Intrinsic production of metal-carbon fiber reinforced plastic hybrid shafts using vacuum-assisted resin transfer molding

Abstract

Over the past decades, the importance of lightweight structures in the aircraft and automotive industries has steadily increased due to regulations aimed at reducing global warming. Work hardened steel alloys are commonly used for lightweight applications, but they face stability issues when the material thickness reaches certain thresholds. Fiber Reinforced Plastics (FRP) offer a viable alternative due to their high strength-to-weight ratio, but they are often expensive due to long production cycles and high material costs. A feasible solution lies in hybrid lightweight designs that utilize expensive FRP materials only in highly stressed areas, achieving a balance between low mass and acceptable cost. These hybrid structures are lighter than metal components and more cost-effective compared to fully FRP structures, without compromising mechanical properties. This study focuses on producing rotationally symmetrical hybrid structures using Resin Transfer Molding (RTM) combined with vacuum assistance in a single-stage process. The research examines the effects of injection pressure, mold temperature, and the interface between metal and FRP. The mechanical characterization of these hybrid structures was conducted to assess their performance under torsion, compression, and interlaminar shear strength (ILSS) loading conditions. The results indicate that hybrid designs can offer a lightweight alternative without compromising mechanical properties.

Keywords

resin transfer molding metal-carbon fiber reinforced plastic (CFRP) hybrid structures mechanical characterizing computed tomography scan (CT-scan)

Introduction

Hybrid manufacturing techniques have gained importance because of their ability to produce lightweight structures with optimized properties. The main concept of hybrid manufacturing is using multiple materials to create structures that compensate for each other’s weakness without compromising their original properties, resulting in superior components that cannot be made from a single material.^1–3 Metal-FRP hybrid structures have been widely studied over the past few decades due to their desirable characteristics, including a high strength-to-weight ratio, improved fatigue performance, corrosion resistance, and vibration damping.^2,4,5 FRP-metal hybrid structures were first introduced for aerospace fuselage structures to replace the aluminum alloys for further weight reduction.⁶ Metal-FRP hybrids are now used in various car body parts due to their lightweight properties, with a potential application in the automotive sector for crashworthiness systems.⁷ Normally, FRP alone cannot be used for crashworthiness because its failure mode is characterized as brittle when loaded in the fiber direction due to unstable interlaminar layers under dynamic conditions. However, the failure behavior of the complete structure can be modified to a ductile failure by integrating metals, leveraging the plastic deformation phase of metals.^8–10 Given these benefits, Metal-FRP hybrids have been widely adopted in high-performance and luxury vehicles for applications requiring high impact resistance, energy absorption, and enhanced mechanical performance.^11,12 These hybrid structures not only meet the stringent safety requirements but also contribute to weight reduction, which is crucial for improving fuel efficiency and reducing emissions in modern vehicles.^13,14 Consequently, extensive research and development have been directed toward optimizing Metal-FRP hybrids for use in automotive systems, making them a promising solution for future vehicle designs including electrical configurations.^15–20

Resin Transfer Molding (RTM) is a well-established technique for manufacturing complex FRP structures and can also be used to produce hybrid structures.²¹ The RTM process begins with preparing a mold, which is coated with a release agent to prevent sticking and ensure easy removal of the final product. A fiber preform, typically made from materials such as glass, carbon, or aramid fibers, is then placed within the mold cavity. Thermosetting resin is injected into the mold under controlled pressure, ensuring thorough impregnation of the fiber preform. This step is crucial to avoid defects such as voids or dry spots. The mold is then maintained at elevated temperature to cure the resin, solidifying it and binding the fibers into a rigid composite structure. After curing, the mold is opened, and the finished composite part is demolded, often requiring further finishing processes like trimming or machining to meet specific requirements. Figure 1 shows the block diagram for the RTM process that produce Metal-FRP hybrid structures. RTM is favored for its ability to produce complex shapes with high fiber volume fractions and excellent surface finishes, though it demands careful management of resin infiltration and curing cycles to ensure high-quality outcomes.²²

Figure 1.

Block diagram showing the process cycle of Resin Transfer Molding (RTM).

