Temperature field analysis for automated placement with thermoplastic composites based on the line-focused infrared heater

Abstract

Infrared heating is one of the most prevalent heat sources utilized in automated fiber placement (AFP), particularly in the context of thermoset composites and the preheating of thermoplastic composites. Traditional infrared heaters often struggle to achieve concentrated heating in these applications. This paper employs a line-focused infrared heater (LIRH) in AFP. A dynamic finite element model was developed using ANSYS to investigate heat transfer during the layup process, and the interrelationship between temperature, laying speed, and LIRH power. The impact of tooling temperature on the cooling rate during layup was also analyzed, and an online measurement system for the temperature field was established. AFP is experimentally performed to analyze the heating performance of the LIRH to determine its relevant parameters. The results of both experiments and simulations indicate the prepreg bundle can achieve temperatures exceeding 400°C at a laying speed of 30 mm/s, with an infrared power setting of 1400 W and a tooling temperature of 50°C. The heat primarily emanates from the focused area of the LIRH, although a minor heating effect is observed from the unfocused infrared light. This research demonstrates the feasibility of employing LIRH for automated placement, particularly from the standpoint of temperature field.

Keywords

Automated fiber placement line-focused infrared heater temperature field CF/PEEK thermal model

Introduction

Fiber-reinforced resin matrix composites have excellent properties such as high-temperature stability, chemical resistance, fatigue endurance, and high specific strength and stiffness, which make them ideal materials for aerospace structural components.^1–6 Carbon fiber-reinforced structures have been found to have excellent thermodynamic properties and a role in lightweighting in aircraft structural applications.^7–9 However, the manufacture of fiber-reinforced plastic-formed components is usually complex, and suitable processing techniques are required to ensure good product performance. In recent years, considerable research has been devoted to developing promising automated manufacturing techniques for thermoplastic composites, such as fiber winding and AFP.¹⁰

Currently, AFP for thermoset composites has been successfully commercialized, but AFP for thermoplastic composites still needs to be sufficiently mature. According to the resin matrix material, fiber-reinforced resin matrix composites are classified as thermoset and thermoplastic composites. Due to the long curing cycle, thermoset composite components can only be manufactured with autoclave curing technology, leading to an increase in manufacturing costs and being limited by the size of the components and the processing site. These shortcomings have limited the application of thermoset composites. Compared with thermoset resins, the molecular chain of thermoplastic resins does not contain any groups that can be chemically reacted within the chains but rather undergo a physical change process of softening by heating and hardening by cooling, which results in a shorter forming cycle and higher production efficiency of thermoplastic composites.¹¹

When used with thermoplastic prepregs, the in situ consolidation is achieved by melting the prepregs and applying pressure. The primary sources of heat commonly used for AFP are hot gas torches,^12–15 infrared,^16–19 lasers,^20–26 and so forth. Compared with other heat sources, the energy utilization rate of hot gas torches is low, and the heat loss is severe.²⁷ Laser heat sources have high energy utilization and density but are physically large, require high cost, and must be placed in a sealed environment. IR lamps are relatively inexpensive and easy to control, allowing the melt zone temperature to be carefully regulated. When used for carbon-based materials, they are a very efficient heat source because the energy is almost completely absorbed.²⁸ Conventional infrared heaters do not provide concentrated energy, leading to heating unwanted materials and areas.¹ In contrast, LIRH can concentrate energy in a single line and increase the energy density several times. Therefore, focused infrared heaters have the potential to become one of the primary heaters for AFP of thermoplastic composites.

For hot gas torches in AFP, the heat load is applied in the form of forced convection. Kim et al.²⁹ developed a finite element code for heat transfer simulation during the layup of nitrogen torch-heated thermoplastic composite tapes, in agreement with experimental data. Cao et al.¹² developed a mathematical heat transfer model during the automated placement of thermoplastic composites and explored the effect of hot gas torch process parameters on the temperature field. Li et al.¹⁴ established a dynamic finite element model of the transient temperature field and compared it with a transient heat transfer model. They explore the temperature field of thermoplastic composite based on the hot gas torches. For laser in AFP, Suemasu³⁰ proposed a thermal history analysis method for evaluating the thermal history of thermoplastic composites during in situ consolidation by an automated layup technique. Kollmansberger³¹ was mainly concerned with modeling the heat transfer into the composite with a laser heat source, the thermal properties of the tape, and the resulting heat distribution in the tooling and compaction roller. Francesco²³ explored laser heater power control for multi-material, variable speed. In addition, Danezis et al.³² established a two-dimensional heat transfer model, which is capable of analyzing the temperature field variations in the time scale of short, high-energy pulses. Although laser and flashlamp heating provide high laying efficiency, their cost is hard to estimate. Lower deposition rates can enable out-of-autoclave manufacturing of medium-sized composite structures for other markets. Therefore, the infrared heater in the middle of the laser and hot air gun is a better choice.

For infrared heaters in AFP, Zhao et al.⁴ established a finite element temperature field model for tooling heating, infrared lamp preheating, and hot air torch heating in the automated layup process. They carried out automated layup experiments to determine the heating performance of the infrared lamp and hot air torch. Wang et al.¹⁸ presented a heater structure-based heating strategy to achieve constant temperature control in a variable speed layup process for thermoset composites. P. Hörmann et al.¹⁹ investigated the effect of the radiation distribution of infrared heaters on neighboring paths. The above authors mainly apply infrared heaters to the heating of thermoset composites and the preheating of thermoplastic composites. Venkatesan is one of the few authors to have applied infrared heaters as primary heaters for the automated layup of thermoplastic composites. However, in his study,^33,34 the thermoplastic composite material CF/PA6 with a melting point of 220°C was selected, and at 40% IR power (100W), it was optimized for a laying speed of just 5.94 mm/s.

From this, more scholars determine the boundary conditions based on the heating method and environment, and most of the heat sources for temperature field analysis come from lasers and hot gas torches. In contrast, most infrared heating temperature field analysis is used in the automated placement process of thermoset composites or preheating of thermoplastic composites. Analysis of the temperature field of the LIRH in the automated placement of CF/PEEK has not been reported in the literature. The temperature field in AFP needs further study based on this heating method.

