Abstract
Keywords
Introduction
Carbon fiber reinforced high-performance thermoplastic composites possess great application potential duo to the short curing cycle, excellent damage tolerance, impact resistance, easy maintenance, and recycling potential, which has attracted wide attention in the aviation field. 1 However, high performance thermoplastic polymers are usually accompanied by high viscosity and the consolidation process requires high temperature and pressure. For traditional processing method such as autoclave, the requirements for equipment and accessory materials are further improved, thus, the manufacturing cost of thermoplastic composites increased. With the increasing demand for composite materials in the aviation field, it is necessary to develop efficient and low-cost composite manufacturing processes.
Automated laying technology is a low-cost composite additive manufacturing process. 2 According to the width of the material used, automated laying technology involves automated tape laying (ATL) and automated fiber placement (AFP). 3 Automated laying technologies greatly improve the manufacturing efficiency of composite and they are particularly suitable for the manufacturing of complex, large-scale components, and can reduce the number of connected components.4,5 This greatly reduces the manufacturing cost of composite and reduces the weight of structural parts, which has a broad application prospect in the aviation field.6,7
For the automated laying of thermoplastic composites, the most commonly used is the “two-step” manufacturing process which combined with autoclave consolidation.8,9 During the step of autoclave consolidation, the problem of high manufacturing cost still exists. Thus, the “one-step” processing called “in situ consolidation (ISC)” is of great significance. During the ISC process, the dwell time of temperature and pressure is extremely short, usually lasts for only few seconds, it’s difficult to reach perfect interlaminar bonding. Thus, it’s necessary to understand the processing mechanisms and optimize the process parameters to improve the performance of the composite.
The processing mechanisms during ISC involves heat transfer,10,11 polymer degradation, 12 intimate contact,13,14 polymer healing,15,16 void dynamics, 17 polymer crystallization,18,19 and residual stress.20,21 To better understand these mechanisms, several theoretical models were established while extensive experimental research were performed. These research aims to investigate the key parameters affected the mechanical properties.
In the process of automated laying in-situ consolidation process, parameters such as lay-down speed, tool temperature, heating temperature, compaction pressure have direct impact on the thermal degradation, interlaminar bonding, void content and crystallinity of the composite, and then affect the mechanical properties of the composite.
Stokes-Griffin and Compston 22 explored the effect of laser heating temperature and lay-down speed on the interlaminar shear strength (ILSS) of CF/PEEK composites prepared by laser-assisted automated fiber placement (LAFP) in situ consolidation, then compared with the composites prepared by autoclave. They found that the ILSS obtained by LAFP in situ consolidation was equal to autoclave at slow placement and the temperature where thermal degradation does not occur. Ray et al. 23 et al. compared the crystallinity and mode І fracture toughness of the CF/PEEK composites manufactured by AFP in-situ consolidation, the results showed that the crystallinity of the composite from AFP in-situ consolidation is far lower than that of autoclave, however, the fracture toughness is 60%~80% higher than that of autoclave. Chu et al. 24 investigated the properties of the GF/PP composites manufactured by ultrasonic-assisted automated fiber placement (UAFP) and hot-press. The results showed that the crystallinity of the composite from UAFP was lower than that of hot-press, and the ILSS of the composite can match the hot-press ones. And the mode І fracture toughness and impact toughness of the UAFP composite were 59.9% and 20.1% lower than the hot-press ones, respectively. Miao et al. 25 reported that the void content decreases with the increase of compaction pressure, which can improve the interlaminar properties.
At present, the properties of automated laying in-situ consolidation can only reach about 80%–90% of that of autoclave. 2 Eliminating the properties gap becomes an urgent problem for further application of automated laying in-situ consolidation. In order to better understand the mechanisms of the properties gap of automated laying technology, in this study, a comparative study of laser-assisted automated fiber placement (LAFP) and hot-press was carried out using high-performance carbon fiber reinforced polyphenylene sulfide (CF/PPS) thermoplastic composites. Several properties and their mechanisms were compared and analyzed.
Processing method
Materials
The continuous carbon fiber reinforced polyphenylene sulfide (CF/PPS) unidirectional tape (Provided by Barrday Co.Ltd) was used for laminates manufacturing. The width and thickness of the tape were 6.35 and 0.14 mm, respectively. The resin content was 34% by weight, and the density was 1620 kg/m3. Also, according to the Technical Data Sheet, the glass transition temperature of the tape was 95℃.
Laminates manufactured by LAFP
Twenty-plies unidirectional laminate ([0]20) was first manufactured by LAFP. The four-tows robot-style laser-assist placement machine was used (Designed by Mtorres, Spanish). A diode laser was integrated in the heating system, and the maximum heating temperature that can be reached was 550℃. The manufacturing process was monitored remotely by CCTV, simultaneously, the temperature history of the layer surface was recorded by an infrared camera. A rigid compaction roller was used.
There were a multitude of processing parameters that affected the laminate quality, like laser temperature, compaction force, lay-down speed and tool temperature, et al. In this study, the relative parameters were chosen as recommended of the equipment supplier, as shown in Table 1. Before placement, a thin polyimide film was fixed on the tool, then [0]20 unidirectional laminate with a size of 300 mm × 150 mm × 3 mm was manufactured. The manufacturing process of CF/PPS laminate by AFP in-situ consolidation was shown in Figure 1. Aluminum foil was inserted between layer-10 and layer-11 to make a 50 mm initial pre-crack for mode Ⅰ fracture toughness test.
Processing parameter of the laminate manufactured by LAFP.