However, there are challenges in the RTM process, especially for parts with high fiber content (typically greater than 40%), which decreases the permeability of the preform, leading to long filling times, incomplete impregnation, and high void content.^23–25 Several strategies have been developed to modify the conventional RTM process.^26,27 High Pressure Resin Transfer Molding (HP-RTM) was introduced to inject resin at pressures as high as 500 bars, reducing filling time but causing voids due to insufficient time for trapped air to escape.²⁸ Vacuum Resin Transfer Molding (VA-RTM) addresses this by evacuating the mold before and during the injection phase. Another improvement is Compression Resin Transfer Molding (CRTM), where the mold is partially opened until the end of resin injection and then closed, squeezing the resin into the preform. Therefore, instead of going through the entire fiber stacking length as in RTM, the CRTM process wets the fibrous reinforcement by penetrating in the thickness direction.^29,30 However, CRTM came with a new challenge, as the wetting of fiber is not homogeneous along the thickness.³⁰ Light Resin Transfer Molding (LRTM) was developed to reduce filling time for large molds by first circulating resin around the mold’s circumference and then impregnating the fibers in a convergent phase. Despite these advancements, the selection of appropriate process parameters remains crucial.³¹ Parameters such as injection pressure, mold temperature, location of inlets and vents, fiber volume, fiber layup, vacuum application, and injection mass flow rate all influence the mechanical properties of the final composite.^32–34 These parameters significantly impact the flow front path of the matrix, which is the most critical aspect of the RTM process.³⁵ Effective management of these variables is essential to achieving high-quality composite structures. To study the relationship between these parameters and the fluid front concerning materials, many mathematical models have been proposed that account for time and temperature dependencies. The relationship between the fluid flow front and process parameters is described by the Arrhenius equation as shown in equation (1).³⁶ Where η(T, t) represents the viscosity at temperature T, at a time t, ʽ $η_{\infty}$ ʼ is calculated viscosity of the material at $T = \infty$ , ʽ ${∆ E}_{n}$ ʼ is the Arrhenius activation energy, ʽRʼ is the gas constant and ʽ $k_{\infty}$ ʼ is an apparent kinetic factor. This equation highlights the critical role of matrix viscosity in determining the fluid flow front during the RTM process, which directly influences the final mechanical properties of the composite structure. The relationship between viscosity and fluid flow is further governed by Darcy’s Law, which describes the flow behavior in porous media. According to Darcy’s Law, the resin flow rate is inversely proportional to viscosity, making the temperature-dependent changes in viscosity, as predicted by the Arrhenius equation, crucial for optimizing process parameters a shown in equation (2). Where Q is volumetric flow rate, k is permeability of the fiber preform, ΔP is pressure gradient, A is cross-sectional area and µ is dynamic viscosity of the resin that corresponds to $η (T, t)$ from Arrhenius equation. Based on this understanding and findings from previous literature, both injection pressure and mold temperature parameters that significantly influence viscosity have been extensively investigated in this study. These parameters were selected to ensure efficient resin infiltration, minimize void formation, and optimize the mechanical performance of the hybrid structures.³⁷

I n η (T, t) = I n η_{\infty} + \frac{∆ E_{n}}{R T} + \int_{0}^{t} k_{\infty} \exp (\frac{∆ E_{k}}{R T}) d t

(1)

Q = \frac{k}{µ} ∆ P A

(2)

Another critical factor in producing hybrid structures is the interface between metal and FRP. The interface plays a major role in metal-FRP hybrids as it is crucial for transferring loads between them, directly influencing the component’s mechanical properties.^38–41 Modifying a suitable polymer to increase bonding strength at the interface between metal and FRP is highly challenging and costly due to the differing physicochemical properties of the two materials. However, increasing the interlocking strength through modified metal surfaces can be easily and efficiently implemented.^42,43 These surface treatments are classified into three categories: mechanical, chemical, and energy methods. Mechanical methods include techniques such as grit blasting, sanding, and creating micro-grooves on the metal surface. These techniques increase the roughness and surface area, providing more contact points for the FRP to adhere to. Chemical methods involve treatments like anodizing or applying chemical primers that enhance the chemical bonding at the interface.⁴⁴ Energy methods, such as laser ablation or plasma treatments, modify the surface properties through energy application, creating a more reactive surface that improves bonding with FRP. For instance, laser ablation, an advanced energy method, uses laser beams to precisely remove material from the metal surface, creating micro-scale features that significantly enhance the bonding area.⁴⁵ This method is highly controllable and can be tailored to create specific surface patterns that optimize the mechanical interlock between metal and FRP. In the current study, mechanical methods based on sandblasting were used for altering the metal surface to increase the surface area and create undercuts. This approach ensures better mechanical interlocking, leading to improved load transfer and enhanced mechanical properties of the hybrid structure.

Materials and methods

In the current study, various structures of an automobile were studied to find the potential application of hybrid structures that results in weight saving and at the same time improved properties. Based on literature, some of potential areas that can use hybrid structures in an automotive are A-pillar, B-pillar, C-pillar, Crash-box, and Drive shaft. One of the common challenges in hybrid structures is residual stresses generated due to different thermal expansion coefficients of metals and FRPs. However, this can be reduced with tailoring the fiber directions or can be used to increase the adhesive force between metal and FRP if the structure is cylindrical. Out of five structures mentioned above, drive shaft cross-section is cylindrical in shape that can use residual stresses for a better adhesiveness between metal and FRP. Based on this approach, a hollow cylindrical shaft geometry that resembles the cross section of vehicle’s drive shaft was selected. The suitability of the RTM process for producing such rotationally symmetrical parts is also highlighted, as these cannot be effectively manufactured using methods like prepreg pressing.

The dimensions of the hybrid drive shaft were selected based on the literature survey. Diameter, thickness, and length are three different dimensions that should be determined. The thickness of the drive shaft should be as minimum as possible so that the inertia is kept to minimum. The typical thickness of shafts ranges from 3 mm to 15 mm depending on the type of vehicle. Diameters range from 42 mm to 60 mm depending on the material used and type of vehicle. Length of the dive shaft depends on the type of drive shaft configuration and material used that ranges from 850 mm to 2000 mm. The configuration of the drive shaft is classified based on the number of segments it’s made from. For example, a three-part drive shaft is dived into three segments and is connected using universal joint to reduce the bending loads when the length of power train is too long. In the current work, the thickness of the drive shaft was kept to minimum that is 3 mm and a mid-value of 54 mm was selected for the outer diameter. The thickness of the metal is 1 mm, whereas CFRP is 2 mm thick. The selected 2 mm thickness ensures proper distribution of stresses across the CFRP layer while maintaining adequate bonding with the underlying metal, contributing to optimal load transfer. This balance of material thickness also helps in reducing manufacturing complexities and achieving a uniform resin flow during the RTM process, which is critical for preventing voids and ensuring consistent fiber wetting. Since the length of the drive shaft has lesser influence as they can be made from, number of segments was kept to a minimum value, which is 320 mm.