This paper focuses on the temperature field analysis during the automated layup for carbon fiber-reinforced thermoplastic composites based on the heating method of the LIRH. The fundamental principles of LIRH are presented. The temperature field heat transfer model is developed, and boundary conditions are determined using CF/PEEK as the research object. A two-dimensional transient thermal model of the temperature field is established to analyze the heating and cooling effects of line-focused infrared heating at different tooling temperatures, the temperature gradient between the layers, and the peak temperature at different process parameters. The heating performance of the LIRH is evaluated by experimentally measuring the temperature field. The feasibility of the LIRH for automated placement processes in thermoplastic composites is analyzed.

Thermal model of thermoplastic fiber placement based on LIRH

Radiation theory of infrared lamps

LIRH is the primary heat source in AFP. LIRH means that the near-infrared light emitted from the halogen lamp is focused on the line through specially designed elliptical mirrors with special gold plating for extremely high reflective efficiency. The principle of the LIRH is shown in Figure 1. Compared with the traditional infrared heater, the focused infrared heater can achieve a better heating effect in a short time, which is suitable for the heating of thermoplastic composites in an automated layup. Due to the focused light, the heating area becomes smaller, and the exchange of heat is high. So, LIRH has a higher heat flux. Assuming that all objects are gray bodies, the intensity of radiation from an infrared lamp filament can be obtained according to Stephen Boltzmann’s law:

E = η σ_{b} T_{i}^{4}

(1)

Figure 1.

Schematic diagram of LIRH.

Where E is the intensity of radiation emitted by infrared lamps. $η$ is the radiation coefficient of the lamp. $σ_{b} = 5.67 \times 10^{- 8} W$ $m^{- 2}$ $k^{- 4}$ is Stefan Boltzmann constant

E = \frac{P_{e}}{A}

(2)

Where $A$ is the effective radiant area of the filament of the focused lamp, where the focusing filament structure is a double helix winding structure, so the filament satisfies the following equation:

A = 2 π γ ξ D L

(3)

Where

ξ

is the density coefficient of helical arrangement of the filament,

D

is the diameter of the filament,

L

is the length of the filament. Substitute (3) into the equation (2) to obtain:

E = \frac{P}{2 π γ ξ D L}

(4)

The radiative heat flow $q_{R A}$ between two surfaces was defined as³⁵:

q_{R A} = \frac{σ_{b} (T_{A_{2}}^{4} - T_{A_{2}}^{4})}{\frac{1 - ε_{1}}{ε_{1} \cdot A_{1}} + \frac{1}{A_{1} \cdot F_{12}} + \frac{1 - ε_{1}}{ε_{2} \cdot A_{2}}} = \frac{η \frac{q_{I R, \max}^{″}}{P_{e, \max}} P_{e} - σ_{b} T_{A_{2}}^{4}}{\frac{1 - ε_{1}}{ε_{1} \cdot A_{1}} + \frac{1}{A_{1} \cdot F_{12}} + \frac{1 - ε_{1}}{ε_{2} \cdot A_{2}}}

(5)

Where

ε_{1}

is the emissivity of the gold-plated surface,

ε_{2}

is the emissivity of the thermoplastic prepreg surface,

A_{1}

is the area of the semi-elliptic mirror,

A_{2}

is the area of the semi-elliptic mirror, and the view factor

F_{12}

is defined as the fraction of radiation that leaves

A_{1}

and intercepts

A_{2}

, the factor is defined as the efficiency with which the electrical power is converted into radiant power from the LIRH. The relevant parameters of LIRH are shown in Table 1. According to Kirchhoff’s law, for a given temperature and wavelength, the emissivity and absorptivity are equal.

ε_{1}

ε_{2}

A_{1}

A_{2}

are all constants. Only the view factor

F_{12}

changes with the position of the heater. Prepreg tows are opaque and highly absorbent. In the above equation, it is assumed that the heat flux q of IR the infrared heater increases linearly with the power

P_{e}

Table 1.

The parameters of LIRH from LONGPRO without documentation.

Parameter	Values/Models	Source
$ε_{1}$	0.95	⁴⁴
$ε_{2}$	0.90	¹⁸
$A_{1}$	7900 mm²	LONGPRO
$A_{2}$	252 mm²	LONGPRO
$q_{I R, \max}^{″}$	400 KW/m²	LONGPRO
$P_{e, \max}$	1400 W	LONGPRO
$η$	0.75
The width of focused light	7 mm	LONGPRO
Filament material	Double spiral tungsten wire	LONGPRO
The length of unfocused light	32.5 mm	LONGPRO
Rated voltage	220 V	LONGPRO
Focusing length	5-10 mm
Maximum response time	0.7s

Heat transfer process during layup

In the thermoplastic composite layup process, the resin in the prepreg tow is heated and melted by the LIRH, and pressure is applied by a pressure roller to make the resin flow, thus realizing the layup forming of prepreg tows. The LIRH-assisted automated layup process of thermoplastic composites is shown in Figure 2. Therefore, whether its temperature can reach the melting point is the key to ensuring the fluidity of the resin. Infrared radiation mainly acts on the upper surface of the paving layer, and the temperature change of each layer mainly relies on the heat transfer process between the layers. During the analysis, the following assumptions were made to simplify the model:

(1) The heat loss by radiation is ignored because the effect of thermal radiation is negligible compared to heat convection and heat conduction.

(2) The deformation in the direction of the thickness is neglected. Each pavement is considered a plate.

(3) It is assumed that the speed is constant throughout the layup so that the heating time is equal at any position of the layup.

(4) The heating width of the LIRH is greater than the width of the prepreg strand, so it is assumed that the prepreg is heated uniformly in the width direction.

(5) The heat flux describes the boundary conditions of the heated region. This discussion specifically refers to Section 4.2.1.

Figure 2.

Schematic diagram of the LIRH-assisted AFP.

Based on the above assumptions, a semi-infinite solid model is used to simulate the heat transfer in the layup process. This model can be described as an infinite extension of the surface in the depth direction, which only considers the heat transfer in the thickness and length directions. The two-dimensional surface values represent the values in the depth direction. The variation in the temperature in the focusing heating region of the materials and the peak temperature of each layer are of interest. The equation for two-dimensional heat conduction is³⁶:

k_{x} \frac{\partial^{2} T}{{\partial x}^{2}} + k_{y} \frac{\partial^{2} T}{{\partial y}^{2}} = ρ c \frac{\partial T}{\partial t}

(6)

Where $k_{x}$ and $k_{y}$ are the thermal conductivities in the respective directions; $ρ$ is the mass density. $c$ is the specific heat During the finite element analysis, the differential equations governing the thermal conductivity of the temperature field are transformed into equivalent finite element control equations for transient heat transfer:

[C] {\dot{T}} + [K] {T} = {Q}

(7)

Where,

C

is the matrices of specific heat,

{\dot{T}}

is derivative of nodal temperature with respect to time.

