Potos : Bruno Häubi/Team Advance
t least two tragic accidents (a hang glider and
the paraglider pilot Yann Espinasse), more
than ten years ago, have resulted in numerous
professionals prohibiting connecting the
reserve to the bridle on the harness via a
lark’s head. In Yann Espinasse’s case, after having a full line
breakage, the pilot opened his reserve in free fall. The lark’s
head which wasn’t centred in the loop burnt and severed
the latter, after which, the pilot was once again in free fall.
It was an extreme case. The PMA wanted to know if there
really was a big risk because numerous pilots or profes
sionals prefer not to use a metal link, albeit a good quality
one (e.g. a maillon made by Péguet) because it also has dis-
advantages: it’s an extra element in the chain of transfer of
forces and, if you forget about it when you are setting up, it
can remain open. A karabiner which turns sideways is also
a weak point (loaded in a way it wasn’t designed for).
Some pilots prefer not to have bits of metal in their back.
In collaboration with Advance and Edelrid, Guido Reusch,
working for the PMA, tested more than 100 connections
which use lark’s heads in modern bridles, particularly in
On Edelrid’s testing benches, the connectors were subject-
ed to a simulation of a 15 G shock with a 120 kg mass. The
connectors which weren’t centred, and with a slippage of
up to 15 cm, generally resisted without any major conse-
quence. The equipment thus tested was then retested with
a static load to discover any possible non visible weakness
es. The result: losses of up to a maximum of 5%, therefore
within the permissible margin of error.
This result is all the more interesting as Dyneema is known
for being sensitive to heat: one would imagine that it would
burn more easily. But according to Guido Reusch, Dynee-
ma is so smooth that the friction and heating remain mini-
mal. On the other hand, the protection sheaths made from
other materials, sometimes present on Dyneema loops,
were marked by the heating. Therefore they could even be
considered counter productive!
The conclusion of the PMA’s test: a lark’s head connector,
properly centred and blocked on the other loop, is at least
as resistant as a connection via a karabiner.
A connection sling/
maillon/sling also
isn’t 100% infallible:
during the tests, in an
atypical configuration
and with a load
slightly above the
maximum authorised
load, a Dyneema loop
was severed by the
During tests with
connections, lark’s
heads which weren’t
centred were subject
to about 15 G for
120 kg and slid
on the other strap
almost without