Materials

Precision steel tube made of E235 + N was used as the metal part for the hybrid drive shaft in the current work. Table 1 shows the mechanical properties of the selected steel tube. Additionally, to increase the contact area, the shafts were sand blasted using SMG 25 Duo with sand particle size in range of 210 μm to 297 μm with a pressure of 4 - 4.5 bars. Dry fiber tubes made of Torayca® T300 J 400 tex (6k) carbon fiber was used for the hybrid drive shaft. The fiber tube has a fiber angle of 45° with a diameter of 60 mm and thickness of 0.27 mm. The diameter of the fiber hose can be adjusted by stretching or compressing, which likewise changes the fiber angle. Figure 2 shows the relationship between fiber angle and the diameter. The figure is sourced directly from the data sheet provided by R&G Faserverbundwerkstoffe GmbH. The diameter in the current study varies from 50 to 54 mm, resulting in the fiber angle of 35° to 37°. The matrix system used in this work is an epoxy resin of type EPIKOTE Resin 05,475 and the amine hardener of type EPIKURE Curing Agent 05,443, both manufactured by Momentive Performance Materials Quartz GmbH. The matrix system shows a relatively low viscosity and a short curing time. The viscosity of the matrix system is highly dependent on temperature as shown in Figure 3. The figure is sourced directly from the data sheet provided by Lange+Ritter GmbH. The viscosity of the resin-hardener mixture with respect to temperature can be seen in Table 2. Based on the data provided in Table 2 and the trend observed in Figure 3, the viscosity at 120°C is calculated to be ∼6 ± 2 mPa.s, while at 140°C, it is ∼3 ± 1 mPa.s reflecting the significant reduction in viscosity at elevated temperatures. These values were estimated using a simple linear interpolation method, based on the experimental data available for nearby temperatures. In addition, the temperature has a great influence on the curing speed. The higher the initial matrix temperature, the faster the matrix system cures. A powder version of EPIKOTE Resin 05,475 was used as a binder to bind the dry fibers in preform stage. The amount of binder used in each sample is calculated based on 1 % - 2 % tolerance between EPIKOTE Resin 05,475 and the amine hardener of type EPIKURE Curing Agent 05,443. Despite all advantages of using hybrids, the biggest issue is when CFRP and metals are joined, the metal structures are prone to become galvanically active resulting in corrosion because of the presence of Carbon element in CFRP that acts as an open-circuit potential (OCP). In humid conditions, a perfect galvanic cell is formed resulting in galvanic corrosion attack on the metal structure. To mitigate the risk of galvanic corrosion, a glass nonwoven layer (Johns Manville ForTex Type SH35/1) was used as an intermediate layer between the steel and CFRP. This layer prevents direct contact between the carbon fibers and the steel tube, reducing the potential for corrosion caused by electrochemical interactions. The glass nonwoven layer, with a thickness of 0.37 mm and a fiber diameter of 13 μm, provides both dimensional stability and good wet strength, ensuring effective isolation without compromising structural integrity. In current study, a few samples with glass nonwoven consisting of glass fibers as an interface layer are used that have an average fiber diameter of 13 μm. Nonwoven is characterized by dimensional stability and good wet strength.

Table 1.

Material properties of Precision steel tube.

E235+C
Yield strength/MPa	235 – 355
Tensile strength/MPa	480 – 640
Elongation at break/%	≥4 - 6

Figure 2.

Angle of fibers with respect to different diameters.

Figure 3.

Viscosity with respect to time at various temperatures of epoxy based Epikote Resin 05,475 and Curing Agent 05,443.

Table 2.

Material properties of epoxy.

EP 05475 + EK 05443 (100:24)
Viscosity/mPa.s	At 25°C	1200 ± 100
	At 80°C	30 ± 5
	At 100°C	13 ± 3
Pot Life/min	At 25°C	120 ± 10
Gel Time/sec	At 80°C	330 ± 30
	At 100°C	210 ± 30
	At 120°C	150 ± 30
	At 140°C	90 ± 30

Hybrid sample production

Sample productions were conducted in two phases: Phase 1 focused on investigating parameters, and Phase 2 examined the interface between steel and CFRP. To reduce the number of samples and get more reliable relations, the design of experiments (D.O.E) was used.⁴⁶ There are two techniques in D.O.E that can be used for two variables with three parameters each, which are the Two Factorial method and Taguchi method. In the current study, the Taguchi method was selected due to its ability to efficiently control the number of experiments while evaluating the influence of parameters. A Taguchi L9 orthogonal array was used, incorporating 2 parameters at three levels, resulting in a total of nine experimental runs, as shown in Table 3. The mold temperatures were selected based on the selected matrix system and manufacturing method used in present work.

Table 3.

Sample types based on Taguchi L9 array for parameter study.

Sample	Mold temperature / °C	Injection pressure / bar
Sample 01PA_01IN	80	5
Sample 02PA_01IN	80	7
Sample 03PA_01IN	80	9
Sample 04PA_01IN	100	5
Sample 05PA_01IN	100	7
Sample 06PA_01IN	100	9
Sample 07PA_01IN	120	5
Sample 08PA_01IN	120	7
Sample 09PA_01IN	120	9

The production of samples starts with cutting the dry fibers based on the dimensions required. A total of six layers of dry fibers are cut to form a 2 mm thick composite layer with 55 % fiber volume. The CFRP are stacked on the steel one by one with binders added for each layer manually. Once the fibers are stacked with binder, they are tapped with a special shrinkage tape of 0.05 mm thickness, as shown in Figure 4. The sample is moved to the mold for preforming, where a temperature of 140°C is applied for 10 min. Due to this step, the binder melts and then solidifies when kept at room temperature creating a bond between dry fibers of different layers. This preform stage prevents fiber deformation or fiber washout during injection phase. After discarding the release foil, the preform is placed in the mold. The mold is now closed, and the injection pressure is adjusted according to the sample number based on Table 3.