K

is the matrices of thermal conductivity,

T

is nodal temperature vector. Q is Nodal heat flow rate vector

Geometry model and boundary conditions

Temperature is one of the most important process parameters in the thermoplastic composite layup, which is used to predict the reduction of voids, the degree of intimate contact and healing of interfaces between the tapes during the consolidation process, and the crystallization of AS4/PEEK composites.^37–40 From the material perspective, temperature is an important factor affecting the crystallization of thermoplastic resins. Controlling the temperature change is conducive to obtaining a higher degree of crystallinity. The increase in temperature causes the thermoplastic resin to change from the glassy state to the rubbery state, and when the temperature continues to rise above the melting point, the thermoplastic resin changes from the rubbery state to the viscous state. According to the manufacturer’s recommendations, 380 °C–400 °C is the appropriate forming temperature range. From the perspective of the layup process, we need to control the heater’s power so that the resin can melt and flow at a certain laying speed and then apply enough pressure to bond the layers. Therefore, it is essential to establish an accurate geometric heat transfer model to analyze the temperature field during the automated layup process.

The variation of the temperature field during the automated placement process is time-dependent. The geometric model of the automated placement process of thermoplastic composites and the corresponding positional relationships and boundary conditions are shown in Figure 3. This coordinate system is located on the bottom surface of the tooling. It does not change with material deposition throughout the layup process. The prepreg tow is laid horizontally to the left, and the speed is set to $v$ . In the model, $S_{i} (i = 1, 2 . . ., 10)$ represents each side of the temperature field model. $y_{1}$ and $y_{2}$ represent the heights of the tooling and layup layer. $d$ represents the thickness of a single layer of prepreg tow. $l$ is the horizontal length of the entire model. $x_{1}$ represents the horizontal distance between the leftmost side of the prepreg tow that is laid out for forming and the origin, $x_{1}$ represents the distance between the right side of the heating area and the compaction point, and $w_{1}$ represents the heating width. During the layup process, the x-coordinate of the forming melting point does not coincide with the center of the nip roll because the pressure deforms the prepreg tow.

Figure 3.

Geometric model of LIRH-assisted automated thermoplastic fiber placement.

Based on the LIRH-assisted AFP, the boundary conditions of the temperature field are determined. Before the start of layup, assume that the surface temperature of the prepreg and the tooling are $T_{p}$ and $T$ , respectively. When $t = 0$ , The bottom surface S of the tooling and the left S2m1 and right S surfaces of the tooling can be defined: $T (x, y) = T_{m}$

The initial temperature of the surface $S_{4}$ , $S_{5}$ , $S_{6}$ , $S_{8}$ of the thermoplastic prepreg tow can bedefined as: $T (x, y) = T_{p}$

In the automated layup process, the boundary conditions for heat transfer according to Fourier’s law and Newton’s cooling equation are described as follows: the bottom surface $S_{1}$ of the tooling and the left $S_{2}$ and right $S_{3}$ surfaces of the tooling are exchanged with the air by natural convection, and the heat transfer process can be described as follows:

- k_{n} \frac{\partial T (x, y)}{\partial n} = h_{1} (T - T_{\infty})

(8)

Where n denotes the normal direction of the heat transfer surface. h1 is the surface heat transfer coefficient of the tooling under natural convection conditions. T is the ambient temperature. kn is the thermal conductivity of the prepreg wire bundle along the normal direction.

On $S_{1}$ , $x$ and $y$ satisfy, respectively: $0 \leq x \leq l, y = 0$ ; On $S_{2}$ , $x$ and $y$ satisfy, respectively: $x = 0$ , $0 \leq y \leq y_{1}$ ; On $S_{3}$ , $x$ and $y$ satisfy, respectively: $x = l$ , $0 \leq y \leq y_{1}$ ;

In the geometrical model, natural convection heat transfer with air takes place on the left $S_{4}$ and right $S_{5}$ surfaces of the pavement, on the upper $S_{8}$ surface of the pavement and on the upper surface $S_{6}$ of the prepreg strip that has been laid. The boundary conditions can be described as:

- k_{n} \frac{\partial T (x, y)}{\partial n} = h_{2} (T - T_{\infty})

(9)

Where h₂ is the surface heat transfer coefficient of the prepreg under natural convection conditions. $T_{\infty}$ is the ambient temperature.

On $S_{4}$ , $x$ and $y$ satisfy, respectively: $x = 0$ , $y_{1} \leq y \leq y_{1} + y_{2}$ ; On $S_{5}$ , $x$ and $y$ satisfy, respectively: $x = l$ , $y_{1} \leq y \leq y_{1} + y_{2}$ ; On $S_{6}$ , $x$ and $y$ satisfy, respectively: $x_{1} + w_{l} \leq x \leq l$ , $y = y_{1} + y_{2} + d$ ; $S_{8}$ , $x$ and $y$ satisfy, respectively: $0 \leq x \leq x_{1}$ , $y = y_{1} + y_{2}$ ;

In AFP, the rollers are made of steel. At the beginning, the initial temperature of the rollers is the same as that of the lower surface of the prepreg tows. The heat transfer coefficient of steel is much higher than that of the prepreg tows, so the rollers usually act as a cooling source and continuously absorb heat. The boundary conditions can be described as

- k_{n} \frac{\partial T (x, y)}{\partial n} = h_{r} (T - T_{r})

(10)

where

h_{r}

is the convective heat transfer coefficient of the pressure roller.

T_{r}

is the surface temperature of the roll at the initial moment.

On $S_{10}$ , $x$ and $y$ satisfy, respectively: $x = x_{1} + w_{l} + x_{2}$ , $y = y_{1} + y_{2}$ ;

At the contact surface $S_{9}$ between the lower surface of the pavement and the upper surface of the tooling, Contact heat transfer between layup and tooling. This boundary condition is described as:

- k_{n} \frac{\partial T (x, y)}{\partial n} = h_{m} (T - T_{m})

(11)

Where

h_{m}

is the thermal conductivity of the tooling.

