J. Eur. Opt. Society-Rapid Publ. 22, 41( 2026) 403
Figure 1.( a) NAPUBEST IR-oven with sample placed at the bottom of the chamber,( b) same arrangement in the simulation model of NAPUBEST IR-oven including the sample placed at the bottom of the chamber,( c) black coated sample. all sides of the sample and is planned to be implemented as anextstep.
For the verification, the CAD model of the specimen used in industrial fabrication and in the NAPUBEST oven was imported into FRED. The full surface of the specimen was used as the detector surface. The absorbed irradiance at the surface was simulated and the resulting irradiance map was exported as a CSV-file for post processing.
The optical properties of all materials in the heating chamber were either taken from vendor data( fused silica transmittance and refractive index, reflectivity of emitter mount and gold coating of emitter backside) or measured using an Ulbricht Sphere for all other materials( see Fig. 2). The black and white coating were found to be perfectly Lambertian: the HDR and DDR measurements had the same result, thus there is no specular fraction. For Alanod the two quantities have quite different values, resulting in a high specular fraction of the reflectivity, which is expected.
Measurements and vendor data were used to create material, coating and scatter data [ 5 ] in FRED for the peak wavelength of the emitters only. This simplification largely reduced the simulation time.
As a first test of the equipment and the correspondence between experiment and simulation, a black and a white coated sample were heated for 2 min each in the NAPUB- EST heating chamber. The sample was placed into the oven chamber as shown in Figure 1 and the chamber was closed by the Alanod-covered front lid. All four emitters were running with full power for 2 min before the lid was re-opened and the temperature distribution was captured by an infrared camera 0, 1 and 2 min later( Testo 883, 640 480 pixel). The corresponding simulation takes into account four emitters at full power and the respective BSDF of the coatings.
Results
The temperatures of the specimen after 2 min of infrared irradiation were measured with the Testo infrared camera. Especial care had to be taken to not measure heat reflections of the lamps or of hot spots which get reflected on colder parts of the sample. The reflecting sheets of the heating chamber create a mirror image of the sample with the advantage of making the down facing surfaces of the sample visible.
The simulation results were displayed in an in-house developed software allowing the readout of surface irradiance point or linewise. The results for maximum and minimum temperature as well as the corresponding irradiance are shown in Table 1. Regarding ratio and over all distribution, the measurement and simulation are in good accordance( see Fig. 3). As expected, the temperature of the black sample was much higher than that of the white sample after the same dwell time in the oven due to the higher infrared absorption.
The temperature distribution though showed differences compared to the irradiance distribution( see Fig. 4) thatwe attribute to heat flow and heat transfer in the system, e. g. at the point of support, the temperature is lower than the mean surface temperature while the irradiance is more or less homogenously distributed over the side surface of the sample.
Discussion
Earlier simulations with a simple T-shaped sample [ 6 ] showed that inhomogeneities in irradiance at the sample surface can be prevented by illuminating the sample from different positions. It allowed to reduce the max / min ratio of the irradiance from 4.2 to 1.9 for the T-shaped sample. In the real world this means to turn the sample in the oven and / or install multiple heating modules in the oven.
The optical parameters of cured powder coating can be measured easily with an Ulbricht sphere and converted into simplified scatter models. The results of this investigation show that this is sufficient to map the simulations to the measurements. Heating results strongly depend on the coating colour and sample geometry as expected. The thermal effects like thermal conduction exceed the expected magnitude and need to be taken into account.