Combustion-related research in Purdue's School of Mechanical Engineering

Professor John Abraham's research concentrates on probing the physics of fuel sprays, droplets and their combustion through applications of computational techniques. In ongoing work, the improvements made in the understanding of the physics are extended to improve the designs of internal combustion engines, fuel reformers and after-treatment devices. In addition to employing traditional finite-volume and finite-difference methods, Professor Abraham and his students also focus on the development and application of emerging computational techniques such as lattice-Boltzmann methods, dissipative particle dynamics, molecular dynamics and combinations of macro, meso and micro methods in multi-scale modeling. These tools are being employed to investigate the physics of liquid jet-breakup, drop-drop interactions, drop impingement on walls, and nano jets. Fuel-air mixing and combustion in turbulent jets, ignition, flame extinction, flame lift-off and flame-wall interactions are also investigated by employing a combination of numerical techniques that includes large-eddy and direct numerical simulations. Reduced and detailed chemical kinetic mechanisms are employed in the combustion studies.



Liquid-jet breakup




Mixture fraction in pulsed injection




Lifted Diesel flames
 

Professor Steve Frankel and his group are active at the forefront of computational combustion modeling.  They are conducting large eddy simulations (LES) of non-premixed turbulent jet flames using the filtered mass density function approach that employs Lagrangian Monte Carlo particle methods to solve the sub-grid-scale joint scalar probability density function.  Both idealized and realistic chemistry, the later being computationally realized through the use of the in situ adaptive tabulation technique and large scale parallel computing, are being pursued to study the dynamics of flame extinction and re-ignition.  Emphasis is being placed on accounting for the effects of length scale and chemistry on the sub-grid-scale mixing frequency.  Validations are being pursued through comparison to detailed measurements of velocity and species statistics in laboratory scale turbulent jet flames.

Professor Paul E. Sojka’s combustion related research is focused on the formation and evolution of super-critical sprays (jets).  That work is motivated by recent advancements in gas turbine technologies that have made flight speeds of up to Mach 7 possible.  At these flight speeds the ambient air surrounding the engine is no longer a sufficient heat sink so an additional source is needed to remove the extra heat generated due to the viscous effects of the high speed air interacting with engine blades and vanes, as well as with airframe control surfaces and the myriad electronics that are part of a typical flight system.  The on-board fuel is being explored as a possible candidate to serve as a heat sink.

By adding heat to the on-board fuel is it quite possible that the temperature and pressure will rise above the critical point.  As a result, the fuel nozzle must be designed to accommodate supercritical fuel (SCF) injection.  Work is progressing along two lines.

In the first case, a unique high-pressure experimental facility was used to determine how ambient pressure influences the spreading angle of swirling liquid jets.  The most recent experimental results extend the pioneering work of DeCorso and Kemeny (1957), as shown in the figure below.  Key conclusions drawn from this data are:

  • A reduction in cone angle with increased ambient pressures was as ambient pressure reached 7.1 MPa.
  • At all levels of Pamb tested, the cone angle was reported to be significantly less than the inviscid pressure swirl prediction.
  • At ambient pressures below ~4.0 MPa the spray width was found to be approximately proportional to Pamb-0.48 (Eqn 5.1); above this pressure, the cone angle was observed to remain constant within the experimental uncertainty.

The practical implication of this work is that designers of high-Mach gas turbine engines can now estimate the fuel distribution produced by swirling injectors.

 

 

In the second case, schlieren photography (figure below left) was used to map concentration fields in super-critical jets produced by swirling flows (figure below right).  Carbon dioxide was used as a surrogate to typical aviation fuels to remove the hazard of explosion or fire.  Typical experimental data are provided in the bottom figure.

  



The following conclusions were drawn from this, and other, data:

  • Initial and final spreading angles were for each injector were found to be insensitive to injectant/environment density ratio or mass flowrate;
  • As expected, spreading angles increased with injector swirl number;
  • All jets demonstrated self-similar mean concentration profiles;
  • The distribution of mass widened significantly with an increase in swirl number.
  • The spread of mass in the jets (width of concentration profiles at half the maximum concentration) was found to vary linearly with axial distance.

Our current efforts focus on methods to prevent vortex breakdown, and the use of actual aerospace propellants at realistic temperatures and pressures.