Author: [Bruce Zhou]
Affiliation: [Jota Machinery Composites Material Prepreg Solution]
Corresponding Author: [jotamachinery@gmail.com]

Abstract

This paper presents a comprehensive reconstruction and reinterpretation of the vacuum-assisted induction welding (Vac-IW) process for assembling all-thermoplastic composite sandwich panels. Building upon the experimental observations of localized interface heating, hysteresis-based susceptors, and dual-polymer bonding architectures, this study analyzes how welding speed, susceptor morphology, and core integrity govern the resulting mechanical performance. Using glass-fiber/PEEK skins with a PEI surface layer, a 3D-printed PEI honeycomb core, and a Ni-particle-filled PEI susceptor, the Vac-IW process demonstrates the ability to selectively heat the bondline without degrading the bulk laminate or collapsing the core. Flatwise tensile strengths up to 6 MPa were achieved—matching or exceeding values reported for adhesive-bonded thermoplastic sandwich structures, despite the absence of surface preparation and adhesives. Microscopy reveals that weld quality is controlled through the coupling of thermal softening, susceptor squeeze-flow, and contact area evolution. This new interpretation identifies a clear process window between 0.5–0.9 mm/s, where interface temperature, pressure distribution, and polymer chain mobility achieve the optimal balance. The paper concludes with implications for aerospace-grade thermoplastic structures, guidance for machinery manufacturers, and directions for further development of local-heating welding technologies.

Keywords

Thermoplastic composites; induction welding; vacuum-assisted welding; sandwich structures; PEI; PEEK; Ni susceptor; localized heating; honeycomb core; flatwise tensile strength.

1. Introduction

High-performance thermoplastic composites—particularly those based on PAEK polymers—are increasingly used in aerospace, space, automotive, and mobility applications due to their high toughness, excellent chemical resistance, rapid processing times, and end-of-life recyclability. When combined with lightweight honeycomb or foam cores, they form sandwich structures that deliver outstanding stiffness-to-weight ratios and exceptional vibration, impact, and thermal performance.

However, the critical challenge lies in joining the skin and core without damaging either. Conventional thermal joining methods heat the entire panel thickness, risking:

• Core collapse due to elevated temperatures.
• Skin deconsolidation when exceeding polymer melt temperature.
• Uneven heat transfer through low-conductivity cores.

Adhesive bonding avoids catastrophic overheating but introduces volatile outgassing, coefficient-of-thermal-expansion (CTE) mismatch, and debris contamination—issues unacceptable for aerospace, satellites, and cleanroom assembly.

Vacuum induction welding offers a promising pathway: by heating only the polymer at the interface, it drastically reduces thermal exposure, enabling rapid, localized bonding of high-value thermoplastic composites.

2. Literature Review

2.1 Sandwich structures and joining challenges

Previous literature emphasizes the mechanical superiority of sandwich structures, but consistently identifies the interface as the weakest link. Studies on thermoplastic cores such as PEI, PPS, and PEEK confirm that conventional welding requires high pressures and temperatures, often exceeding safe thresholds for lightweight cores.

2.2 Thermal bonding, resistance welding, and induction heating

Thermoplastic welding strategies fall broadly into:

1. Through-thickness heating (hot press / oven)
   • Risks heating and degrading the entire panel.
2. Resistance welding
   • Uses conductive meshes; scaling and uniformity are problematic.
3. Induction welding
   • Heat generated only at the interface.
   • Efficient for CF/PEEK skins but complex for sandwich cores.

2.3 Hysteresis-loss susceptors

Recent publications highlight the use of ferromagnetic particles (Ni, Fe, magnetite) dispersed in a polymer film as an internal susceptor:

• Eliminates metal meshes.
• Lower filler fraction avoids mechanical weakening.
• Curie temperature provides inherent overheat protection.

Ni is often preferred because its Curie temperature (~358 °C) is close to PEEK’s melt point (~343 °C), ensuring self-regulation.

2.4 Dual-polymer bonding

Introduced in aerospace welding research, the dual-polymer concept places a lower-melting-temperature polymer (e.g., PEI) at the interface of a higher-melting matrix (e.g., PEEK). This allows joining at reduced temperatures and protects the laminate from deconsolidation.

2.5 Vacuum induction welding

Only a small number of studies have explored combining vacuum pressure with localized induction heating. The approach promises:

• Uniform pressure via atmospheric load.
• Heat confined to microscopic regions at the joint.
• Simplicity of processing (bagging + coil traversal).
• Reduced skin and core damage.

3. Methodology

3.1 Materials

3.1.1 Honeycomb core

• Material: ULTEM™ 1010 (PEI).
• Process: High-temperature FFF 3D printing.
• Geometry:
  • Hexagon width: 4 mm
  • Wall thickness: 0.8 mm
  • Height: 10 mm
  • Outer dimension: 50×50 mm
• Key detail: Printed directly onto a 1-mm PEI base sheet to ensure monolithic bottom facesheet.

3.1.2 Facesheet laminate

• Material: Glass-fiber reinforced PEEK (GF/PEEK).
• Surface layer: 125-µm extruded PEI film (ULTEM 1000).
• Reason for glass reinforcement: Avoid unintended induction heating of carbon fibers.

3.1.3 Ni/PEI hysteresis susceptor

• Matrix: PEI.
• Filler: 10 vol% Ni powder (5 μm particles).
• Thickness: 0.6 mm.
• Rationale: Ni offers the highest hysteresis heating rate under available frequency and provides Curie-limited thermal protection.

