Investigation of the interaction between injected water droplets and flames

In addition to basic research, droplet-flame interaction is also an interesting topic for a variety of industrial and practical applications. In gas turbines, for example, temperature peaks and thus the formation of thermal nitrogen oxides can be significantly reduced by targeted water injection. Also in piston engines (gasoline, diesel), water injection is used in various designs to specifically influence combustion. In addition to reducing pollutant emissions, this also helps to increase performance. Intake air cooled by water droplets requires less specific compression work, which enables a higher air flow rate. Thus, a greater fuel throughput can be achieved with the same amount of air, which in turn increases power [Steinhilber, 2007].

However, the real impetus for the work arose from fire and explosion protection. Both encounter the same problem in principle, but the time scales in explosion protection are dramatically smaller, which is why a distinction should be made here. The main extinguishing effect is the cooling effect of water, which has a high specific heat capacity in the liquid phase and an even higher enthalpy of vaporization. The mass of extinguishing water limits the maximum possible heat absorption, while the evaporation rate limits the real amount of heat that can be absorbed. The latter depends to a large extent on the total droplet surface area, which increases with decreasing droplet size in relation to the total mass. Finer droplets can therefore lead to a more efficient extinguishing effect, but they are also more susceptible to flow conditions and sometimes do not reach the desired location of most effective heat absorption, namely the flame front near the reaction core. While the cooling effect is fairly well understood in the overall balance, more detailed knowledge at smaller laboratory scales is still lacking (see for example [Sasongko et al., 2011]). What other effects are important for flame quenching? How can these be described or predicted? Can such results be transferred to the real scale?

Schematic of burner nozzle with isothermal flow field from Merkle (2006).

To this end, the droplet-flame interaction is to be investigated experimentally on a laboratory scale. In order to reproduce a highly turbulent diffusion flame as close to reality as possible, a tube combustion chamber fired with natural gas and equipped with the institute's own burner nozzle [Merkle, 2006] is used (see Figure 1). This also provides a semi-stationary flame front to study the effects of water droplets. The droplets are expected to be in the microscale range between 10 and 100 µm and to reach the flame via a polydisperse spray from different nozzle positions. Thus, unlike previous studies, the combustion air is not preloaded with water [Sasongko et al., 2011]. The experiments are to be supported by a detailed literature study as well as by basic numerical investigations of a validation nature. A quantitative description of the interaction effects between water droplets and flame directly at their front as well as in dependence of the respective spray and droplet characteristics is expected. Therefore, an extensive spray investigation should be carried out in advance, so that this can be varied accordingly in the later course. Non-intrusive optical measurement methods such as LDA/PDA and shadowgraphy are conceivable for this purpose. A thermo- and velocimetric measurement of the relevant areas (TE, hot wire) and an exhaust gas investigation support the flame stability investigation in a meaningful way. Optimally, the results would finally be transferable to a real scale.


[1] T. W. Steinhilber, Influence of water or emulsion injection on homogeneous diesel combustion, Technical University of Munich, Chair of Thermodynamics, 2007.

[2] M. N. Sasongko, M. Mikami, and A. Dvorjetski, "Extinction condition of counterflow diffusion flame with polydisperse water sprays," Proceedings of the Combustion Institute, vol. 33, no. 2, pp. 2555-2562, 2011.

[3] K. Merkle, Influence of co-rotating and counter-rotating combustion air streams on the stabilization of turbulent double-twist diffusion flames, Universität Karlsruhe (TH), Faculty of Chemical Engineering and Process Engineering, 2006.



Experimental study on the effects of sprinkler activation on the stratified smoke layer inside a 36 m² enclosure
Vetter, M.; Dinkov, I.; Schelb, D.; Trimis, D.
2021. Fire safety journal, 121, Article no: 103313. doi:10.1016/j.firesaf.2021.103313
Experimental and numerical study on the interaction between sprinkler water spray, fire plume and smoke layer
Dinkov, I.; Vetter, M.; Schelb, D.; Trimis, D.
2019. Proceedings of the European Combustion Meeting – 2019, April 14-17, Lisboa, Portugal, S2_R1_89