Figure 4.

Schematic diagram showing the production cycle of Hybrid sample; (a) adding glass fleece; (b) adding CFRP; (c) adding binders; (d) tapping the sample with shrinkage tape for preforming and (e) Hybrid shaft.

Another parameter that was controlled is the initial mass flow rate of the matrix system before the process reaches the injection pressure. Once the injection pressure is reached, the mass flow rate is automatically reduced to maintain the given injection pressure. An offset of 0.5 bar was given in the machine so that the pressure can range from 6.5 bar to 7.5 bar for 7 bar injection pressure. Figure 5 show the injection pressure and mass flow rate recorded over the time during the injection phase. The initial mass flow rate was kept at 0.15 kg/min for all hybrid samples. The machine has a record frequency of 1 Hz. Additionally, for all the hybrid samples, the machine was programmed to inject 0.21 kg of resin into the mold and stop automatically. Based on theoretical calculation, 0.14 kg of matrix system should be injected to maintain 55 % fiber volume for the hybrid samples produced. However, an extra 0.07 kg was added to offset the matrix wasted because of inlets head volume and to maintain a minimum of 90 sec on injection time. The excess resin is collected at the outlet vents that are equipped with resin trap pot. During the entire production cycle, the mold was connected to a vacuum pump via outlet vents at both ends of mold and maintained at ∼0.015 bar (absolute pressure). Figure 6 shows the production setup of the current study.

Figure 5.

Production data points when injection pressure was set to 5 bar and initial mass flow rate of 0.15 kg/min.

Figure 6.

Schematic diagram showing the production setup for VA-LRTM.

In addition to the parameter study, the interface between CFRP and steel was investigated in the current work, referred to as Phase 2. Three unique interface modifications were applied to the samples: first, the steel tube underwent sandblasting; second, the contact layer of steel was replaced with a glass fleece made from glass fibers; and third, a combination of both modifications was applied. In the first modification, the steel tube was sandblasted to assess whether changes in surface roughness could influence the ILSS between CFRP and steel. The surface roughness values for Ra (average roughness) increased from 1.41 μm to 2.83 μm, and for Rz (peak-to-valley roughness) from 2.13 μm to 4.27 μm. This corresponds to an approximate depth alteration of 15 – 20 μm, attributed to the particle size and applied pressure during the process. In the second modification, the outermost layer of the CFRP was replaced with a glass fleece layer. This glass fleece has a greater thickness, and a lower fiber density compared to CFRP, allowing for more resin at the interface, potentially improving bonding. The layer thickness of CFRP is 0.27 mm, while the glass fleece is slightly thicker at 0.35 mm. For the third modification, both treatments were combined: the steel tube was sandblasted, and the last layer of CFRP was replaced with glass fleece, aiming to maximize the interface modifications for improved bonding. These modifications were evaluated using the optimized parameters from Phase 1, with all other experimental conditions held constant, focusing solely on the effect of the interface changes. Table 4 outlines the specific interface samples that were analyzed.

Table 4.

Sample types used in interface investigation.

Sample	Steel type	Glass fleece	Parameters
Sample 10PA_01IN	Normal	No	▪ Mold temperature – 120°C
Sample 10PA_02IN	Normal	Yes	▪ Injection pressure – 5 bar
Sample 10PA_03IN	Sandblasted	No	▪ Mass flow rate - 0.15 kg/min
Sample 10PA_04IN	Sandblasted	Yes

Results and discussion

Interlaminar shear strength (ILSS)

Samples according to Table 3 were produced with respective injection pressure and mold temperature. To evaluate these samples, the interlaminar shear strength (ILSS) for the cylindrical shafts was assessed, based on the shear strength testing of flat composite specimens. The hybrid shafts were sectioned into 20 rings, each 5 mm thick, and subjected to shear testing using a shear-edge device, which pressed the steel ring out of the CFRP structure as shown in Figure 7. The recorded force was normalized against the steel/CFRP interface area to determine shear strength, reflecting the interfacial bonding quality. By designing the die and punch precisely, the force flow was directed through the interface, establishing a defined stress state, which accurately reflects the interfacial shear strength rather than simply adhesive strength.

Figure 7.

Schematic diagram showing sample sliced across its length into specimens for testing their respective interlaminar shear strength (ILSS).

The samples were tested and grouped based on injection pressure and mold temperature as shown in Figure 8. The results are based on an average value of 20 specimens on each sample and their standard deviation is recorded as an error bar. A difference of 26 % in ILSS can be noted between stronger samples (injection pressure – 5 bar & mold temperature 120°C) and weaker samples (injection pressure – 9 bar & mold temperature 80°C). Based on mold temperature, a liner trend can be spotted across the nine samples. Even with different injection pressures, higher mold temperature has better ILSS values. This phenomenon can be derived from the temperature versus viscosity of the matrix system. Higher temperature results in lower viscosity that makes the matrix flow better between fiber tows and increases its wetting property of the fibers. Further increase in mold temperature from 120°C can’t be considered for the selected resin as the Gel time is reduced to 90 ± 30 sec at 140°C affecting the injection time. Injection time is the total time the machine needs to pump the 0.21 kg of resin into the mold.

