On the assumption that the roof is fairly high in this building, the smoke from a fire at floor level is likely to be relatively cool by the time it has ascended to roof level. The size of the panels will play a part in their softening and falling out - the larger the panel the less resiliant it will be due to the stresses induced within its own structure by its own mass and consequential turning moments. Small ones, like the typical 0.5m square ones you see so often in these buildings, would have little weight of their own to assist them in dropping out quickly when heated.
I suspect that you would probably need a pretty large fire to get these panels to drop out and by that time a lot of the damage may have been done to the valuable contents of the building.
You may be able to find data online about the melting temperature of the material involved (if you can establish the material!) but what you really need is data about how the material softens with increasing temperature and then, for an accurate model of the failure of these units, you would have to tie that data up with the size, thickness and mass of the units. Also, you'd have to consider the specific heat capacity of the material and the means by which heat is transferred from the hot gases to the units as well as other potential simultaneous loads such as the wind. It's all getting a bit complicated and I understand now why you asked your initial question. Good one.
I further suspect that the model you're seeking doesn't exist but, having said that, I hope it does and I hope someone comes up with it. It would be costly to develop now - it's almost bound to be cheaper and more effective to put in new vents.
Stu
