Till the time when Sujith's work became known, combustion instabilities were studied in the conventional manner, where stability of individual eigenmodes has been taken to imply stability of the system. However, this approach is inadequate for systems governed by non-normal operators which have non-orthogonal eigenvectors. Sujith discovered that thermoacoustic interactions are governed by a non-normal operator and has worked out in a series of break-through papers the possible implications for the stability of combustion systems. He was the first to provide a convincing mathematical and physical explanation for subcritical transition to instability (known in thermoacoustics as triggering) from small but finite amplitude initial conditions. Strong transient growth in disturbance energy due to non-normality, when combined with nonlinear effects, can destabilize a system which is predicted to be stable by classical linear stability theory.

Sujith demonstrated that non-normality can lead to the failure of the traditional controllers that were designed on the basis of classical linear stability analysis. He pioneered the establishment of mathematical machinery to analyse transient growth and its consequences in thermoacoustic systems. He introduced the analysis of the non-normal nature of thermoacoustic systems using pseudospectra. He showed that singular values (and not eigenvalues) characterise energy growth during thermoacoustic instabilities. He established that the internal dynamics of the flame plays a significant role in thermoacoustic evolution. Sujith's recent papers on unified theory for thermoacoustic systems have made a significant impact in the understanding of the fundamental coupling mechanisms. Thus, Sujith's work has led to a paradigm shift in the analysis of thermoacoustic instability. The thermoacoustic community has now adopted this approach of study.

Sujith is also credited with the establishment of routes to chaos in thermoacoustic systems. Traditionally, it is believed that when the acoustic driving provided by the flame is balanced by the losses in the system, one attains limit cycle oscillations. In recent years there is an interest in predicting the amplitudes of the limit cycle oscillations. Our recent findings indicate that limit cycle is just one of the possible end states of the system. A thermoacoustic system can undergo further bifurcations and attain states such as quasi-periodic, period doubling, intermittency, frequency locked and chaotic states as observed in both experiments and computations.

Traditional analysis and modeling of the low amplitude combustion noise as well as its transition to high amplitude combustion instability often neglects or averages out the unsteady irregular fluctuations observed in the measured data or treat them as a stochastic background. Sujith demonstrates that a detailed analysis of the irregular fluctuations can provide information that is of both diagnostic as well as prognostic value. Combustion noise was identified to be the result of chaotic dynamics of the global system comprising turbulence, combustion and the chamber acoustics. Further, he shows that these chaotic fluctuations display scale invariance and multifractality, which provides evidence for the multiple scales involved in the energy transfer. Sujith has developed several representative measures that can act as early warning signals to impending instability in practical gas turbine combustors. He successfully tested these measures on laboratory scale combustors, identifying the proximity of the operating conditions to the onset of instability, well before the combustor encounters them.

Sujith contributed to the understanding of fundamental processes that control acoustically excited jets and sprays, swirl flow from vane swirlers and the interactions between spray, swirl flow and acoustic oscillations using optical diagnostics. He has conducted experimental and theoretical investigations and has obtained elegant analytical solutions for acoustic and shock wave propagation in inhomogeneous media.