Manufacturing process of CF/PPS laminate by AFP in-situ consolidation.
Laminates manufactured by hot-press
Twenty-plies unidirectional laminate ([0]20) were produced by hot-press (HBSCR, Huabo Machinery, China) using standard manufacturing procedures provided by the tape supplier. The consolidation temperature and pressure were 330℃ and 1.5 MPa. During the consolidation process, the laminate was first heated to 330℃ and then held for 30 min. Simultaneously, the laminate and tool were cooled down to the room temperature by furnace cooling. The cooling rate during cooling stage was calculated by the temperature profile recorded by the thermocouple embedded between the laminate and tool. The manufacturing process of CF/PPS laminate by hot-press was shown in Figure 2.

Manufacturing process of CF/PPS laminate by hot-press.
Characterization
Thermogravimetric analysis (TGA) of the CF/PPS tape was performed using TGA 8000 (PE, America) under a nitrogen atmosphere. Samples weight was 3–5 mg and the purge gas flow was 20 ml/min. The samples were heated from 30℃ to 800℃ with heating rates varied from 5℃/min to 50℃/min. Before testing, all samples were dried for 6 h in a vacuum oven to remove water to ensure the accuracy of the experimental data.
The surface profile of the laminates was scanned by Bruker DektakXT roughness meters to assess the smoothness. A minimum of five samples for the laminates manufactured by different method were tested and the surface roughness values were estimated at five locations. Then the average deviation of surface profile height Ra was used to to represent the roughness value.
Micro morphology of the laminate was observe using Hitachi SU 8010 field emission scanning electron microscope (SEM). The acceleration voltage was 1 kV. The samples were sprayed with gold before the test.
The crystallinity of the laminates was calculated from differential scanning calorimetry (DSC) curves of the samples. DSC tests were performed on DSC PE 8500 under a nitrogen atmosphere with a heating rate of 10°C/min. Samples were weighed about 5 mg into aluminum pans and then heated from 40°C to 330°C. To ensure the consistency of data, the samples were cut at the same position in the center of the laminates. Then, the crystallinity of the laminates was calculated using following equation:
Where
The 2D cross-sectional microscopy was performed to evaluate the void content of the laminates manufactured by LAFP and hot-press. The cross section of the samples were observed and photographed using the microscope (Olympus BX41M-LED, Japan) after polishing and polishing. Then the void content was calculated through image analysis software. Five locations were selected for analysis of each laminate.
Mechanical test
Several mechanical tests were carried out to evaluate the mechanical properties of the laminates manufactured by LAFP and hot-press: (a) short beam shear test for interlaminar shear stress (ILSS, ASTM D-2344), (b) three-point bending test for flexural strength (ASTM D-7264), (c) double cantilever beam test for mode Ⅰ fracture toughness (ASTM D-5528). All tests were performed on a ETM 204C electro-mechanical universal machine.
For the short beam shear test, the dimension of the samples was 18 mm × 6 mm × 3 mm. The displacement rate of the cross-head was 1 mm/min. Five samples were tested for each laminate, the ILSS was determined as follows:
Where
For the three-point bending test, the dimension of the samples was 140 mm × 13 mm × 3 mm. The span thickness ratio was 32:1. The displacement rate of the cross-head was 2 mm/min. Five samples were tested for each laminate, the ILSS was determined as follows:
Where
For the double cantilever beam (DCB) test, the dimension of the samples was 140 mm × 25 mm × 3 mm. The displacement rate of the cross-head was 2 mm/min. Five samples were tested for each laminate, the mode Ⅰ fracture toughness
Where
Results and discussion
In order to better explain the influence of different processing methods on the properties of the laminates, the properties of CF/PPS prepreg and the temperature history during the process of LAFP in situ consolidation and hot-press were first discussed.
Figure 3 shows the thermogravimetric curves and first-order differential thermogravimetric curves of CF/PPS composites in nitrogen at different heating rates. It can be seen from the figure that at a constant heating rate, the CF/PPS composite is a one-step weight-loss process. With the increase of the heating rate, the initial weight-loss temperature of the composite moves to the high temperature zone. When the heating rate increases from 5℃/min to 50℃/min, the initial thermal degradation temperature increases from 409℃ to 493℃. At the same time, the maximum weight loss rate temperature of the composite also increases with the increase of the heating rate, which was between 500℃ and 600℃. The maximum weight loss at different heating rates is basically the same, about 23%. For present study, both processing temperatures of LAFP in situ consolidation (380℃) and hot-press (330℃) were lower than the initial thermal degradation temperature of the prepreg, thus, the performance of the laminate will not be affected by thermal degradation.