Figure 8.

ILSS results of samples with respect to mold temperature.

Like mold temperature, a linear trend can be observed across all nine samples when grouped based on injection pressure. Change in ILSS value with respect to injection pressure can be possibly related to two factors, i.e., fill time and fiber deformation near injection area. Fill time is the time recorded when resin enters the mold and spotted at the outlet vents for first time. Figure 9 shows the fill time recorded for the samples with respect to injection pressure and mold temperature. The data is averaged based on three samples for each case. The sample with the longest fill time of 58 sec has an ILSS value of 11.4 MPa, which is 15 % less than the maximum ILSS value recorded for a sample that has a fill time of 52 sec. However, sample with shortest fill time of 43 sec has an ILSS value of 11.4 MPa that is same ILSS value of the sample with longest fill time. Based on this statistical analysis, a proper relation could not be concluded on the fill time. Another factor that depends on the injection pressure is fiber deformation that occurs near injection area. To check this influence factor, the ILSS value across each sample’s length was investigated. Three different trends were observed based on this analysis. The first noticeable factor is that for all samples, the ILSS value near the injection area is lowest, that lies between specimen 10 and specimen 11. The second notable factor is the standard deviation near the injection zone is high in comparison with the end specimens of the sample, which implies that the values near the injection zone are inconsistent. The third notable factor is that the ILSS value across the sample tends to decrease, and the standard deviation tends to increase with decrease in mold temperature. Figure 10 shows the ILSS value recorded across the two samples that had maximum ILSS value and minimum ILSS value. The values are based on three samples for each location and standard deviation is recorded as the error. Based on these data, it was concluded that a pressure of 5 bar and a mold temperature of 120°C for current geometry and mold design is suitable based on ILLS property.

Figure 9.

Fill time of samples with respect to injection pressure.

Figure 10.

ILSS values across sample length based on three specimens averaged for data point.

Similarly, the interface-modified samples were tested, and the results are presented in Figure 11. A significant increase of approximately 40 % in ILSS was observed when both sandblasting and glass fleece were used in combination, compared to the baseline sample without any modifications. When comparing the effects of glass fleece and sandblasting separately, the glass fleece-modified sample showed a 33 % increase in ILSS, while the sandblasting-only approach resulted in a 14 % increase, both relative to the unmodified sample. The increase in ILSS for the sandblasted sample can be attributed to improved mechanical interlocking between the steel and CFRP surfaces due to the roughened texture. In the case of glass fleece modification, the higher ILSS is likely due to the increased resin content at the interface, which improves adhesion. The fiber content in the glass fleece sample is 51 % (46 % CF + 5 % GF), lower than the 55 % fiber content (all CF) in both the unmodified and sandblasted samples. This reduction in fiber content is due to the glass fleece layer replacing part of the CFRP, which may also contribute to reduced galvanic corrosion at the metal-composite interface.

Figure 11.

ILSS results of interface samples with same production parameter.

In conclusion, the combination of sandblasting and glass fleece significantly improves the ILSS, demonstrating the effectiveness of optimizing both mechanical interlocking and resin-rich interfaces for enhancing the performance of CFRP-steel hybrid structures. Building on these findings, further investigations were conducted to assess the behavior of the hybrid components under compression and torsion tests. The compression tests were designed to evaluate the energy absorption capacity of hybrid structures, which is critical for applications requiring high impact resistance or crashworthiness. By analyzing the hybrid components to absorb and dissipate energy, the study aimed to provide insights into their performance under dynamic loading conditions. Additionally, torsion tests were performed to evaluate the shear stress distribution and behavior under rotational loads, further assessing the structural integrity and mechanical performance of the components under torsional stress. These tests were conducted on two types of samples: the first being a baseline sample with an unmodified interface, and the second incorporating both sandblasting and glass fleece modifications. A steel-only sample was also included as a reference to compare the behavior of the hybrid structures against a traditional material under similar loading conditions.

Compression test

Compression tests were conducted using a servo motor spindle press capable of applying up to 300 kN of force. The samples, originally 320 mm in length, were cut down to 150 mm using a saw machine to ensure consistency. For each test, the same part of the sample, relative to the mold marking, was consistently positioned in the press, with the center of the sample always aligned with the moving part of the press, as illustrated in Figure 12. A specialized load cell was installed on the press to accurately measure both displacement and corresponding forces. The test speed was set to 5 mm/min, and samples were compressed until they were reduced by 90 – 100 mm of their length. Two types of samples were selected for the compression tests based on the interface study. The first, referred to as the normal hybrid sample, consisted of unmodified steel and six layers of CFRP. The second, known as the modified hybrid sample, had sandblasted steel combined with a glass fleece layer and five layers of CFRP. Additionally, a steel tube without CFRP was tested as a reference. The steel tube exhibited buckling and diamond buckling deformation modes (Figure 13(a)), while the hybrid shafts displayed diamond buckling along with shear cracking and wrinkling (Figure 13(b)). In addition, a common observation across all hybrid samples was the complete debonding of the steel and CFRP.

Figure 12.

Compression test setup of Hybrid shafts.

Figure 13.

Samples after compression test (a) Steel; (b) Normal Steel + CFRP and (c) Sandblasted Steel + Glass Fleece + CFRP.