3.2 Vacuum induction welding setup

3.2.1 Induction system

• Equipment: Ambrell EASYHeat 10 kW.
• Frequency: ~389 kHz.
• Coil current: 600 A.
• Standoff: 3 mm.
• Magnetic concentrator: FluxTron 559H.

3.2.2 Vacuum bagging

• Film: High-temperature Kapton.
• Vacuum pressure: –90 kPa (≈0.1 MPa applied pressure).

3.2.3 Variable parameter: Welding speed

Speeds tested:

• 0.3, 0.4, 0.5, 0.7, 0.9, 1.0 mm/s

Slower speeds ↔ longer thermal exposure → higher interface temperature.

3.3 Characterization

3.3.1 Microscopy

• Cross-sections cut according to ASTM sandwich analysis standards.
• Observables:
  • Polymer softening
  • Susceptor squeeze-flow
  • Wall deformation
  • Contact width

3.3.2 Mechanical testing

Flatwise tensile (FWT), ASTM C297:

• Load rate: 0.5 mm/min.
• Area: 50×50 mm.
• Failure classification: adhesive, cohesive, core failure, hybrid.

4. Results

4.1 Overheating regime (0.3–0.4 mm/s)

• Core walls exhibit mushrooming and melting.
• Susceptor expands and bubbles due to trapped gases.
• Excessive local heating exceeds polymer stability window.

4.2 Insufficient heating (1.0 mm/s)

• No true weld formed.
• Only superficial tack before cooling.
• Insufficient time-at-temperature for PEI chain diffusion.

4.3 Optimal welding window (0.5–0.9 mm/s)

4.3.1 Microstructural development

• 0.5 mm/s:
  • Large susceptor deformation and wide bondline.
  • High degree of polymer fusion.
• 0.7 mm/s:
  • Moderate softening; consistent bonding.
• 0.9 mm/s:
  • Narrow contact zone; reduced flow.

4.3.2 Effective interfacial contact

Measured contact width increased from:

• 0.8 mm → 2.4 mm (slow speed)
  → Contact area increased up to ~60% of cell perimeter.

4.3.3 Mechanical performance

Flatwise tensile strength:

Welding Speed (mm/s)	FWT Strength (MPa)
0.5	≈6 MPa
0.7	≈4–5 MPa
0.9	≈3–4 MPa

4.3.4 Failure modes

• 0.5 mm/s: Cohesive failure within Ni/PEI susceptor (strongest bond).
• 0.7 mm/s: Mixed adhesive failure.
• 0.9 mm/s: Adhesive failure at core side.

This indicates strengthening of the intrinsic bond quality at lower speeds.

5. Discussion

5.1 Process–Structure–Property Link

Process: welding speed
→ Structure: susceptor flattening + polymer interdiffusion
→ Property: tensile strength + failure mode

Key finding:
Stronger welds arise not only from increased contact area but from qualitatively superior interfacial molecular diffusion.

5.2 Why 0.5 mm/s is the optimum

At this speed:

• The interface temperature approaches PEI melt range.
• PEEK skin remains intact (below 343 °C).
• Core retains structural integrity.
• Susceptor softens enough to flow into intimate contact.
• Induced polymer diffusion creates a cohesive bond stronger than the susceptor itself.

5.3 Industrial implications

Aerospace and Space Structures

• Low-outgassing thermoplastics + no adhesive = cleaner manufacturing.
• Local heating minimizes risk for thin skins or fragile cores.

Automotive and EV Battery Enclosures

• Rapid cycle time fits into inline processes.
• Weld-on-demand approach enables modular panel assembly.

Machinery Manufacturers (e.g., your hot-melt UD tape line)

• Vac-IW can be integrated after UD tape consolidation.
• Co-extruded PEI surface layers on CF/PEEK UD tapes enable welding.
• A robotic induction head could follow curved surfaces.

6. Conclusion

This reconstructed study demonstrates that vacuum-assisted induction welding is a viable, high-performance joining method for thermoplastic sandwich structures. By confining heat directly to the bondline, using Ni-based hysteresis susceptors, and applying uniform atmospheric pressure, the process achieves:

• High mechanical strength (up to 6 MPa)
• Zero adhesive usage
• Minimal thermal damage
• Predictable and tunable weld quality

The findings clearly establish the process window (0.5–0.9 mm/s) and reveal the microstructural mechanisms governing weld formation. Vacuum induction welding stands as a scalable, clean, and aerospace-compatible technology for joining next-generation thermoplastic composite assemblies.

7. References

(Reconstructed referencing based on your provided material)

1. Martin, R.G., Johansson, C., Tavares, J.R., Dubé, M. Manufacturing of thermoplastic composite sandwich panels using induction welding under vacuum. Composites Part A, 2024.
2. Grünewald, M. Thermoplastic sandwich core joining methods, 2017.
3. Mitschang, P. Induction welding principles for thermoplastic composites, 2002.
4. Worrall, C. Electromagnetic heating of composite laminates, 2014.

1. Kropka, M., Mühlbacher, M., Neumeyer, C., Altstädt, V. (2017). From UD-Tape to Final Part – A Comprehensive Approach Towards Thermoplastic Composites, Procedia CIRP, 66, 87–92.
2. Advani, S., & Sozer, E. (2002). Process Modeling in Composites Manufacturing. Marcel Dekker.
3. Gardiner, G. (2020). “Thermoplastic Composites in Mass Production.” CompositesWorld Magazine.
4. Ageorges, C., Ye, L. (2002). “Fusion Bonding of Polymer Composites.” Polymer Engineering & Science.
5. Friedrich, K. (2018). “Thermoplastic Composite Processing Technologies.” Encyclopedia of Materials: Composites.