On $S_{9}$ , $x$ and $y$ satisfy, respectively: $0 \leq x \leq l$ , $y = y_{1}$ ;

In the geometric model, $w_{l}$ is the width of the infrared focusing heating area, and the upper surface $S_{7}$ of the layup is the main heated area. Due to the complexity of the infrared radiation heating mechanism, the load is applied in the form of heat flow density in order to simplify the model. Thus, this boundary condition is described as:

- k_{n} \frac{\partial T (x, y)}{\partial n} = \frac{q}{k_{n}}

(12)

Where q is the heat flux converted to infrared radiation. On

S_{7}

x

and

y

satisfy, respectively:

x_{1} \leq x \leq x_{1} + w_{1}

y = y_{1} + y_{2} + d

Thermal convection under natural conditions

Convection loads should be applied to all surfaces exposed to the ambient environment. The convective heat transfer coefficient under natural conditions increases as the surface temperature increases, and the following empirical equations apply to that condition³⁵:

h = 1.32 {(\frac{∆ T}{L})}^{1 / 4}

(13)

The

L = 0.07 m

is the characteristic length of the natural convection region, and the heat transfer coefficients of thermoplastic composites under ambient environment at different temperatures are calculated as listed in Table 2.

Table 2.

Convective heat transfer coefficients under ambient environment.

T (°C)	h (W/m²·°C)
20	5.42
50	6.82
100	8.11
150	8.98
200	9.65
250	10.20
300	10.68
350	11.10
400	11.48
450	11.82
500	12.14

Finite element simulation based on ANSYS

Finite element heat transfer modeling and solution

In the automated layup process, the infrared heater predominantly focuses its emitted light on the upper surface of the prepreg tow being laid down. During this process, the prepreg transitions into a molten state, enabling it to bond together under the pressure of the roller. Throughout the layup, both the heater and pressure roller continuously move alongside the layup head, while the prepreg tow is deposited onto the substrate. Consequently, the temperatures of the prepreg bundles and the layups vary continuously with both time and position, aligning well with transient heat transfer analysis. The temperature of the prepreg bundles directly influences the temperature of the substrate, and given that PEEK has a melting temperature of 343°C, it is particularly sensitive to temperature variations. To understand the impact of temperature on the substrate and prepreg bundles during the infrared-assisted heating layup, it is essential to analyze the temperature changes occurring in each layer throughout the process. This study employed the ANSYS Workbench platform in conjunction with a parametric language to address the thermal field during thermoplastic automated placement. Additionally, cyclic loading and birth-death element strategies were utilized to simulate the continuous movement of the laying head and the material deposition process. The procedural steps for analyzing the transient temperature field during the automated placement of thermoplastic composites are illustrated in Figure 4.

Figure 4.

Flow chart for parametric language design.

In the preprocessing module, the element type for the model was defined, selecting Element PLANE293 in ANSYS Workbench for 2D transient thermal analysis. This planar element, characterized by four nodes, is capable of discretizing orthogonal anisotropic materials that exhibit temperature-dependent specific heat capacities and thermal conductivity coefficients. The relevant material properties for CF/PEEK and the tooling were detailed in Table 3 and Table 4, respectively. The dimensions of a single layer of prepreg were specified as 70 mm in length and 0.2 mm in thickness. The AFP process was simulated from the first ply through to the sixth ply. The boundary conditions and associated input parameters for the transient temperature field model of the LIRH-assisted automated placement process are listed in Table 5. Following the principles of infrared radiation, a linear relationship was established between the power output of the LIRH and the corresponding heat flux. The actual laying process was simulated by adjusting the applied heat flux, while the laying speed was set within a specific range based on the processing temperature of the prepreg tows. It is noteworthy that the thermal loads implemented in this study were based on the line-focused infrared layup experiments that were conducted. Due to the short focusing length, the thermal loads were applied not at the nip point but rather at the upper surface of the prepreg. The contact lengths between the roller and prepreg were minimal. Consequently, the cooling effect of the pressure roller on the prepreg was neglected in the finite element model. Furthermore, given that the tooling temperature significantly influences the cooling rate throughout the laying process, the initial temperature of the tooling was treated as a variable in order to investigate its effect on the temperature field of the prepreg tows during deposition.

Table 3.

Tooling physical parameters.

Parameter	Value
Specific heat C (J/(kg·K))	905
Density ρ (kg/m³)	2700
Thermal conductivity K (W/(m·K))	237

Table 4.

Temperature-dependent material thermal properties (CF/PEEK)²⁶.

Temperature (°C)	Density(kg/m³)	Specific heat(J/(kg·K))	Conductivity (W/(m·K))
Temperature (°C)	Density(kg/m³)	Specific heat(J/(kg·K))	Axial	Transverse
0	1601	800	3.5	0.42
50	1598	930	4.6	0.52
100	1593	1040	5.1	0.6
150	1586	1260	5.9	0.7
200	1575	1300	5.9	0.7
250	1563	1400	6.1	0.7
300	1551	1550	6.7	0.75
350	1537	1650	6.8	0.68
400	1524	1700	7.0	0.65

Table 5.

Input parameters and boundary conditions.

Parameter	Unit	Value
Width of prepreg	mm	6.35
Single-ply thickness of the prepreg	mm	0.2
Length of the mandrel l	mm	70
Total number of plies n		6
Thickness of the mandrel y₁	mm	20
Thermal conductivity of the tooling h_m	W/(m²·K)	2000
Thermal conductivity between layers	W/(m²·K)	5500
Convective heat transfer coefficients of composite under free convection h₁ h₂	W/(m²·K)	Table 2
Initial temperature of prepreg T_p	°C	20
Ambient temperature T_∞	°C	20
Initial temperature of the tooling T_m	°C	50,100,140
The power of the LIRH P	W	0-1400
Moving speed of heater v	mm/s	10-80
Heating width w_l	mm	72

In the automated layup process, the heat source continuously moved in the layup direction. The moving heat source needed to be discretized into a specified time step in the finite element model. Assuming that the length of the pavement is 70 mm, the LIRH moved at 30 mm/s, and the total time for the pavement of a single layer was 2.1 s. The heating width was 72 mm and the width of focused light was 7 mm. The length of the single layer prepreg tows was divided into 100 elements, and the size of single element was 0.7 mm. The red and circle shown in Figure 5 was the heating region. In the finite element model, the dead element that was about to be heated was activated, and the heat source moved one node for each additional time step. The time taken by the moving heat source to move two neighboring nodes was the distance between the nodes divided by the speed of movement (Δt = 0.7 mm/30 mm/s = 0.023s). This meant that the current residence time in the heated area was 0.023 s. The results of the calculations in step 1 were applied as initial conditions in step 2, and each node advanced by the moving heat source was considered to cycle through the calculations until the end of the cycle of n calculations to the single-layer placement (step 4).