In terms of peak force, the hybrid tubes outperformed the steel reference tube, showing a 60 % increase in peak force. The average peak force for the hybrid samples increased by 88 % compared to the steel tube. The normal hybrid sample recorded a maximum peak force of 120 kN, while the modified hybrid sample achieved a slightly lower peak force of 113 kN, representing a 6 % decrease. However, the average force for the modified sample was 64 kN, which is 8 % higher than the normal sample’s mean force of 59 kN. Figures 14 and 15 shows the forces versus distance recorded during the compression test.

Figure 14.

Compression results of interface sample (Normal steel + CFRP) with respect to steel tube.

Figure 15.

Compression results of interface sample (Sandblasted steel + Glass Fleece + CFRP) with respect to steel tube.

Key performance metrics for these compression tests were energy absorption and specific energy absorption, both of which are calculated using equation (3) and equation (4) that are plotted in Figure 16. Where EA is energy absorption, $μ$ is mass of the sample, d is displacement and $P_{mean}$ is mean force. The modified hybrid sample exhibited an 8 % higher specific energy absorption and total energy absorption compared to the normal sample. However, in some instances, the results for both sample types were very close, as indicated by the error bars derived from the standard deviation calculations. To compare the hybrid samples with the steel tube, specific energy absorption was used due to the difference in mass between the steel tube and the hybrid samples. The hybrid samples showed an 89 % improvement in specific energy absorption over the steel tube.

E A = \int_{0}^{d} F (x) ⅆ ϰ

(3)

S E A = \frac{E A}{μ d} = \frac{P_{m e a n}}{μ}

(4)

In conclusion, while the interface modifications did not significantly impact specific energy absorption or total energy absorption, hybrid samples exhibited a considerable increase in energy absorption compared to the steel tube. This suggests that hybrid structures, regardless of interface modifications, can substantially enhance energy absorption, making them ideal for applications requiring lightweight, impact-resistant materials.

Figure 16.

Total energy absorption and specific energy absorption of interface samples with respect to steel tube.

Torsion test

The torsional stiffness of the hybrid sample was determined using torsion setup based on the real load situation. A hydraulic rotary cylinder based on the double vane principle was used to introduce the torsional moment on the test samples. These valves can generate a torque of ±1000 Nm at a torsion angle of ±50°. The structure of the torsion test setup consists of rotating cylinder equipped with an angle of rotation and torque sensor. The clamps and load introduction are connected to the test specimen by means of a force-fit with the aid of annular springs as shown in Figure 17. The left side of the samples are fixed, whereas the right side of the samples is subjected to twist which is controlled by the sensors and the cylindrical values. Quasi-static tests were performed at 5°/min, for both normal and modified sample and data was recorded as shown in Figures 18 and 19. The points represent the data values recorded, and the line represents trend line. Torsional stiffness values of both hybrid samples are same with a small deviation of 3 % in their value i.e., normal hybrid samples had a torsion stiffness of 427 Nm/deg., whereas modified samples had a torsion stiffness of 440 Nm/deg. This indicates that interface modifications, such as sandblasting or the inclusion of glass fleece, have negligible influence on the torsional stiffness of the hybrid structures. For comparison, a reference torsion test was conducted on a steel tube without any CFRP reinforcement. The torsional stiffness recorded for the steel sample was 324 Nm/deg., which is 35 % lower than the hybrid samples. To validate the experimental findings, the torsional stiffness of the steel tube was calculated using equation (5) and equation (6). Where K is the torsional stiffness (Nm/rad); G is shear modulus (Pa); L is Length (m); J is polar moment of inertia (m⁴); D is outer diameter (m) and d is inner diameter (m). The torsional stiffness calculated for the steel tube was 18,488 Nm/rad, which can be equated to 323 Nm/deg. The input values used to calculate torsional stiffness are, polar moment of inertia of 92,440 mm⁴; Shear modulus of 70 GPa. In conclusion, while interface modifications did not significantly affect the torsional stiffness of the hybrid samples, the overall performance of the hybrid structures in torsion far exceeded that of the steel tube, demonstrating the superior mechanical properties imparted by the CFRP reinforcement. This makes the hybrid structures more suitable for applications requiring high torsional rigidity.

K = \frac{G J}{L}

(5)

J = \frac{π}{32} (D^{4} - d^{4})

(6)

Figure 17.

Torsion test setup for interface samples.

Figure 18.

Torsion results of Normal sample (Normal steel + CFRP) with respect to steel tube.

Figure 19.

Torsion results of modified sample (Sandblasted steel + Glass Fleece + CFRP) with respect to steel tube.

CT investigation

As the next step in the study, a non-destructive test was performed using CT scanning to investigate the porosity within the CFRP-steel hybrid samples. CT scanning was selected due to its ability to provide high-resolution 3D models of internal structures without damaging the samples, making it ideal for detailed porosity analysis. The CT scan was conducted using a x|cube 225 device from GE, equipped with VG Studio Max 2022.1 software for the reconstruction and analysis of 3D models. The scans were performed at a voltage of 200 kV and a current of 2 mA. The X-ray gun used had a focus diameter of 0.4 mm, with an exposure time of 100 ms. To filter out weaker X-rays, a 2 mm copper filter was applied. The CT scan consisted of 1440 rotational steps to ensure optimal reconstruction of the 3D model. Based on the CT models, the total porosity in the hybrid samples was calculated to be 2.47 %, with a standard deviation of 0.04. The observed pore diameters ranged from 0.03 mm to 2.01 mm. Notably, most of the porosity was concentrated near the interface between the steel and CFRP layers, as shown in Figure 20. This suggests that factors such as resin flow or bonding processes may need further refinement to reduce void formation. Identifying porosity at the interface is critical, as it can negatively impact mechanical properties like shear strength and fatigue resistance. These results demonstrate that further optimization of the interface, through improved bonding techniques or material modifications, could enhance the overall structural integrity and reduce porosity in key areas.