Thermogravimetric curves (a) and differential thermogravimetric curves (b) of CF/PPS composites at different heating rates.
Figure 4 was surface temperature record by infrared camera during LAFP in situ consolidation. The result showed that the surface temperature during LATP of the layer changed quite quickly. The heating and cooling process was completed within the time scale of the seconds. The heating rate and cooling rate exceed 1500℃/min and 600℃/min, respectively. Simultaneously, the time when the surface temperature of the layer above its melting temperature was less than 0.1 s. Since the laser heat source and the compaction roller moved at the same time during the consolidation process, the holding time of the pressure was close to the temperature. This result indicated that there was not enough time for the interlaminar bonding and defects eliminating during the process LAFP in situ consolidation. In addition, the external, internal structure and various properties such as crystallinity and void content of the laminate will also be greatly affected, which would be discussed in the following section. Figure 5 was the surface temperature during cooling stage recorded by thermocouple during hot-press. The difference in the temperature history between the two processing methods will lead to the difference in the properties of the laminates.

Surface temperature record by infrared camera during LAFP in situ consolidation.

Surface temperature during cooling stage recorded by thermocouple during hot-press.
Characterization of the laminates
In order to better understand the influence of processing parameters of LAFP in-situ consolidation on properties and its mechanism, the properties of the laminates manufactured by two forming methods were compared and analyzed.
Firstly, the surface quality of the laminates was analyzed. The surface profile and average roughness of the laminate manufactured by LAFP and hot-press were shown in Figure 6. It was clear that the surface profile height and the corresponding surface roughness of the laminate manufactured by LAFP were significantly higher than that of the hot-press. This was mainly because that during the LAFP in situ consolidation process, the resin flows after melting, and the time of heating and compaction of the layer was very short, resulting in the layer cannot form a smooth and flat surface. Chanteli et al. 27 et al investigated the influence of repass treatment and autoclave treatment on the surface roughness of the CF/PEEK laminate manufactured by laser-assisted automated tape placement (LATP). The repass treatment refers to applying heat and pressure to the laminate through the ATP head without adding new tapes. They found that after once and twice repass treatment, the surface roughness of the laminate reduced from 4.88 to 2.36 and 2.12 μm, respectively. The result indicated that the surface quality of CF/PEEK composite can be significantly improved by repass treatment. However, compared with the sample treated by autoclave (with a surface roughness of 0.48 μm), even the sample after twice repass treatment has a rougher surface. This result was similar to that of present study. For the process of autoclave and hot-press, the longer heat preservation and pressure retention time were conducive to forming a flat surface.

Surface profile (a) and average roughness (b) of the laminate manufactured by LAFP in situ consolidation and hot-press.
The DSC curves of the laminates manufactured by AFP in-situ consolidation and hot-press were shown in Figure 7. It was obvious that the DSC curves of the laminates manufactured by AFP in-situ consolidation existed a cold crystallization peak at the temperature range of about 120℃–150℃. The result indicated that the crystallization of PPS was incomplete. On the contrary, the cold crystallization peak disappeared for the laminate manufactured by hot-press. According to the temperature history of the processes (Figures 5 and 6), the cooling rates of the two processes were quite different, as shown in Table 2. For the two processes, the cooling rates were >10,000℃/min and ~2.5℃/min. At an extremely high cooling rate, PPS molecular chains did not have enough time to arrange regularly. The crystallinity of the two laminates were 21.6% and 46.3%. It was reported that 19 during the process of AFP in situ consolidation, the substrate under the heat source would undergo a cold crystallization process due to the repeated heating, resulting in an increase of crystallinity. However, in this study, the crystallinity of the laminate was still low.