Figure 5.

Schematic diagram of the process of heat transfer analysis of a moving heat source in Workbench using ANSYS Parametric Design Language.

Simulation results of the temperature field

Variation of the temperature field in the layers

Figure 6 illustrates the 2D temperature distribution contours for selected locations in the third and sixth layers, under conditions of a tooling temperature of 50°C, a laying speed of 30 mm/s, and an infrared power of 1400 W. To enhance clarity in distinguishing between active and inactive elements, the elements that had not yet been activated are omitted from the figure.

Figure 6.

Temperature distribution contours in the third and sixth layers at 30 mm/s and an infrared power of 1400W.

The temperature distribution history throughout the AFP process, encompassing both heating and cooling phases, had been elucidated through numerical simulations. Figure 7 displayed the temperature histories recorded at six distinct points during the entire laying process. Six temperature probes designated TP1 to TP6, were positioned on the upper surfaces of layers 1 to 6, with each probe located 35 mm from the left end of the model. The thermal conductivity between the tooling and the first layer was set at 2000 W/(m²·K), while the thermal conductivity between subsequent layers was defined as 5500 W/(m²·K).¹⁸ A rapid temperature rise was observed as the focused light reached the area directly above each temperature probe, followed by a swift decrease in temperature as the moving heat source passed beyond the probes. This effect diminished with increasing distance from the heated surface, indicating that the surface layer experienced heating and cooling at a significantly faster rate than the layers below it.

Figure 7.

Temperature histories of points at TP1-TP6 on the top surfaces of the 1-st to 6-th tows.

A gradual flattening of the temperature wave peaks was observed as the distance from the heated surface to the substrate increased. By the time the heater passed over the fourth layer, the temperature of the first layer had already dropped below the glass transition temperature, indicating that the maximum number of layers that could be cooled to below the glass transition temperature was three. Although there was sufficient time for the first two layers to cool completely, heat accumulation occurred. This heat accumulation resulted in four layers remaining above the glass transition temperature when the moving heat source passed over the sixth layer. This observation suggests that the number of layers exceeding the glass transition temperature increased with the deposition of additional layers due to the effects of heat accumulation.

Effect of tooling temperature on cooling rate

The literature has reported that the cooling rate significantly influences the crystallinity of thermoplastic composites.⁴¹ Accurate calculation of the cooling rate during lay-up is crucial for studying crystallization kinetics under non-isothermal conditions.³⁷ This section investigates the effect of tooling temperature on the cooling rate during the layup process. Figure 8 illustrates the heating and cooling profiles of the peak temperature for the first layer at tooling temperatures of 50°C, 100°C, and 140°C. The heating and cooling intervals were determined by tracking the movement of the nodes. Initially, the unfocused light provided weak heating to the material. Subsequently, the temperature rose rapidly when the focused light reached temperature probe TP1. Following this, the influence of unfocused light introduced an inflection point in the cooling period, indicating that unfocused light can effectively slow down the cooling process.

Figure 8.

Effect of the boundary condition of the tooling on temperature histories of the first layer when the speed is 30 mm/s, and the power is 1400W.

The interval between the glass transition temperature and the peak temperature was selected to calculate the cooling rate at temperature probe TP1. As the tooling temperature increased, it was observed that the cooling rate from the peak temperature to the glass transition temperature decreased. Specifically, when the tooling temperature was set at 50°C, the calculated cooling rate was 289°C/s, whereas this rate slowed to 168°C/s when the tooling temperature was raised to 140°C Furthermore when the tooling temperature exceeded 143°C, isothermal crystallization of the substrate occurred. This isothermal crystallization, sustained for a certain duration, enhanced the degree of crystallinity. This analysis highlighted the critical role that tooling temperature plays in modulating the cooling rate, which in turn affects the thermal properties and crystallinity of the thermoplastic composite layers.

Effect of laying speed and focused lamp power

LIRH power and laying speed are the two primary parameters of the automated laying process that directly influence the temperature field. To achieve optimal laying quality, it is essential to maintain the material temperature at a constant value during the layup process. The relationship between material temperature, laying speed, and LIRH power was established through finite element simulations. This section conducted a series of comprehensive simulation tests to investigate these interrelationships. Specifically, a tooling temperature of 50°C was used, while laying speeds of 10, 20, 25, 30, 35, 40, 45, 50, 60, and 80 mm/s were combined with LIRH powers of 140, 280, 420, 560, 700 ,840, 980, 1120, 1260, and 1400 W. A total of 100 simulation tests were performed to explore these dynamics, with the resulting relationships depicted in Figure 9.

Figure 9.

The correlation between material temperature, laying speed, and focused lamp power when the tooling temperature is 50°C.

As the laying speed increased, the influence of LIRH power gradually diminished. At a laying speed of 10 mm/s, the temperature difference resulting from varying LIRH power was 400°C. In contrast, at a higher laying speed of 80 mm/s, this temperature difference decreased to only 150°C. At lower laying speeds, the prepreg had sufficient time to absorb heat, allowing for greater temperature variation with changes in LIRH power. From the finite element simulations, we concluded that when the focused lamp was operated at its maximum power of 1400 W, a laying speed of 40 mm/s was barely adequate to achieve the melting temperature of PEEK, but it did not reach the higher forming temperature necessary for optimal processing. Conversely, a laying speed of 30 mm/s was found to attain a more suitable forming temperature for the material.

