Aluminum (Al) powder has been proposed as a promising dense energy carrier to store and transport abundant renewable power for a future carbon-neutral society. In this work, a detailed model for simulating Al particle cloud combustion with multiple oxidizers in the context of large eddy simulation is developed. The Al particles are tracked in the lagrangian framework with various sub-models including phase change, heterogeneous surface reaction, evaporation and radiation. The model is then employed to simulate turbulent jet flames with Al particles (d(32) = 25 mu m) in a hot co-flow. Tunable diode laser absorption spectroscopy (TDLAS) is used to measure the temperature and H2O distribution. The present model is first validated by comparing the predicted condensed Al2O3 distribution with the measured Mie scattering signals. The results indicate that the simulation results are in good agreement with the experimental measurements. Further comparisons of the temperature and H2O profile are then conducted. The numerical simulation predicts weaker mass and heat transport in the upstream and middle of the flame along the centerline of the burner, compared with the reconstructed TDLAS results, but the downstream scalar profiles of the jet flame agree well with the measurements. Discrete characteristics are observed in the simulated instantaneous temperature and condensed Al2O3 snapshots, which are due to the nature of heterogeneous combustion of metal fuel particles under fuel-lean conditions. Finally, the influence of the oxidizer mole fraction and the initial temperature of the primary jet on the ignition distance is analyzed. The results demonstrate that the ignition distance can be decreased effectively by increasing the oxidizer concentration of the hot co-flow or the initial temperature of the Al particles and carrier gas.