You can read it here: http://maeresearch.ucsd.edu/~mgollner/publications/2012_Gollner_PhD_Dissertation_Final.pdf
Abstract of the Dissertation:
Experimental
techniques have been used to investigate three upward flame spread phenomena of
particular importance for fire safety applications. First, rates of upward
flame spread during early-stage burning were observed during experiments on
wide samples of corrugated cardboard. Results indicated a slower acceleration
than was obtained in previous measurements and theories. It is hypothesized
that the non-homogeneity of the cardboard helped to reduce the acceleration of
the upward spread rates by physically disrupting flow in the boundary layer
close to the vertical surface and thereby modifying heating rates of the solid
fuel above the pyrolysis region. The results yield alternative scalings that
may be better applicable to some situations encountered in practice in
warehouse fires.
Next,
a thermally thick slab of polymethyl methacrylate was used to study the effects
of the inclination angle of a fuel surface on upward flame spread. By
performing experiments on 10 cm wide by 20 cm tall fuel samples it was found
that the maximum flame-spread rate, occurring nearly in a vertical
configuration, does not correspond to the maximum fuel mass-loss rate, which
occurs closer to a horizontal configuration. A detailed study of both flame
spread and steady burning at different angles of inclination revealed the
influence of buoyancy-induced flows in modifying heat-flux profiles ahead of
the flame front, which control flame spread, and in affecting the heat flux to
the burning surface of the fuel, which controls fuel mass-loss rates.