Experimental verification of the temperature field

Experimental setup

An online temperature field measurement platform was developed to assess the changes in the temperature field during the placement of thermoplastic composite fibers with the assistance of LIRH. The temperature measurement platform consisted of a temperature sensor and an intelligent paperless recorder. The paperless recorder efficiently processes signals and stores relevant data for analysis. The temperature sensor utilized was a sheet-type K thermocouple, with the front end secured to the pavement using high-temperature tape, while the other end was connected to the paperless recorder. The experimental setup associated with the temperature field measurements is depicted in Figure 10. The Type K thermocouples (Model PB-HH-K-24) featured leads made of fiberglass and a metal shielding net. The diameter of the wire was 0.5 mm, while the outer diameter was 2.1*2.9 mm. The wire material was a nickel-chromium alloy, and the thickness of the sheet thermocouple was 0.15 mm. The temperature range of the paperless recorder (Model KPS04A0R) was −200 to 2400°C, with an acquisition time set at 0.1 seconds.

Figure 10.

Physical diagram of the online temperature field measurement system.

In this study, thermocouples were employed to measure the temperature field at various locations on the same pavement and on the upper surface of the first pavement layer during the layup of six layers. The investigation focused on the peak temperature under different laying speeds and infrared power settings, reflecting the unique characteristics of the LIRH system. Additionally, the effects of focused and unfocused light on the temperature field were explored. The material used in the experiments was CF/PEEK prepreg produced by Heng Bo Composites, which had a resin mass fraction of 34% ± 3%. Detailed parameters associated with the LIRH system are described in Section 2.

Results of the experiment

Thermal effects of unfocused light

The images captured by the focusing effect camera for the LIRH at focusing lengths of 5, 10, 15, and 20 mm are shown Figure 11. At a focusing length of 5 mm, the focused area was smaller, resulting in a stronger focusing effect. With a focusing length of 10 mm, the linear focusing effect became more pronounced; however, the heat flux at the focal point was slightly diminished. At a focusing length of 15 mm, the focusing effect began to deteriorate, leading to an increase in the extent of unfocused light. Finally, at a focusing length of 20 mm, the focusing effect was nearly imperceptible.

Figure 11.

Camera picture of focusing effect at focusing lengths of 5 mm(a),10 mm(b),15 mm(c),20 mm(d).

Figure 11 illustrates that the heating area of the LIRH was not limited to focused light; unfocused light also had the potential to affect the temperature field. To analyze the impact of unfocused light on the temperature field, thirteen sets of Type K thermocouples were arranged on the pavement with a spacing of 6 mm. The configuration of the temperature measurement devices is depicted in Figure 12. To ensure that the increased temperatures recorded by thermocouples TC1 through TC13 were due to unfocused light and not attributable to heat conduction from the prepreg tow in the focal area, each experimental trial measured the temperature at a single location.

Figure 12.

Layout of the focused infrared lamp.

Figure 13 presents the experimental results of the LIRH at a focusing length of 10 mm. It is evident from the figure that the experimental results align with the characteristics of the LIRH system. The temperature in the focused area reached 416°C, while the temperatures in the unfocused area ranged from 50°C to 110°C. This indicates that the primary heating effect originated from the focused light, with a weaker heating effect resulting from the unfocused light. As the unfocused light moved away from the center of the focus, its heating effect diminished. These findings can serve as a reference for determining the numerical values of heat flux in simulations.

Figure 13.

Temperature distribution on prepreg tow at the focusing length of 10 mm.

Rapid response is essential for temperature control during variations in speed.⁴² When the power of the IR heater is increased from 0 W to 1400 W, the thermocouple records a temperature of 416°C in 0.9 seconds. Once the IR heater stabilizes its output at 1400 W, a temperature of 400°C is reached in 0.2 seconds. A rough estimate of the maximum response time of the line-focused IR heater is approximately 0.7 seconds. This response time accommodates the low-speed variations encountered during layup.

Effect of laying speed and lirh power on the temperature field

Considering laying speeds ranging from 20 mm/s to 80 mm/s, with the maximum LIRH power set at 1400 W and tooling temperature maintained at 50°C, the relationship between peak temperature of the second layer and laying speed was analyzed. As illustrated in Figure 14(a), it can be concluded that the peak temperature of the pavement decreases as the laying speed increases. In the simulation, an increase in laying speed from 20 mm/s to 30 mm/s resulted in a significant reduction in peak temperature of approximately 90°C. Conversely, when the laying speed increased from 70 mm/s to 80 mm/s, the peak temperature of the pavement only decreased by about 20°C. However, the experimental data did not exhibit the same clarity of trend as the simulation results. Notably, the experimental measurements were consistently lower than the simulated values, particularly at higher laying speeds. This discrepancy may be attributed to several factors. The thermocouples may exhibit hysteresis, and data were sampled at relatively large intervals. At higher laydown speeds, the duration of exposure to the focused light at the thermocouple collection points was reduced, leading to a lag in the thermocouples’ response to temperature changes. Additionally, the paperless recorder encountered difficulties recording the instantaneous fluctuations in thermocouple temperature. These factors likely contributed to the observed lower experimental values compared to the simulation predictions at elevated laying speeds.

Figure 14.

The influence of infrared power and speed on the peak temperature of the second layer. (a)Comparison of the peak temperature under different speeds. (b) Comparison of the peak temperature under different LIRH power.

In the case of LIRH power ranging from 420 W to 1400 W, with a consistent speed of 30 mm/s and a tooling temperature maintained at 50°C, the relationship between peak temperature of the second layer and LIRH power was illustrated Figure 14(b). At higher power levels, the experimental measurements aligned well with the simulation results. However, discrepancies were observed at lower LIRH power levels. A significant factor contributing to this mismatch was the variable resistance of the tungsten filaments in line-focused infrared heaters, which changed with increasing temperature. This variation led to a nonlinear relationship between heating power and temperature peaks, particularly affecting the accuracy of predictions at lower power settings. Understanding this nonlinearity was crucial for improving the correlation between experimental data and simulation outcomes across the range of LIRH power.

Heat transfer between layers

Setting the LIRH power to 1400 W, with a speed of 30 mm/s and a tooling temperature of 50°C, we compared the simulation results of the temperature field of the first layer throughout the laying process with the experimental measurements. Figure 15(a) illustrates the temperature history of the bottom surface as the heater passes over the first layer. As observed in the figure, the heating phase of the experimental results aligns well with the simulation data. Notably, the cooling rate significantly slowed down as the experimental temperature decreased from the peak to 100°C, which corresponds with the trend observed in the simulation results. However, discrepancies were noted between simulation and experimental data. Several factors may have contributed to these differences. First, the focusing distance of the LIRH influenced the heat flux distribution; even minor errors could result in inaccurate heat flux assessments. Second, the thermocouple sampling interval of 0.1 seconds posed challenges in accurately capturing the temperature values during the rapid heating and cooling phases. Consequently, the thermocouple may not have adequately tracked the instantaneous changes in temperature, thereby affecting the measurement of heating and cooling rates. Third, while the simulation incorporated interlayer thermal resistance based on existing literature, the actual contact thermal resistance is influenced by the degree of surface contact, surface roughness, and transverse thermal conductivity.⁴³ An inaccurate setting of contact thermal resistance can also result in errors.

