Numerical investigation of direct ammonia injection and combustion

A numerical investigation of ammonia direct injection is presented in this work. Injection is initially conducted through a single-hole ECN-like nozzle into a constant volume chamber. The performance of break-up and turbulence models is assessed and the mesh is validated in non-reacting conditions against experimental data. The results, in terms of spray tip penetration and liquid length, show good agreement with experiments, although an overprediction is observed in the initial transient phase of the spray tip penetration. For the reacting case, a flamelet tabulated manifold (FGM) model is used. 1D flamelet studies and 3D Spray D-like studies, provide the initial conditions favoring the auto-ignition of ammonia. All cases analyzed, ranging in temperature from 1200 to 1300 K and in pressure from 79 to 113 bar, show ignition, in very different ranges. However, only the marine-engine derived case (“Case 3”) is providing a stable lifted flame. Therefore, Case 3 is chosen to study a marine-like combustion chamber (SCC), in which ammonia is injected via a multi-hole injector. To replicate the conditions in the real application, the pilot diesel injection is treated in a simplified manner, with only the temperature and pressure rise taken into account. The conditions studied refer to two positions of the pilot injection with respect to the main ammonia injection. The results are encouraging, as both solutions show ignition. The second, considering a pilot very close to the ammonia injector, also shows a stable lifted flame for ≈8 ms. However, it should be noted that no experimental data are currently available for any of the reacting conditions of both the Spray D-like and marine-like cases. The study is therefore purely exploratory, mostly providing best practices and approaches, and suggesting the conditions that can be tested in real hardware in the future.