The single body theory of meteoroid interaction with the atmosphere in
the form of l = l(t), where l is the relative distance along the mete
oroid trajectory and t the relative time, is generalized by allowing f
or points of sudden gross-fragentation. This theory can be directly ap
plied to photographic observations of meteors, comparing the measured
distance, l(obs), with the model computed distance, l, by means of the
least-squares method. The no-fragmentation case has 4 free parameters
to be determined from observations (l(o), v(o), v(oo) sigma), distanc
e and velocity at time t = O, velocity at t = -oo and ablation coeffic
ient, respectively. The case of one gross-fragmentation point adds two
more free parameters to the problem, v(oo)2 and sigma(2), velocity at
t = -oo and ablation coefficient, both for the second part of the tra
jectory after fragmentation. Moreover the position of the gross-fragme
ntation point and relative amount of fragmented mass can also be deter
mined from photographic observations by searching for the best fit for
different choices of the fragmentation point. The single body model a
pplied to theoretically computed data for a meteoroid with fragmentati
on points enabled insight into time dependence of residuals, their rel
ation to position of fragmentation point and influence on the resultin
g ablation coefficient, initial velocity and mass. Moreover, the model
was checked by means of 11 PN (Prairie Network) photographs, where th
e splitting into individual fragments was directly visible and the fra
gmentation point has been determined geometrical way from intersection
s of trajectories of individual fragments. A computer program for auto
matic search of gross-fragmentation points was applied to 80 records o
f PN fireballs. Fireballs can be separated according to their sudden f
ragmentation into NF (no-fragmentation), 1F(one fragmentation point),
MF (many fragmentation points) and LA (low accuracy, cannot decide) cl
asses. The dynamic pressure rho v(2) (rho the air density) at the frag
mentation points of 1F and MF fireballs tends to cluster at several se
parate values. This is used for proposing a classification into severa
l strength categories (a: 0.8 Mdyn/cm(2), b: 2.5, c: 5.3, d: 8.0, e: 1
1). The ablation ability groups I, II, IIIA and IIIB proposed earlier
and these new proposed strength categories form a possibility of two d
imensional classification for 1F and MF fireballs separating meteoroid
composition and structure. Ablation coefficients proved to be lower t
han the previously determined statistical values, while the bulk densi
ties are approximately the same as statistically determined. Ablation
coefficients determined for meteoroids with gross-fragmentation are ha
lf of the values, which would result for the same fireball without tak
ing the fragmentation effect into consideration. Gross-fragmentation a
t a point is more frequent for stronger types of fireballs. Typical am
ount of fragmented mass at the fragmentation point for 1F and MF fireb
alls is 60%. 1F fireballs have another typical value of the amount of
fragmented mass, i.e. between 95% and 99%, which corresponds to almost
complete disruption of the body. Initial velocities and masses for 1F
and MF fireballs have to be determined by means of our model: their v
alues without taking the gross-fragmentation effect into account are s
ignificantly different. We give also list of classified fireballs cont
aining all PN fireballs, to which we applied our model. We discuss als
o precision of our results with special attention to the Lost City fir
eball. Our method needs very precise fireball records. PN fireballs ar
e the most precise data available, but