Definition of scaling strategy and isothermal scaling of a lean-burn combustion system to different air mass flows using CFD

Abstract

The demand for shorter development times is becoming increasingly important in today's industry. The adaptation or scaling of an existing engine to different mass flow rates would permit to deal with the shorter development time requirements. This scaling method applied to a combustion system, consists in adapting a new configuration in order to obtain the same flow field characteristics as in the reference configuration. The development of the lean-burn combustion systems is still in its initial phase. Thus, this technology would especially benefit from the corresponding scaling procedure. For this reason, an automatic procedure for scaling of a lean-burn burner-combustion chamber configuration to a modified mass flow rate is developed in this work. The burner geometry is adapted in order to comply with scaling goals. The burner parameters that have a significant influence on the flow field are identified. An automatic process chain is developed with which the effect of the variation of the different burner geometrical parameters on the flow field is investigated. It comprises several processes that are performed sequentially in an automatized way and can be divided into three major blocks: geometry and mesh, simulation and postprocessing. The burner-combustion chamber configuration is scaled under the condition of self-similar flow field. Different criteria to judge about flow field similarity are investigated. Finally, the profiles of circumferentially averaged (CA) velocities are established as a suitable criterion and constitute the objective function for the numerical scaling. Isothermal scaling is performed with different definitions of objective functions based on RANS simulations. When values and positions of characteristic points of CA velocity profiles are included in the objective function, large deviations in the tangential velocity with respect to BK160 are observed. The reason is that the region with large deviation is not included in the objective function and the process causing that behaviour of the curve cannot be represented by the points included in the objective function. When equally distributed points are considered to represent the axial, radial and tangential profiles, the tangential velocity profiles are adjusted to BK160 but with a deviation between the axial and radial profiles. When only the tangential profile is included in the objective function with equally distributed points, the tangential velocity curve accurately represents the BK160 and the axial and radial velocity curves are fully reproduced. For this test case, adjusting the tangential behaviour is important for reproducing the global combustion chamber flow field.