Heat reaches the food in an oven by three routes, and each has a different distribution:
Radiation. From the elements and the hot walls. It depends on view factor: surfaces that face an element directly receive far more. That is why the edge of a tray under the top element burns.
Convection. Dominant in fan ovens; regions of higher air velocity cook faster.
Conduction. Through the tray or mould into the base.
Even baking means the sum of the three being balanced over the food surface — and the three peak in different places.
Why a fan oven is more even
The fan circulates the hot air, making the convective component dominant and more uniform, and relatively reducing the effect of radiation's directional dependence. That is why fan mode is preferred for multi-shelf baking.
But fan mode has its own distribution problem: regions near the fan outlet keep a higher velocity. Air guides and perforated panels exist to balance that and are directly an analysis subject — a classic distribution problem.
Humidity should not be ignored
Vapour leaving the food raises the relative humidity inside the oven. Humidity delays crust formation and changes the heat transfer. In steam ovens this is used deliberately. A model that does not solve moisture transport cannot predict crust and colour formation. See the drying note.
Modelling
Radiation is mandatory. View-factor based surface-to-surface radiation is required; neglect it and the distribution comes out entirely wrong [1]. See the radiation note.
Surface emissivity governs. There is a factor between a bright aluminium tray and a dark enamel one. That also explains the difference users experience.
The food is a thermal mass, not a boundary condition. Because the question is the temperature at the centre of the food over time, the problem is conjugate and transient.
References
- Modest, M. F. & Mazumder, S. (2021). Radiative Heat Transfer, 4th ed. Academic Press. ISBN 978-0-323-98406-5.
- Ranz, W. E. & Marshall, W. R. (1952). Evaporation from drops. Chemical Engineering Progress, 48, 141–146 (Part I) and 173–180 (Part II).