SAMPE Journal - January/February 2021 - 26

FEATURE / REINFORCED POLYMER HANGERS
Figure 4.

High strength polyester yarns tie the two layers of
plies, each of rectangular cross-section, in lateral
direction together into one circular cross-section.
After the application of additional reinforcement in
the region of the thimbles the hanger is cured in an
oven at 140°C. Carbo-Link produced in autumn 2016
for the static and fatigue experiments three identical
CFRP hangers of 2.8 m in length and 33 mm diameter on the free length (Figure 4) numbered A to C for
the " single case approval " (ZiE).
SIMULATION OF 100 YEARS OF TRAIN
OPERATION
The design methods in steel and concrete construction are at a very high level today due to great global
research efforts at least since a century. The construction sector is applying CFRP only since 1991.
The design methods are therefore not yet as sophisticated as in steel and concrete construction. Fullscale experiments serve the calibration of the finite
element and the analytical calculations.
100 years of train operation correspond to 11
million load cycles between 149 and 328 kN for the
highest loaded hangers. It was originally planned to
subject the CFRP " hanger A " to cyclic loading with
a minimum load of 115 kN and a maximum load of
575 kN at a frequency of 4.2 Hz. After approximately 70 minutes of cyclic loading in the fatigue-test rig
(Figure 5), the temperature raised up to 80°C at the
outer surface of the CFRP layers around the thimbles. This was an important reason to interrupt the
experiment. It was assumed that the temperatures

Figure 5. Fatigue-testing rig.

at the titanium/CFRP interfaces would be higher
than the glass transition temperature Tg of the CFRP
matrix of 138°C and that the " hanger A " would fail
prematurely if the fatigue tests were continued under the same conditions. The ISO standard 13003
allows a maximum temperature increase DT of 10°C
for fatigue tests on fibrous composites. After further
short tests with different frequencies and load limits as well as intensive discussions, it was decided to
accept the fatigue test with a minimum load of 193
kN and a maximum load of 462 kN. The time- and
cost-pressure enforced to stay with the frequency of 4.2 Hz but to reduce the load amplitude and
the maximum load. The temperature at the outer
surface of the CFRP layers around the thimbles stabilized under the new conditions at 45 to 50°C. By
this time, the " hanger A " had already experienced
816,000 load cycles and was pre-damaged. The fretting between the titanium thimbles and the CFRP
layers produced from the beginning of the fatigue
tests, dust-like black microparticles observed on the
terminations of the hanger and on the surrounding
components of the loading device. These were clearly CFRP particles, wear products. The relative movement between titanium and CFRP and the still too
high temperature caused a non-quantifiable material removal, especially on the flanks of the CFRP
loops. Adhesive and abrasive wear caused this abrasion. The main reason was the high frequency of 4.2
Hz and the associated heating of the CFRP hanger
in the loop areas. In reality, the load frequency will
reach in the worst case 0.003 Hz under the assumption that a train crosses the bridge every three minutes. Under such loading frequencies, there will be
no thermal damage and the wear greatly reduced.
The identical CFRP " hanger B " was installed in
the vertical test rig connected at the bottom to the
tensioning device and at the top to the original steel
arch section of the bridge. This " chain " was also
subjected to a cyclic load with a minimum load of
193 kN, a maximum load of 462 kN and a frequency
of 4.2 Hz. In this test too, the temperatures on the
CFRP loops rose to 45 to 50° C. There was no previous damage compared to CFRP hanger A, there was
significantly less wear.
LOAD/ELONGATION BEHAVIOR OF CFRP
HANGERS
The load/elongation behavior of the CFRP hangers
A and B under quasi-static loading was determined
according to11 before and after the fatigue tests. The
CFRP hangers showed an almost perfect linear elastic behavior over the entire load range. During unloading resulted a very narrow hysteresis loop. There
were no permanent deformations.

SAMPE JOURNAL

J A N UA R Y/ F E B R UA R Y 2 0 2 1

w w w. s a m p e . o r g

http://www.sampe.org

SAMPE Journal - January/February 2021

Table of Contents for the Digital Edition of SAMPE Journal - January/February 2021