Papers with impact factor

1. Hughes KJ, Tomlin AS, Hampartsoumian E, Nimmo W, Zsely IG, Ujvari M, Turanyi T
An investigation of important gas-phase reactions of nitrogenous species from the simulation of experimental measurements in combustion systems
Combustion and Flame 124 (4): 573-589 MAR 2001

Simulated results from a detailed elementary reaction mechanism for nitrogen-containing species in flames consisting of hydrogen, C1 or C2 fuels are presented, and compared with bulk experimental measurements of nitrogen-containing species in a variety of combustion systems including flow reactors, perfectly stirred reactors, and low pressure laminar flames. Sensitivity analysis has been employed to highlight the important reactions of nitrogenous species in each system. The rate coefficients for these reactions have been compared against the expressions used in three other recent reaction mechanisms: version 3.0 of the GRI mechanism, the mechanism of Glarborg, Miller and co-workers, and that of Dean and Bozzelli. Such comparisons indicate that there are still large discrepancies in the reaction mechanisms used to describe nitrogen chemistry in combustion systems. Reactions for which further measurements and evaluations are required are identified and the differences between the major mechanisms available are clearly demonstrated.

2. Zsély I, Zádor J, Turányi T
Similarity of Sensitivity Functions of Reaction Kinetic Models.
Journal of Physical Chemistry 107: 2216-2238 (2003).

Local sensitivity functions @Yi/@pk of many chemical kinetic models exhibit three types of similarity: (i) Local similarity: ratio Ďij ) (@Yi/@pk)/(@Yj/@pk) is the same for any parameter k. (ii) The scaling relation: ratio Ďij is equal to (dYi/dz)/(dYj/dz). (iii) Global similarity: ratio (@Yi/@pk)/(@Yi/@pm) is constant in a range of the independent variable z. It is shown that the existence of low-dimensional slow manifolds in chemical kinetic systems may cause local similarity. The scaling relation is present if the dynamics of the system is controlled by a one-dimensional slow manifold. The rank of the local sensitivity matrix is less than or equal to the dimension of the slow manifold. Global similarity emerges if local similarity is present and the sensitivity differential equations are pseudohomogeneous. Global similarity means that the effect of the simultaneous change of several parameters can be fully compensated for all Variables, in a wide range of the independent Variable by changing a single parameter. Therefore, the presence of global similarity has far-reaching practical consequences for the “validation” of complex reaction mechanisms, for parameter estimation in chemical kinetic systems, and in the explanation of the robustness of many self-regulating systems.

3. Zsély IGy, Turányi T
The influence of thermal coupling and diffusion on the importance of reactions: The case study of hydrogen-air combustion
Physical Chemistry Chemical Physics 5: 3622-3631 (2003)

Detailed chemical kinetics mechanisms are usually developed on the basis of spatially homogeneous calculations, but utilized in the simulation of very complex physical models. A fundamental question is if the importance of reactions is determined solely by the temperature and the actual concentration set or if it is also influenced by the thermal and diffusion couplings present in the physical model. A 46-step detailed mechanism of hydrogen oxidation was studied at equivalence ratios 0.5, 1.0, 2.0, and 4.0. Six physical models were designed (homogeneous explosion, burner-stabilized and freely propagating laminar flames, with and without thermal coupling), which provided very similar concentration curves as a function of temperature, while the local sensitivity functions revealed that the couplings in these models were very different. The importance of the reactions in every model was investigated by the principal component analysis of the rate sensitivity matrix F (PCAF method), exploiting that the results of this method depend only on the concentrations and temperature. A fundamentally different method, the principal component analysis of the local sensitivity matrix S (PCAS method) was used to extract information on the importance of reactions from the sensitivity functions. The PCAF and PCAS methods selected identical reduced mechanisms at all conditions, which shows that these are equally effective methods for determining a minimal reduced mechanism. The good agreement between the results of the two methods in the case of all models demonstrated that the importance of reactions was independent of the physical model into which the mechanism had been embedded. Thermal coupling did not have any effect on the selection of the reduced mechanisms. The difference between the importance of reactions in explosions and flames was caused by the difference of the concentrations in the low-temperature regions and not by the presence of diffusion. The reduced mechanisms contained 15 to 28 reaction steps, depending on the equivalence ratio and the type of the model. All species were retained in models of the combustion of lean and stoichiometric mixtures, while species H2O2 could be eliminated at rich conditions. Description of near stoichiometric conditions required more reaction steps, while rich combustion could be described by few reactions. An overall reduced mechanism, applicable in a wide range of conditions, contained 31 reaction steps. Results of the PCAS method revealed the global similarity relations of the sensitivity matrices of adiabatic explosions.