Figure 20.

Pores distribution in Hybrid sample.

Conclusions

This study investigated the development and mechanical characterization of CFRP-steel hybrid structures produced via Resin Transfer Molding (RTM) with vacuum assistance. The influence of key process parameters, including injection pressure, mold temperature, and interface modifications, was evaluated in terms of interlaminar shear strength (ILSS), compression performance, torsional stiffness, and porosity. Optimization of the RTM process revealed that an injection pressure of 5 bar and a mold temperature of 120°C resulted in the highest ILSS values. This was attributed to better resin flow and fiber wetting due to lower viscosity at elevated temperatures. Furthermore, interface modifications combining sandblasting of the steel surface and the inclusion of glass fleece improved ILSS by 40 %, enhancing mechanical interlocking and increasing resin content at the interface. Compression testing showed that hybrid structures exhibited 60 % higher peak force, and 89 % greater specific energy absorption compared to traditional steel tubes. Torsional stiffness testing demonstrated that hybrid samples provided a 35 % improvement in stiffness over steel tubes, though interface modifications had minimal effect on torsional rigidity, suggesting that CFRP layers primarily contribute to increased torsional strength. Computed tomography (CT) scanning revealed that most of the porosity was concentrated near the interface, with an average porosity of 2.47 % and pore sizes ranging from 0.03 mm to 2.01 mm. This highlights a need for further refinement of the resin flow and bonding processes to minimize porosity at critical interfaces, thus improving overall structural integrity and fatigue resistance.

In conclusion, CFRP-steel hybrid structures present a highly effective alternative to traditional metal components, offering significant advantages in weight reduction, energy absorption, and torsional stiffness without sacrificing mechanical performance. The findings of this study provide a solid foundation for further optimization of hybrid structures, with a focus on reducing porosity and improving interfacial properties to enhance durability and performance in automotive and aerospace applications.

Footnotes

Acknowledgments

The authors would like to express their gratitude to the University of Paderborn for providing the research facilities and resources necessary for the successful completion of this study. Special thanks are extended to the technicians at LiA chair, for their invaluable technical support and assistance throughout the experimental phase of this research. Without their expertise and contributions, this work would not have been possible.

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) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Deviprasad Chalicheemalapalli Jayasankar

Thomas Kordisch

Data availability statement

All data generated or analyzed during this study are included in this published article*.

References

Kruckenberg

Paton

. Resin transfer Moulding for aerospace structures. Springer Science, 1998.

Sinmazçelik

Avcu

Bora

MÖ

Çoban

. “A review: fibre metal laminates, background, bonding types and applied test methods”. Mater Des 2011; 32: 3671–3685, Bd.

Lee

J-M

Min

B-J

Park

J-H

, et al. “Design of lightweight CFRP automotive part as an alternative for steel part by thickness and lay-up optimization”. Materials 2019; 12(Nr. 14): 3.

Min

Sun

, et al. “Fabrication processes of metal-fiber reinforced polymer hybrid components: a review”. Advanced Composites and Hybrid Materials 2021; 2: 3.

Lin

Min

Teng

, et al. “Flexural performance of steel–FRP composites for automotive applications”. Automot Innov 2020; 3(3): 280–295.

Kavitha

Vijayan und

Sathishkumar

. “Fibre-metal laminates: a review of reinforcement and formability characteristics”. Mater Today: Proceedings 2020; 22(Nr. 3): 601–605, Bd.

Lin

Zhang

, et al. “Crushing characteristics comparison between aluminum/CFRP and aluminum/CFRP/aluminum hybrid tubes”. Polymers. 2022. https://doi.org/10.3390/polym14194034

Sun

Wang

Hong

, et al. “Experimental investigation of the quasi-static axial crushing behavior of filament-wound CFRP and aluminum/CFRP hybrid tubes”. Compos Struct 2018; 194: 208–225, Bd.

Yang

Lei

, et al. “Energy absorption and failure pattern of hybrid composite tubes under quasi-static axial compression”. Compos B Eng 2020; 198: 108217.

10.

Han

, et al. Crushing characteristics of aluminum/CFRP/aluminum hybrid tubes prepared by spinning forming. Compos Struct 2020; 249: 112551.

11.

Boria

Belingardi

. “Numerical investigation of energy absorbers in composite materials for automotive applications”. International Journal of Crashworthiness 2011; 17(Nr. 4): 4, Bd.

12.

Boria

Pettinari

Giannoni

, “Theoretical analysis on the collapse mechanisms of thin-walled composite tubes”. Compos Struct 2013; 103: 43–49, Bd.

13.

Kurihara

. “Polymer matrix composite materials in automobile industries”. Adv Compos Mater 1995; 4(Nr3): 209–219.

14.

Seydibeyoğlu

MÖ

Doğru

Kandemir

Aksoy

. Lightweight polymer composite structures. CRC Press, 2020.

15.

Wazeer

Das

Abeykoon

, et al. “Composites for electric vehicles and automotive sector: a review”. Green energy and Intelligent Transportation. 2023. https://doi.org/10.1016/j.geits.2022.100043.

16.

Wilson

. “Vehicle weight is the key driver for automotive composites”. Reinforced Plastics 2017; 61(Nr2): 7, Bd.

17.

Marmol

Ferreira

Fangueiro

. “Automotive and construction applications of fiber reinforced composites”. Fiber Reinforced Composites 2021; 6: 785–819.