Figure 15.

(a) The heating and cooling history in the first layer (b) The peak temperature of the first layer from the first to sixth layer under different interlaminar contact conditions. (c) Temperature history of the 1st layer during automated fiber placement (d) Temperature history of the 2nd layer during automated fiber placement.

In the present study, thermal conductivity coefficients (TCC) of 3500, 5500, 7500, 10000 W/(m²·K) and perfect contact are considered. Figure 15(b) displays the peak temperature of the first layer as the heater moves through the first to sixth layers under varying interlaminar contact conditions. A high contact heat transfer coefficient results in a higher temperature in the first layer. It is evident that contact thermal resistance hinders heat transfer in the thickness direction, resulting in a reduction in peak temperature. Contact thermal resistance is a critical factor influencing interlayer heat transfer and is significantly affected by the degree of intimate contact.³⁸ In future work, intimate contact will be considered in automated placement, with a focus on simultaneously solving for heat transfer and the evolution of intimate contact.

Figure 15(c) and (d) illustrate the temperature profiles of the first and second layers of prepreg tows throughout the layup process. As shown in the figures, the simulated and experimental temperature histories exhibit similar trends. However, the post-cooling temperature values in the experiments gradually increase with the number of layup layers. This phenomenon can be attributed to the continuous heating of the surrounding air by infrared radiation and the tool, which reduces the temperature difference between the tows and the ambient environment. Additionally, the exothermic crystallization of the matrix also affects the cooling rate and the ambient temperature of the surrounding environment. The current model does not account for these factors, leading to certain discrepancies in the results.

Conclusion

In this study, the variations in the temperature field during fiber placement were investigated using the LIRH with CF/PEEK as the primary material of interest. The influence of various process parameters on the transient temperature field was thoroughly analyzed. A two-dimensional heat transfer model of the automated layup process was developed, with its boundary conditions established based on heat transfer theory. A continuous dynamic finite element analysis of the layup model was conducted using ANSYS to examine the temperature distribution under the specified heating parameters of the LIRH and to derive a range of temperature values corresponding to different process settings. Through both experimental and finite element method simulations, the following conclusions can be drawn:

(1) Material temperatures of CF/PEEK can reach up to 400°C, achieved by utilizing an infrared heater with a laying speed of 30 mm/s, an infrared power of 1400W, and a tooling temperature of 50°C.

(2) The finite element analysis revealed that an increase in tooling temperature leads to a reduction in the cooling rate, while unfocused light can also contribute to a decrease in the cooling rate.

(3) Experimental results demonstrate that focused light significantly elevates the surface temperature of the material, whereas unfocused light exhibits a minimal heating effect.

This work highlights the feasibility of using LIRH as a primary heating method for applications in AFP from the perspective of the temperature field. Future research will aim to provide a comprehensive and integrated characterization of the properties of the resulting products based on this heating method.

Footnotes

Author contribution

All authors contributed to the study conception and design, all authors commented on previous versions of the manuscript, and all authors read and approved the final manuscript.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Natural Science Foundation of China [grant number 52075361]; Shanxi Province Science and Technology Major Project [grant number 20201102003]; Shanxi-Zheda Institute of Advanced Materials and Chemical Engineering [grant number 2022SX-TD021]; Lvliang Science and Technology Guidance Special Key R\&D Project [grant number 2022XDHZ08].

ORCID iD

Jialin Li

Data Availability Statement

Date will be made available on request.*

References

Yassin

Hojjati

. Processing of thermoplastic matrix composites through automated fiber placement and tape laying methods: a review. J Thermoplast Compos Mater 2018; 31: 1676–1725.

Zhang

Wang

Pei

, et al. Review of automated fibre placement and its prospects for advanced composites. J Mater Sci 2020; 55: 7121–7155.

Oromiehie

Prusty

Compston

, et al. Automated fibre placement based composite structures: review on the defects, impacts and inspections techniques. Compos Struct 2019; 224: 110987.

Zhao

Liu

Chen

, et al. The effect of temperature field on the characteristics of carbon fiber reinforced thermoplastic composites in the laying and shaping process. Int J Adv Manuf Technol 2022; 121: 7569–7589.

Zhang

Duan

Xiao

, et al. The effects of processing parameters on the wedge peel strength of CF/PEEK laminates manufactured using a laser tape placement process. Int J Adv Manuf Technol 2022; 120: 7251–7262.

Zhao

Liu

Yue

, et al. Determination of thermal degradation for high-performance thermoplastic composites manufactured by laser-assisted automated fibre placement. High Perform Polym 2022; 34: 149–161.

Soutis

. Carbon fiber reinforced plastics in aircraft construction. Materials Science and Engineering: A 2005; 412: 171–176.

Schmidt

Damerow

Lauter

, et al. Manufacturing processes for combined forming of multi-material structures consisting of sheet metal and local CFRP reinforcements. KEM 2012; 504–506: 295–300.

Jin

Sun

Liu

, et al. Forming method, crystallinity, and mechanical performance of buckypaper/carbon fiber/polyether ether ketone multi-scale thermoplastic composites. High Perform Polym 2024; 36: 119–129.

10.

Hosseini

SMA

Schäkel

Baran

, et al. Non-uniform crystallinity and temperature distribution during adjacent laser-assisted tape winding process of carbon/PA12 pipes. Int J Adv Manuf Technol 2020; 111: 3063–3082.

11.

Van Rijswijk

Bersee

HEN

. Reactive processing of textile fiber-reinforced thermoplastic composites – an overview. Compos Appl Sci Manuf 2007; 38: 666–681.

12.

Cao

Dong

Liu

, et al. Temperature field in the heat transfer process of PEEK thermoplastic composite fiber placement. Materials 2020; 13: 4417.

13.