4. Zádor J, Zsély IGy, Turányi T
Investigation of the correlation of sensitivity vectors of hydrogen combustion
International Journal of Chemical Kinetics 36: 238-252 (2004)

A well-established method for the analysis of large reaction mechanisms is the calculation and interpretation of the sensitivity of the kinetic model output Yi to parameter changes. Comparison of the sensitivity vectors s(i)={dY(i)/dp) belonging to different model outputs is a new tool for kinetic analysis. The relationship of the sensitivity vectors were investigated in homogeneous explosions, freely propagating and burner-stabilized laminar flames of hydrogen-air mixtures, using either calculated adiabatic or constrained temperature profiles, for fuel-to-air ratios phi = 0.5-4.0. Sensitivity vectors are called locally similar, if the relationship s(i)=lambda(ij)*s(j) is valid, where lambda(ij) is a scalar. In many systems, only approximate local similarity of the sensitivity vectors exists and the extent of it can be quantified by using an appropriate correlation function. In the cases of adiabatic explosions and burner-stabilized flames, accurate local similarity was present in wide ranges of the independent variable (time or distance), and the correlation function indicated that the local similarity was not valid near the concentration extremes of the corresponding species. The regions of poor similarity were studied further by cobweb plots. The correlation relationships found could be interpreted by the various kinetic processes in the hydrogen combustion systems. The sensitivity vector of the laminar flame velocity is usually considered to be characteristic for the whole combustion process. Our investigations showed that the flame velocity sensitivity vector has good correlation with the H and H2O concentration sensitivities at the front of the adiabatic flames, but there is poor correlation with the sensitivity vectors of all concentrations in homogeneous explosions.

5. Zsély IGy, Zádor J, Turányi T
On the similarity of the sensitivity functions of methane combustion models
Combustion Theory and Modelling (in press)

It is widely known that detailed kinetic mechanisms with identical reaction steps but with very different rate parameters may provide similar simulation results in combustion calculations. This phenomenon is related to the similarity of sensitivity functions, which arises if low-dimensional manifolds in the space of variables, and autocatalytic processes are present. We demonstrated the similarity of sensitivity functions for adiabatic explosions and burner-stabilized laminar flames of stoichiometric methane-air mixtures. The cause of similarities was investigated by calculating the dimension of the corresponding manifolds, and the pseudo-homogeneous property of the sensitivity ODE. The methane explosion model showed global similarity, which means that different parameter sets could provide the same simulation results. This was demonstrated by numerical experiments, in which two significantly different parameter sets resulted in identical concentration profiles for all species. This is important from a practical point of view in the fields of the 'validation' of complex reaction mechanisms and the parameter estimation of chemical kinetic systems.
calculated peak OH concentration.


10. Lovrics A, Zsély IGy, Csikász-Nagy A, Zádor J, Turányi T, Novák B
Analysis of a budding yeast cell cycle
model using the shapes of local sensitivity functions
International Journal of Chemical Kinetics 40 710-720 (2008)


The Chen et al. (Mol Biol Cell 2000, 1 1, 369-391) budding yeast cell cycle model is a biochemical kinetic model that describes how the controlling protein concentrations change during a proliferation cycle. Time dependence of local sensitivity coefficients was calculated for all variables and parameters of the model. Some of the local sensitivity coefficients-time functions could also be obtained from another one by multiplying it with a constant, which means that these functions exhibit global similarity. Local similarity of the sensitivity functions was also detected. The distance of the shapes of two scaled sensitivity functions was defined by the integrated squared difference of these functions. The distance matrices of function shapes were interpreted by a clustering method, and the shapes could be sorted to two main groups for each model variable. The presence of the global similarity of sensitivity functions means that the change of some enzyme activities car be fully compensated by changing the activity of other enzymes. This feature can be related to the robustness of living organisms.


