Impact problems in geotechnical
engineering typically encountered by the Defense Nuclear Agency (DNA) can be
modeled by DYNA, the general purpose finite element codes of Lawrence Livermore
National Laboratory (LLNL). The DYNA cap model is one of the geological
material models in DYNA codes. This cap model is based on classical incremental
theory of work hardening plasticity. It was developed as part of a constitutive
model to control the excessive dilation that is predicted by the earlier
elastic perfectly-plastic models. Several versions of cap model have been
successfully developed with a number of features which can be included or
excluded in specific problems. However, the DYNA cap model has deficiencies and
can give deceptive results. Fixing or replacing the DYNA cap model was the main
objective of this study.
In this study, a modified cap model
with new features such as Koiter’s generalized flow rule, Kuhn-Tucker
optimization conditions, and recent developments in return mapping alogotithms
was developed. It is based mainly on Hofstetter, Simo, and Taylor’s 1993 cap
model. A subroutine containing the modified cap model was written that can fit
the general purpose finite element code DYNA-3D. Furthermore, two applications
were carried out to check the applicability of the modified cap mode and
compare it to the existing one in DYNA-3D. In the first application uniaxial
strain tests were simulated using the Material Model Driver of DYNS-3D. In the
second application the results of finite element analysis were compared with laboratory
experiments of a free-falling weight penetrating a soil specimen in a confined
model. Two results wee compared; the maximum penetration depth caused by the
falling weight and the deceleration of the falling weight while penetrating the
soil. When compared with the existing model, the modified cap model is shown to
be faster, requires less computer time, is simple and easy to follow from
programming point of view, contains new features, and predicts heave, which the
existing cap model could not do. In general, the modified cap model is shown to
be an improvement over the existing model. Some of the bugs in the general
purpose finite element code DANA-3D are described. However, more detailed and
various laboratory and/or field applications may be required.