile loads $F$ of the width reduced milled specimens are plotted in the graph. Their original width of $w_0$  = 10 mm was reduced by $\mathit{\chi }$  = 0–2.2 mm in steps of $\mathrm{\Delta }\mathit{\chi }$  = 0.2 mm. The width reduction $\mathit{\chi }$ is drawn on the x-axis. Additionally, a ratio of the reduced width to the original width is calculated in equation (3) and displayed in a second x-axis in Figure 8.

$\mathit{\theta }=\frac{w_0-\mathit{\chi }}{w_0}$

3

This is performed to compare the impact of the width reduction $\mathit{\chi }$ , due to thermal damage induced by the laser radiation, with the influence of a width reduction on the tensile load $F$ of milled specimens.

It is obvious that the lines of best fit due to the mean width ${\mathit{\alpha }}_1$ and ${\mathit{\alpha }}_3$ do not run parallel to the line of best fit of the milled specimens, while for ${\mathit{\alpha }}_2$ it does (Figure 8). The line of best fit of ${\mathit{\alpha }}_2$ is also closest to the milled specimens. Consequently, it can be assumed that ${\mathit{\alpha }}_2$ best represents the width of laser cut CFRP that do not transmit tensile loads.

In Figure 9, the reduction in the load bearing area due to the subtraction of the heat-influenced area ${\mathit{\alpha }}_2$ is illustrated. With the maximum tensile loads, the ${\mathit{\varphi }}_{\mathrm{T}}$ ratio of static tensile strength of laser-treated ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{L}}$ to milled specimens ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{M}}$ is calculated. For increasing feed rates $v$ , a rising ratio of static tensile strength properties ${\mathit{\varphi }}_{\mathrm{T}}$ , up to ${\mathit{\varphi }}_{\mathrm{T}}$  = 0.91 for feed rates of v = 13.1 m/min, can be observed (Figure 10).

Figure 9.

Reduction in load bearing area of specimens due to laser impact.

Figure 10.

Ratio of tensile strength and ratio of actual tensile strength for laser machined specimen to those of milled samples as a function of feed rate.

Due to the reduction of the load bearing area, resulting from the mean width of the heat-influenced area ${\mathit{\alpha }}_2$ (Figure 9), an effective ratio of the tensile strength ${\mathit{\varphi }}_{\mathrm{T}}^{\ast }$ between the lasers processed on milled specimen can be expressed as

${\mathit{\varphi }}_{\mathrm{T}}^{\ast }=\frac{{\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{L}}^{\ast }}{{\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{M}}}$

4

where ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{L}}^{\ast }$ represents the actual tensile strength of laser-treated specimens, and ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{M}}$ the tensile strength of the milled specimens (equation (5)). With the maximum load applied at failure $F$ and with s as the thickness of the laminate, ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{L}}^{\ast }$ can be formulated as

${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{L}}^{\ast }=\frac{F}{s\cdot w^{\ast }}$

5

The relationship between the real width $w_0$ of the specimens and the effective width $w^{\ast }$ is given in equation  (6).

$w^{\ast }=w_0-2\cdot {\mathit{\alpha }}_2$

6

Taking only the effective width $w^{\ast }$ of the laser cut laminates into account, when calculating the ratio of actual tensile strength ${\mathit{\varphi }}_{\mathrm{T}}^{\ast }$ of laser cut to milled specimens, values between ${\mathit{\varphi }}_{\mathrm{T}}^{\ast }$ = 0.92 and ${\mathit{\varphi }}_{\mathrm{T}}^{\ast }$ = 0.98 are obtained (Figure 10). Hence, the actual tensile strength ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{L}}^{\ast }$ of laser processed specimens is almost at the level of milled laminates ${\mathit{\sigma }}_{\mathrm{T}\mathrm{,}\mathrm{M}}$ , when referring to the tensile loads only on the undamaged, load transmitting areas as limited by the mean width ${\mathit{\alpha }}_2$ .

Conclusions

In this article, investigations on whether thermal damage caused by high-power laser processing reduces the static tensile strength properties of continuous carbon fibre reinforced thermoplastics are given. To this end, laminates consisting of 5-harness satin carbon weaves in conjunction with a semicrystalline PPS matrix were cut using a multimode fibre laser.

It was demonstrated that the thermal cutting process induces a HAZ within the composite laminate, starting at the cutting edge. It turned out that the HAZ could be described as a modified composite structure, consisting of three different areas, each of them representing a zone where damage occurs in a different manner. Within the frame of the investigations, it was found that the cutting velocity significantly determines the extent of all three areas of the HAZ. The higher the feed rate, the smaller the thermal damage at the cutting edge.

The resulting HAZ was correlated with corresponding static tensile strength measurements. In this context, specimens which were cut by laser at varying feed rates were tested with respect to tensile load. A comparison with samples prepared by milling revealed the reduction in the load bearing area due to thermal influence as the main factor for a reduction in tensile strength found for laser-treated composite laminates. In order to evaluate the decrease in tensile strength, an effective width was introduced, taking into account the remaining load transmitting width of laser cut specimens.

For future investigations, it is intended to further reduce or even to eliminate the HAZ. Taking into account that high feed rates decrease the extent of the HAZ, an additional augmentation of the velocity of the laser beam seems to be promising. Therefore, multiple crossings of the cutting outline are necessary to achieve the required energy per unit length to cut the CFRP. The application of a quasi-simultaneous cutting process using a laser scanning head could lead to the desired result.