Abstract:
This study aims to develop tools for the design, optimisation and control of automated
tape placement (ATP) that integrates flashlamp heating. A thermo-optical simulation of
ATP is developed, combining a 2D finite element model of the heat conduction with 3D
ray tracing analysis. The methodology is validated against measurements acquired during
ATP trials of AS4 carbon/PEEK composites, presenting deviations up to 20°C. Flashlamp
operation at low frequency and long pulses (25 Hz /4.75 ms) results in up to 150°C higher
irradiation temperatures and increased thermal penetration depth, 100 over 50 μm,
compared to high frequency and short pulses (100 Hz/1.1 ms) or continuous operation.
Consolidation temperatures under the roller are identical for pulsing scenarios of
equivalent average power, including continuous operation. To increase computational
efficiency, an 1D simulation of ATP is put forward comprising distinct models
representing the tow, deposited material, and consolidated stack with transfer of
temperature information to ensure field continuity. The 1D solution requires only 1-2%
of the computational effort of the 2D model with a minor trade-off in accuracy, up to
14°C. Based on the efficient 1D solution, an optimisation scheme of ATP is developed
by integrating models of material degradation, interfacial bonding and a Genetic
Algorithm. The optimisation scheme identifies the Pareto front of the multi-objective
problem accurately in 25% of the computational effort required for an exhaustive search.
Strong trade-offs exist between bonding, thermal degradation and productivity, limiting
the average bonding value in the stack to 0.35 before matrix degradation exceeds the 1%
threshold typically set for aerospace applications. A monitoring strategy for ATP of
thermoplastic prepregs is proposed combining 1D analytical solutions and temperature
data acquired on the tool side of the deposited material, allowing estimation of nip point
temperatures in real time. The method integrates an inverse solution to determine the
heater power input from temperature data and enhance the nip point estimation accuracy.
The monitoring scheme presents good accuracy for a wide range of velocities, substrate
thickness and tooling materials, with an average error of 15°C even in the presence of
significant measurement noise.