18.

Sateesh

Subbiah

Nookaraju und

, et al. “Achieving safety and weight reduction in automobiles with the application of composite material”. Mater Today: Proceedings 2022; 6: 4469–4472, Bd. 62, Nr.

19.

Arenas

Z. L. B . “Composites and biocomposites for noise and vibration control in automotive structures”. Biocomposite and Synthetic composites for automotive applications, 2021; 1: 305–345.

20.

Sarfraz

Hong

Kim

. Recent developments in the manufacturing technologies of composite components and their cost-effectiveness in the automotive industry: a review study. Compos Struct 2021; 266: 113864.

21.

Potter

. Resin transfer Moulding. Chapman & Hall, 1997.

22.

Hsiao

K-T

. “Vacuum assisted resin transfer molding (VARTM) in polymer matrix composites”. Manufacturing Techniques for Polymer Matrix Composites (PMCs) 2021; 6: 310–347.

23.

Naik

Sirisha

Inani

. “Permeability characterization of polymer matrix composites by RTM/VARTM”. Prog Aero Sci 2014; 65: 22–40, Bd.

24.

Salvatori

Caglar

Teixidó

, et al. “Permeability and capillary effects in a channel-wise non-crimp fabric”. Compos Appl Sci Manuf 2018; 108: 41–52. Bd.

25.

Zarandi

MAF

Arroyo

Pillai

, “Longitudinal and transverse flows in fiber tows: Evaluation of theoretical permeability models through numerical predictions and experimental measurements”. Compos Appl Sci Manuf 2019; 119: 73–87, Bd.

26.

Teresa

. Kruckenberg und Rowan Paton. Resin Transfer Moulding for Aerospace Structures; 199: 1–24.

27.

Baskarana

Ortiz de Mendibil

Sarrionandia

, et al. Manufacturing cost comparison of RTM, HP-RTM and CRTM for an automotive roof. In European conference on composite materials. Spain, 2014.

28.

Chaudhari

Karcher

Elsner

, et al. “Characterization of high pressure RTM processes for manufacturing of high performance composites”. In 15th European Conference on Composite Materials. Venice, 2012, pp. 24–28.

29.

Wen-Bin

Cheng-Wey

. “Study on compression transfer molding”. J Compos Mater 1995; 29: 2180–2191, Bd.

30.

Bhat

Merotte

Simacek

, et al. “Process analysis of compression resin transfer molding”. Compos Appl Sci Manuf 2009; 40: 431–441, Bd.

31.

Ageyeva

Sibikin

Kovács

. “A review of thermoplastic resin transfer molding: process modeling and simulation”. polymers 2019; 11(Nr10): Bd. https://doi.org/10.3390/polym11101555.

32.

Lin

Murphy

Hahn

. “Resin transfer molding process optimization”. Compos Appl Sci Manuf 2000; 21(Nr4): 361–371, Bd.

33.

Haywond

Harris

. “Effect of process variables on the quality of RTM mouldings”. SAMPE J 1990; 26(Nr1): 39–46, Bd.

34.

Lundstrom

Gebart

. “Influence from process parameters on void formation in resin transfer molding”. Polym Compos 1994; 15(Nr1): 25–33, Bd.

35.

Matsuzaki

Todoroki

SKA

Todoroki

. “Flow control by progressive forecasting using numerical simulation during vacuum-assisted resin transfer molding”. Compos Appl Sci Manuf 2013; 45: 79–87, Bd.

36.

Roller

. “Characterization of the time-temperature-viscosity behavior of curing B-staged epoxy resin”. Polymer Enginerring and Sciecne 1975; 15(Nr6): 7, Bd.

37.

Kaynak

Kas

. Effects of injection pressure in resin transfer moulding (RTM) of Woven carbon fibre/epoxy composites. Polymers & Polymer Composites 2006; 14(Nr1), Bd.

38.

Heckert

Zaeh

. “Laser surface pre-treatment of aluminium for hybrid joints with glass fibre reinforced thermoplastics”. Phys Procedia. 56: 1171–1181, Bd.

39.

Steinert

Dittes

Schimmelpfennig

, et al. Design of high strength polymer metal interfaces micro structured surfaces. IOP Conf Ser Mater Sci Eng 2018.

40.

Min

Wan

, et al. “Effect of adhesive type on mechanical properties of galvanized steel/SMC adhesive-bonded joints”. Int J Adhesion Adhes 2020; 97: 102482, Bd.

41.

Carey

Cantwell

Dearden

, et al. “Towards a rapid, non-contact shaping method for fibre metal laminates using a laser source”. Int J Adv Manuf Technol 2010; 47: 557–565.

42.

Shanmugam

Kazemi

Rao

, et al. “On the metal thermoplastic composite interface of Ti alloy/UHMWPE-Elium® laminates”. Compos B Eng 2020; 7: 181.

43.

Zinn

Bobbert

Dammann

, et al. “Shear strength and failure behaviour of laser nano-structured and conventionally pre-treated interfaces in intrinsically manufactured CFRP-steel hybrids”. Compos B Eng 2018; 151: 173–185, Bd.

44.

Sun

Yang

, et al. “Advanced high strength steel (AHSS) development through chemical patterning of austenite”. Scripta Mater 2018. 146(Nr. 15): 60–63, Bd.

45.

Min

Wan

Carlson

, et al. Application of laser ablation in adhesive bonding of metallic materials: a review. Opt Laser Technol 2020; 128: 106188.

46.

Evans

, MINITAB manual, 2009.