[kbhamidipa3]

I am trying to model heat transfer in a battery pack. The pack consists of some cells and each cell is separated by a 1 mm thick mica insulation. The cell width is about 30 times higher than this thickness. I simulated two cases keeping everything the same except the following;
1. I included 1 mm mica as part of the domain separating every cell in the first case
2. I am capturing the mica spacing and its thermal resistance using shell conduction in the second case. I identified the end cell faces as walls and checked shell-conduction and specified 0.001 m for thickness.

My understanding from reading FLUENT manual is that there should be no difference whether we take path 1 or path 2. However, when I compared the cell temperatures at the end of simulation, the differences in the temperatures are higher when mica is included in the domain versus simulation using shell conduction. In other words, when actual mica was included, it seems to be capturing additional thermal resistance. I am unable to understand the reason for this discrepancy. Any help with this situation is greatly appreciated.