[an error occurred while processing this directive]
Experiments in PMMA
Jens Hauch and Michael Marder performing a highly controlled experiment
In an experiment we introduce seed cracks into the plates and
load them in extension, i.e. we pull on them perpendicular to the anticipated
crack path. To measure the crack velocity we coat the plates with a 30nm thick
coating of Aluminum. We attach elctrodes on either side of the seed crack, as
the crack propagates it breaks the coating such that the resistance between
the electrodes is a linear function of the cracklength (to see the
aesthetically very pleasing results of simulations to determine the current
density everywhere on the plate click here [11.1K]).
This approach allows
us to measure the crack velocity with an accuracy of 10m/s over a bandwidth
of 10MHz. The fracture velocities range from 200m/s to 600m/s in PMMA, and
from 900m/s to 2000m/s in Glass.
Schematic of the experimental setup
Our experiments reveal several new characteristics of crack dynamics. Initially
there is a very fast velocity jump, to roughly 200m/s in PMMA and 900m/s in
Glass, that is independent of loading and stress at fracture initiation. We
also found that large velocity fluctuations set in at about 45% of the
soundspeed in the material.
Typical time series obtained from the velocity measurements.
Note the initial velocity jump, and the onset of velocity fluctuations at
roughly 400m/s.
The velocity fluctuations are accompanied by
increased roughness of the fracture surface which in PMMA also has a periodic
component that disappears at higher velocities. The velocity fluctuations and
increased surface roughness are the signature of a bifurcation in the dynamics
due to a microcracking instability. The mechanism of this bifurcation is such
that as the crack reaches a critical velocity it will attempt to branch, this
branching slows down the crack, it propagates straight again and the process
repeats.
Detailed photographs of the fracture surface for different
crack velocities.
From the left: 460m/s, 570m/s, 620m/s.
Both these dominant features of the dynamics, the velocity jump and
the microcracking instability are qualitatively reproduced by several lattice
models of fracture that were also developed here at the Center for Nonlinear
Dynamics. The models indicate that the velocity jump is due to a forbidden
band of velocities for the steady state propagation of cracks in crystalline
materials. The models also revealed the mechonism of the microcracking
instability. Future work will focus on reproduction of these results in
various materials, and further experiments to determine if the initial velocity
jump is in fact due to a forbidden band of velocities.
For more detailed information we have a
detailed list of publications that
originated here at the center for nonlinear dynamics.
Return to fracture homepage.
Special thanks to Steve Gross, Jay Fineberg, Mike Marder, and Xiangming Liu
for their contributions to this project
For more information or comments about the project,
send email to Michael Marder at
marder@chaos.ph.utexas.edu