12. Mészáros R, Zsély IGy, Szinyei D, Vincze Cs, Lagzi I
Sensitivity analysis of an ozone deposition model
Atmospheric Environment 43 663-672 (2009)


In this study, sophisticated sensitivity analyses of a detailed ozone dry deposition model were performed for five soil types (sand, sandy loam, loam, clay loam, clay) and four land use categories (agricultural land, grass, coniferous and deciduous forests). Deposition velocity and ozone flux depend on the weather situation, physiological state of the plants and numerous surface-, vegetation-, and soil-dependent parameters. The input data and the parameters of deposition-related calculations all have higher or lower spatial and temporal variability. We have investigated the effect of the variability of the meteorological data (cloudiness, relative humidity and air temperature), plant-dependent (leaf area index and maximum stomatal conductance) and soil-dependent (soil moisture) parameters on ozone deposition velocity. To evaluate this effect, two global methods, the Morris method and the Monte Carlo analysis with Latin hypercube sampling were applied. Additionally, local sensitivity analyses were performed to estimate the contribution of non-stomatal resistances to deposition velocity. Using the Monte Carlo simulations, the ensemble effect of several nonlinear processes can be recognised and described. Based on the results of the Morris method, the individual effects on deposition velocity are found to be significant in the case of soil moisture and maximum stomatal conductance. Temperature and leaf area index are also important factors; the former is primarily in the case of agricultural land, while the latter is for grass and coniferous forest. The results of local sensitivity analyses reveal the importance of non-stomatal resistances.

Papers without impact factor


1. Zsély IGy, Turányi T
Investigation and reduction of two methane combustion mechanisms
Archivum Combustionis, 21, 173-177 (2001)



4. Zsély IGy, Zádor J and Turányi T
Local and global similarity of sensitivity vectors of combustion kinetic models
Proceedings of the 3rd Mediterranean Combustion Symposium pp. 849-859, Marrakech, Morocco, June 8-13, 2003, Editors: F. Beretta and A. Bouhafid


7. Zsély IGy, Zádor J, Turányi T
Uncertainty analysis of updated hydrogen and carbon monoxide oxidation mechanisms
Proceedings of the Combustion Institute 30/1 1273-1281 (2005)

Uncertainty analysis was used to investigate H2/air and wet CO/air combustion mechanisms. The hydrogen/carbon monoxide submechanism of the Leeds Methane Oxidation Mechanism was updated on the basis of the latest reaction kinetics and thermodynamics data. The updated mechanism was tested against three hydrogen oxidation and two wet CO bulk experiments. Uncertainties of the simulation results, caused by the uncertainties of the kinetic parameters and the heat of formation data, were analysed. The methods used were the local uncertainty analysis and Monte Carlo analysis with Latin hypercube sampling. The simulated flame velocity had relatively large uncertainty in both hydrogen-air and wet CO flames. In the case of ignition experiments, for both fuels the uncertainties of the simulated ignition delay times were small and comparable with the scatter of the experimental data. There was a good agreement between the simulation results and the measured temperature and concentration profiles of hydrogen oxidation in a flow reactor. However, accurate ignition delay is not a result of the flow reactor experiments. The uncertainty of the required time correction for matching the simulated 50% consumption of H2 to that of the experimental one (corresponding to the simulated ignition delay) was found to be very large. This means that very different parameter sets provide very different ignition delays, but very similar concentration curves after the time correction. Local uncertainty analysis of the wet CO flame revealed that uncertainties of the rate parameters of reactions O2 + H (+M) = HO2 (+M), and CO + OH = CO2 + H cause high uncertainty to the calculated flame velocity, temperature, and peak concentrations of radicals. Reaction H + HO2 = H2 + O2 also causes high uncertainty for the calculated flame velocity. The uncertainty of the enthalpy of formation of OH is highly responsible for the uncertainty of the calculated peak OH concentration.


Web pages of I. Gy. Zsély were visited times.

This page was last modified on