Zhao

Hoa

Gao

. Thermal stresses in rings of thermoplastic composites made by automated fiber placement process. Sci Eng Compos Mater 2011; 18: 35–49.

14.

Yang

. Dynamic finite element simulation and transient temperature field analysis in thermoplastic composite tape lay-up process. J Thermoplast Compos Mater 2015; 28: 558–573.

15.

Zacherl

Shadmehri

Rother

. Determination of convective heat transfer coefficient for hot gas torch (HGT)-assisted automated fiber placement (AFP) for thermoplastic composites. J Thermoplast Compos Mater 2023; 36: 73–95.

16.

Pan

Yang

, et al. As-built FE thermal analysis for complex curved structures in automated fiber placement. Simulat Model Pract Theor 2022; 118: 102561.

17.

Lichtinger

Hörmann

Stelzl

, et al. The effects of heat input on adjacent paths during Automated Fibre Placement. Compos Appl Sci Manuf 2015; 68: 387–397.

18.

Wang

, et al. Thermal management for thermoset automated fiber placement based on infrared heater structure arrangement. Chin J Aeronaut 2022; 35: 173–183.

19.

Hörmann

Stelzl

Lichtinger

, et al. On the numerical prediction of radiative heat transfer for thermoset automated fiber placement. Compos Appl Sci Manuf 2014; 67: 282–288.

20.

Stokes-Griffin

Kollmannsberger

Ehard

, et al. Manufacture of steel–CF/PA6 hybrids in a laser tape placement process: effect of first-ply placement rate on thermal history and lap shear strength. Compos Appl Sci Manuf 2018; 111: 42–53.

21.

Stokes-Griffin

Compston

. Investigation of sub-melt temperature bonding of carbon-fibre/PEEK in an automated laser tape placement process. Compos Appl Sci Manuf 2016; 84: 17–25.

22.

Stokes-Griffin

Compston

. A combined optical-thermal model for near-infrared laser heating of thermoplastic composites in an automated tape placement process. Compos Appl Sci Manuf 2015; 75: 104–115.

23.

Di Francesco

Veldenz

Dell’Anno

, et al. Heater power control for multi-material, variable speed Automated Fibre Placement. Compos Appl Sci Manuf 2017; 101: 408–421.

24.

Baho

Ausias

Grohens

, et al. Simulation of laser heating distribution for a thermoplastic composite: effects of AFP head parameters. Int J Adv Manuf Technol 2020; 110: 2105–2117.

25.

Weiler

Emonts

Wollenburg

, et al. Transient thermal analysis of laser-assisted thermoplastic tape placement at high process speeds by use of analytical solutions. J Thermoplast Compos Mater 2018; 31: 311–338.

26.

Stokes-Griffin

Compston

Matuszyk

, et al. Thermal modelling of the laser-assisted thermoplastic tape placement process. J Thermoplast Compos Mater 2015; 28: 1445–1462.

27.

Yousefpour

Hojjati

Immarigeon

J-P

. Fusion bonding/welding of thermoplastic composites. J Thermoplast Compos Mater 2004; 17: 303–341.

28.

James

Black

. Thermal analysis of continuous filament-wound composites. J Thermoplast Compos Mater 1996; 9: 54–75.

29.

Kim

Lee

. A study on heat transfer during thermoplastic composite tape lay-up process. Exp Therm Fluid Sci 1996; 13: 408–418.

30.

Suemasu

Aoki

Hoshi

, et al. Thermal analysis for in-situ consolidation in the AFP process. Compos Appl Sci Manuf 2023; 172: 107577.

31.

Kollmannsberger

Lichtinger

Hohenester

, et al. Numerical analysis of the temperature profile during the laser-assisted automated fiber placement of CFRP tapes with thermoplastic matrix. J Thermoplast Compos Mater 2018; 31: 1563–1586.

32.

Danezis

Williams

Edwards

, et al. Heat transfer modelling of flashlamp heating for automated tape placement of thermoplastic composites. Compos Appl Sci Manuf 2021; 145: 106381.

33.

Venkatesan

Velu

Vaheed

, et al. Effect of process parameters on polyamide-6 carbon fibre prepreg laminated by IR-assisted automated fibre placement. Int J Adv Manuf Technol 2020; 108: 1275–1284.

34.

Venkatesan

Zulkifli

Silva

. Effects of processing parameters of infrared-based automated fiber placement on mechanical performance of carbon fiber-reinforced thermoplastic composite. Compos Struct 2023; 309: 116725.

35.

Incropera

DeWitt

Bergman

, et al. (eds). Principles of heat and mass transfer. 7 International student version. Hoboken, NJ: Wiley, 2013.

36.

Sonmez

Hahn

. Modeling of heat transfer and crystallization in thermoplastic composite tape placement process. J Thermoplast Compos Mater 1997; 10: 198–240.

37.

Tierney

Gillespie

. Crystallization kinetics behavior of PEEK based composites exposed to high heating and cooling rates. Composites Part A Applied Science & Manufacturing 2004; 35: 547–558.

38.

Levy

Heider

Tierney

, et al. Inter-layer thermal contact resistance evolution with the degree of intimate contact in the processing of thermoplastic composite laminates. J Compos Mater 2014; 48: 491–503.

39.

Tierney

Gillespie

Jr . Modeling of in situ strength development for the thermoplastic composite tow placement process. J Compos Mater 2006; 40: 1487–1506.

40.

Tierney

Gillespie

. Modeling of heat transfer and void dynamics for the thermoplastic composite tow-placement process. J Compos Mater 2003; 37: 1745–1768.

41.

Zhao

Zhang

, et al. Effect of the cooling rate on the fracture toughness of carbon fiber reinforced thermoplastic composites. J Thermoplast Compos Mater 2023; 36: 1680–1693.

42.

Heider

Piovoso

Tierney

. Artificial neural network modeling of the automated thermoplastic composite tow-placement system. J Mater Process Manuf Sci 1999; 7: 360–379.

43.

Barasinski

Leygue

Soccard

, et al. Identification of non uniform thermal contact resistance in automated tape placement process. Int J Material Form 2014; 7: 479–486.

44.

Hohlfeld

Müller

Wellershoff

S-S

, et al. Time-resolved thermoreflectivity of thin gold films and its dependence on film thickness. Appl Phys B Laser Opt 1997; 64: 387–390.