The climate crisis has compelled the scientific community to explore new methods of energy generation, storage, and transportation. In addition to the energy carrier, consideration must also be given to how it is converted into usable energy. For hydrogen (H 2 ) and ammonia (NH 3 ), fuel cells are potentially the overall efficient technology for converting these gases into electricity. Among these, hydrogen solid oxide fuel cells (SOFCs) demonstrate high efficiency, although they operate at elevated temperatures. In direct ammonia solid oxide fuel cells (DA-SOFCs), ammonia can be fed directly into the cell, as the anode catalyses the decomposition reaction: 2NH 3 → 3H 2 + N 2 . This is typically done using common nickel-based materials such as Ni-YSZ and Ni-GDC. While experimental methods are the conventional means to analyse the performance of a cell, they often present challenges that are difficult to overcome, mostly due to budget and/or time constraints. As a result, computational simulation has emerged as a practical alternative for understanding the physical processes involved in SOFC operation. In this study, the material properties serve as adjustable parameters to represent different materials, allowing for the evaluation of physical fields anywhere within the system. In this work we utilise a DA-SOFC to measure the effect of temperature and fuel composition on power output through computational and experimental methods. The cell consists of a 31 cm tall vertical, 2.55 mm thick alumina tube with a 14.4 mm diameter. Tangent to its interior left wall there is an alumina fuel inlet, 29.96 cm tall, 0.6 mm thick and a diameter of 2.6 mm. At the top of the large tube lies a shorter, 2.8 mm tall alumina tube of the same diameter and thickness. Inside its opening a 200 μm thick LSGM electrolyte disc acts as a sealed lid. There is a 50 μm thick, 1 cm by 1 cm porous square Ni-GDC anode at the center of this disc, on its lower face. Similarly, above it there is a porous square cathode of the same dimensions made of SrCo 0.95 Re 0.05 O 3-δ . This ensemble is placed inside a cylindrical furnace 29 cm tall with a diameter of 22 cm so that the electrolyte remains at the centre. The top and bottom of the furnace are exposed to the atmosphere. The experimental setup can be seen in Figure 1 (A). For a visual representation of the cell (not to scale) see Figure 1 (B). Simulations are implemented in the software COMSOL Multiphysics ® . The stationary model consists of fluid dynamics and species transport in both free flow and porous regions; ionic and electronic current in the active cell phase; heat transfer in the entire domain, and finally electrochemical and ammonia cracking reactions. Boundary conditions are used to prescribe furnace temperature, operational voltage and fuel composition. Physical parameters are measured in-house when possible, otherwise they are obtained from literature or estimated. Equations are coupled and solved with the finite element method using a multi-step procedure. The model is summarized in Figure 1 (C). For a given furnace temperature, a room-temperature mixture of hydrogen and ammonia is supplied through the inlet. A voltage difference is applied between the anode and the cathode to generate polarization and power curves in the range 0.22 to 1.12 V. In the case of the simulation, a step of 0.1 V is used; in this way, a simulated polarization curve has 10 points. This procedure is repeated for furnace temperatures of 650 °C, 700 °C and 750 °C and inlet ammonia fractions of 0, 0.5, 1 (pure hydrogen, 50% ammonia and pure ammonia). Given that electrochemical reactions consume hydrogen only, it is expected that pure hydrogen fuel has better performance. Preliminary results for pure H 2 and pure NH 3 shown in Figure 1 (D) demonstrate that H 2 provides greater power output, as hypothesized. Power peaks using NH 3 are approximately half of those using H 2 . The simulated curves are comparable to experimental results without resorting to prior curve fitting. It can also be observed in Figure 1 (E) that average anode temperatures are lower than furnace temperatures, and this phenomenon is more apparent when pure NH 3 is used as fuel. The loss of efficiency when transitioning from H 2 to NH 3 as fuel can be attributed to the endothermic nature of the reaction 2NH 3 → 3H 2 + N 2 , which reduces the temperature in the electrochemically active region which in turn negatively affects the electronic and ionic conductivity of components. Finally, the opportunities for optimum functioning of the cell are further explored computationally considering various operating conditions and potential new materials. Figure 1: (A) Experimental setup, (B) Model geometry (not to scale), (C) Scheme of multiphysics coupling, (D) Experimental and computational power curves for 750 °C, (E) Average anode temperature in relation of furnace temperature. Figure 1