DSC curve of CF/PPS laminates manufactured by LAFP in-situ consolidation and hot-press.
Crystallinity obtained by DSC of the laminates manufactured by AFP in situ consolidation and hot-press.
The 2D cross-sectional morphology of the two type laminates were shown in Figure 8. The calculated void content was 2.75% and 0.63%, respectively. It was also found that the void size was bigger and the distribution range was larger for the laminate manufactured by LAFP in situ consolidation compared to the laminate manufactured by hot-press. For the thermoplastic composites, the void was flows with the fiber-resin mixture after melting and then discharged under pressure during the processing, which means filling the interlayer pores requires macro flow of the fiber-resin mixture. During the LAFP in situ consolidation, on the one hand, the processing time was quite short, and the melting viscosity of the resin was high, thus, the macro flow of the fiber-resin mixture was difficult to complete in such a short time. On the other hand, the compaction roller was directly above the heating area, which the elimination of pores was limited. These two reasons lead to high void content of the laminate manufactured by LAFP in situ consolidation. On the contrary, for the laminate manufactured by hot-press, the fiber-resin mixture has enough time to flow adequate, resulting in a low void content. Table 3 shows the void content made by AFP/ATP situ consolidation process in the literature.

Cross-sectional morphology and related void content of the laminates manufactured by (a) LAFP in situ consolidation and (b) hot-press.
Void content made by AFP/ATP* situ consolidation process in the literature.
ATP: Automated Tape Placement.
Mechanical properties
Figure 9 showed the load-displacement curves and interlaminar shear strength (ILSS) of the laminates manufactured by LAFP in situ consolidation and hot-press.

Load-displacement curves (a) and flexural strength (b) of the laminates manufactured by LAFP in situ consolidation and hot-press.
It can be seen that the ILSS of the laminate manufactured by LAFP in situ consolidation was 45.8 MPa, which was 34.9% lower than that of hot-press (70.4 MPa). According to above results, the crystallinity of the laminate manufactured by LAFP in situ consolidation was lower than the laminate manufactured by hot-press while the void content was higher. Simultaneously, the interlaminar properties of the laminate was strongly affected by the interlaminar bonding during the manufacturing process. The interlaminar bonding including two main process, intimate contact and polymer healing. The theoretical model of intimate contact was establish by Lee and Springer. 34 The algebraic expression of the Lee’s model was:
Where
The process of polymer healing was described using a non-isothermal model 35 which algebraic expression was:
Where
From the above analysis, it can be concluded that several key parameters affecting interlayer bonding were pressure, temperature and its applied time, as well as polymer viscosity. For the process of LAFP in situ consolidation, although the processing temperature (380℃) was higher than that of hot-press (330℃), but its applied time was only a few seconds. Therefore, the degree of interlayer bonding was far lower than that of hot press.
Figure 10 showed the load-displacement curves and flexural strength of the laminates manufactured by LAFP in situ consolidation and hot-press. The result was similar to the result of ILSS. The flexural strength of the laminates manufactured by LAFP in situ consolidation and hot-press were 870 and 1265 MPa.

Load-displacement curves (a) and flexural strength (b) of the laminates manufactured by LAFP in situ consolidation and hot-press.
The ability of composite materials to resist delamination can be characterized using fracture toughness. Figure 11 showed was the R curve of the DCB test and Mode Ⅰ fracture toughness of the laminates manufactured by LAFP in situ consolidation and hot-press. The result revealed that the Mode Ⅰ fracture toughness of the laminates manufactured by LAFP was 2.18 kJ/m2, 103.7% higher than that of hot-press with 1.07 kJ/m2. It was reported that the deformation of the matrix contributes more than 75% of the interlaminar fracture toughness of the thermoplastic composites 36 and the deformation of the matrix was mainly depended on the crystallinity. During the process of LAFP, the rapid cooling leads to incomplete crystallization of the matrix, thus the plasticity was strengthened and when the delamination occurs, more energy was absorbed. The mechanism was further confirmed by the fracture morphology of DCB sample, as is shown in Figure 12. The ductile drawings on the fracture surface of DCB samples manufactured by LAFP in situ consolidation, indicating that significant plastic deformation of the matrix happened during the crack propagation. On the contrary, for the samples manufactured by hot-pressed, the matrix cracking was the main failure mode.

The R curve (a) of the DCB test and Mode Ⅰ fracture toughness of the laminates manufactured by LAFP in situ consolidation and hot-press (b).

Fracture morphology of DCB samples manufactured by LAFP in situ consolidation (a) and hot-press (b).
Conclusion
In this study, the properties of laminates manufactured by LAFP in situ consolidation and hot-press were compared and analyzed. Because of the tremendous difference in processing, the properties of the two laminates were quite different. Due to the rapid heating and cooling process of LAFP in situ consolidation, the dwell time of temperature and pressure of the layer was extremely short, resulting in lower crystallinity and higher void content of the laminate compared to the laminate manufactured by hot-press. For the mechanical properties, the ILSS and flexural strength of the laminate manufactured by LAFP in situ consolidation was lower than that of hot-press due to the lower crystallinity and interlaminar bonding and higher void content. On the contrary, the Mode Ⅰ fracture toughness was higher for the laminate manufactured by LAFP in situ consolidation because of the plasticity and the ability of deformation were strengthened.
