izpis_h1_title_alt

Matematično modeliranje požarov v naravi
ID Kušar, Iztok (Author), ID Schnabl, Simon (Mentor) More about this mentor... This link opens in a new window

.pdfPDF - Presentation file, Download (4,94 MB)
MD5: 61E98F3CBE7DA636DFA2D6408C9B8C5D

Abstract
V magistrskem delu je predstavljen matematični model za analizo in simulacijo požarov v naravi. Računalniški program je namenjen ponazoritvi razvoja talnih požarov. Sprogramiran je v programskem orodju Wolfram Mathematica po metodi celične avtomate (CA), ki spada med diskretne načine matematičnega modeliranja. Osnova algoritma CA je semiempirični matematični model, ki je izpeljan iz splošne energijske bilance, poenostavljene na podlagi osmih predpostavk. Koncept prenosa toplote je empirične narave. Uporablja geometrijski model, ki je analogen točkastemu modelu, vendar je velikost toplotnega toka plamena (Q̇c) odvisna od kalibracijskega faktorja (η), umerjenega na osnovi eksperimentalnih podatkov iz literature. Osnovnemu algoritmu CA je dodan še t. i. razširitveni del, ki je namenjen simulaciji požarnih preskokov in v nasprotju z osnovnim delom temelji na stohastičnem principu ter uporablja večjo sosesko CA. Obnašanje umerjenega matematičnega modela smo preizkusili s parametrično študijo in z analizo občutljivosti. Rezultate parametrične študije, v kateri smo spremljali odvisnost hitrosti požara (ROS) od vrednosti posamezne spremenljivke, smo primerjali z ugotovitvami iz literature. Analizo občutljivosti smo izvedli po principu polne dvostopenjske faktorialne analize po metodi DOE (angl. Design Of Experiments). Na podlagi standardnih odklonov ROS (s) smo primerjali velikosti vplivov testiranih spremenljivk v matematičnem modelu. Eksperimentalne rezultate smo analizirali tudi v programu Minitab, kjer smo s pomočjo t-testa določili, ali so testirane spremenljivke v matematičnem modelu statistično signifikantne. Na podlagi rezultatov analize matematičnega modela smo ugotovili, da je le-ta primeren za ponazoritev razvoja talnih požarov, saj se simulacije v različnih pogojih obnašajo skladno z opažanji iz literature. Pokazali smo, da je za uspešno širjenje požara potrebno minimalno razmerje gorljivih in negorljivih celic (p), ki je pri pogojih numeričnega eksperimenta znašalo 0,54. Kritična vrednost masnega deleža vlage v gorivu (Hu,k), pri kateri širjenje požara ni več možno, je odvisna od površinske gostote goriva (W). Vse testirane spremenljivke v matematičnem modelu so statistično signifikantne, pri čemer med velikostma vplivov začetne temperature goriva (T0) in temperature vžiga (Tvž) ni statistično signifikantne razlike. Najvplivnejši spremenljivki matematičnega modela sta konstanta hitrosti reakcije (kr) in kurilna vrednost goriva (Hcom), ki primarno vplivata na velikost toplotnega toka plamena (Q̇c) in višino plamena (Hf). Najmanjši vpliv med vsemi spremenljivkami pa je imela specifična toplota suhe snovi (cs).

Language:Slovenian
Keywords:požari v naravi, matematično modeliranje, celična avtomata
Work type:Master's thesis/paper
Typology:2.11 - Undergraduate Thesis
Organization:FKKT - Faculty of Chemistry and Chemical Technology
Year:2022
PID:20.500.12556/RUL-141342 This link opens in a new window
COBISS.SI-ID:132061443 This link opens in a new window
Publication date in RUL:28.09.2022
Views:569
Downloads:55
Metadata:XML DC-XML DC-RDF
:
Copy citation
Share:Bookmark and Share

Secondary language

Language:English
Title:Mathematical modeling of wildfires
Abstract:
The master's thesis presents a mathematical model for analysis and simulation of wildfires. The computer program is intended to emulate the progression of ground fires. It is programmed in the Wolfram Mathematica software tool using the cellular automata (CA) method, which belongs among discrete techniques of modelling. The fundament of the CA algorithm is a semiempirical mathematical model. It is derived from the general energy balance, which is simplified based upon eight assumptions. The concept of heat transfer is empirical in nature. It uses a geometric model that is analogous to the point source one, but the magnitude of the flame heat flux (Q̇c) is contingent on the calibration factor (η), which is calibrated based on experimentally obtained data from the scientific literature. Moreover, to the primary CA algorithm an extension section is added, which is designed to simulate firebrand spreading. Unlike the elementary part of the program, it is founded on the stochastic principle and uses a larger CA neighbourhood. The response of the calibrated mathematical model was tested by a parametric study and sensitivity analysis. The results of the parametric study, where we observed the dependence of rate of fire spread (ROS) on the value of individual model parameters, were compared with other published research data. The sensitivity analysis was performed on the principle of two level full factorial analysis in accordance with the DOE (Design Of Experiments) method. Based on the standard deviations (s), we determined which variables our mathematical model depended on the most. Data of the full factorial analysis was also evaluated in the Minitab program, where we used a t-test to assess whether influence of model parameters is statistically significant. Considering the behaviour of simulations in different circumstances and results of the analysis of the mathematical model we have concluded, that the computer program is suitable to represent the development of ground fires. Furthermore, we have shown that a minimal ratio of combustible to non-combustible cells (p) is required for successful fire spread, which in the conditions of our numerical experiment was equal to 0,54. Moreover, the critical value of mass fraction of moisture in the fuel (Hu,k) is dependent upon the fuel load (W). All the tested variables in the mathematical model were statistically significant without any statistically significant difference between the influences of the initial fuel temperature (T0) and ignition temperature (Tvž). The predominant variables of the mathematical model are the reaction rate constant (kr) and the heat of combustion (Hcom), which primarily affect the heat release rate of the flame (Q̇c) and flame height (Hf). However, the specific heat of dry fuel matter (cs) had the smallest effect among all tested variables.

Keywords:wildfires, mathematical modelling, cellular automata

Similar documents

Similar works from RUL:
Similar works from other Slovenian collections:

Back