We analyse in detail the role of additional hadron–hadron
interactions in elastic photon–initiated (PI) production at the LHC,
both in pp
and heavy ion collisions. We first demonstrate that the source of
difference between our predictions and other results in the literature
for PI muon pair production is dominantly due to an unphysical cut that
is imposed in these latter results on the dimuon–hadron impact
parameter. We in addition show that this is experimentally disfavoured
by the shape of the muon kinematic distributions measured by ATLAS in
ultraperipheral PbPb collisions. We then consider the theoretical
uncertainty due to the survival probability for no additional
hadron–hadron interactions, and in particular the role this may play in
the tendency for the predicted cross sections to lie somewhat above
ATLAS data on PI muon pair production, in both
pp
and PbPb collisions. This difference is relatively mild, at the
\sim 10\%∼10%
level, and hence a very good control over the theory is clearly
required. We show that this uncertainty is very small, and it is only by
taking very extreme and rather unphysical variations in the modelling of
the survival factor that this tension can be removed. This underlines
the basic, rather model independent, point that a significant fraction
of elastic PI scattering occurs for hadron–hadron impact parameters that
are simply outside the range of QCD interactions, and hence this sets a
lower bound on the survival factor in any physically reasonable
approach. Finally, other possible origins for this discrepancy are
discussed.