TUCer / Diploma thesis · Technical University of Crete
Composite Curing Oven.
A purpose-built oven to produce the CFRP wheel's carbon-fibre components — thermal design modelled and sized in Simulink before the first panel was cut.
FIG. 01 — A rim curing under vacuum bagging inside the finished oven
Requirements & thermal model
The brief: hold up to 120°C, ramp at up to 3°C/min, and stay within ±1°C once stable — tight enough to cure the wheel's resin reliably without over- or under-curing any part of it.
Before building anything, I modelled the oven as a physical network in Simulink/Simscape's Foundation Library thermal domain, not an abstract lumped circuit: five thermal masses named for the real materials in the build — Coil, Air, Insulation Board, Rock Wool, and OSB — linked by three conduction paths and two convection paths, with an ambient temperature source and a controlled heat-flow source driving the network. The loop simulated is the loop that got built: commanded cure profile → smoothing → PID → PWM → the plant, with measured temperature fed back to close it.
A Thermal Power block fixed at 1600 W, integrated to kWh, turned the model into a sizing tool for insulation and installed power rather than a simple temperature check. Different insulation materials and thicknesses were traded off against that power budget in simulation before the first panel was cut — landing on 1.6 kW installed (2×800 W elements) against about 300 W average draw for a typical cure profile, for ramp margin and headroom to scale up later.
Build
The chamber is 940×970×600 mm (0.55 m³ working volume): an 18 mm OSB shell with seams sealed in aluminum tape, lined with 30 mm of chimney-grade rockwool insulation. Two 1Ω, 800W DC resistance elements provide heat, with forced-air circulation and two K-type thermocouples (air and wall) feeding a PID loop that drives a MOSFET power stage via PWM.
In production use
Simulated performance met spec: both temperature plateaus (80°C and 120°C) held with small steady-state error, and the target ramp rate was achievable within the installed power budget.
In real use, the oven ran every one of the wheel's cure cycles — 16 hours at 60°C per the resin manufacturer's spec — and the molds it cured against showed no deformation or wear after many thermal cycles. Measured thermocouple logs to pair directly against the simulated profile are still on the list to capture.
Every part it cured went on to the instrumented test bench for physical validation against the wheel's FEA predictions.