wood pyrolysis simulation won't run

[poppy.fryer]

I am trying to recreate the simulation in the paper: 'A model for the oxidative pyrolysis of wood', Lautenberger, Fernandez-Pello, 2009. This can be found here: https://www.sciencedirect.com/science/article/pii/S0010218009000832 

I have put in all of the parameters listed in tables 1-5 in the report (see attached spreadsheet) and used the general parameters used in the 1D pyrolysis simulation in the Gpyro work book, however when I try and run the simulation, the CMD pops up briefly and then disappears and nothing happens. Initially when I tried to convert the excel file to a data file I kept getting the error message: 'Run time error '52': Bad file name or number', but eventually it seemed to let me convert it. I'm not sure if I've just put something into the spreadsheet incorrectly, or if there are more fundamental errors in the way I've set up the simulation.

Any help on getting the simulation to run would be greatly appreciated. 

Thanks very much, 

Poppy

[poppy.fryer]

Dear all, 

I have solved the initial problem with getting the simulation to run. The simulation will run in a nitrogen atmosphere, producing very similar results to those published in the paper by Lautenberger. 

However, when I try and run the simulation in an air atmosphere however, the simulation runs to appox. 130s, where the Gas Yj residual is the order of 10^-5 and then suddenly increases to a residual of 10 and the simulation will not converge. I have tried excluding each of the reactions individually and changing the i_O2 rxn index to 1 for the appropriate reactions. 

I would be very grateful if anyone could provide any insight into why the simulation will not converge and have attached the excel file. 

Thanks very much,

Poppy 

[frarich]

Dear all,

it seems that the problem only appears in the 0.8 version, while the model runs well in the 0.7 version. I noticed that the problem does not occur when one switches of oxygen diffusion at the front surface (set hm=0) and allows the numerical code more tolerance for conversion. I hope that helps for now to progress with the problem.

Franz