Month: June 2020

June the 24th – PhD Defense Nicolas Gartner

Abstract

Identification of hydrodynamic parameters by simulation with Smoothed Particle Hydrodynamics

This thesis focuses on techniques that allows the simulation of dynamic interactions between an underwater vehicle and the surrounding water. The main objective is to propose a satisfactory solution to be able to test control algorithms and hull shapes for underwater vehicles upstream of the design process. In those cases, it would be interesting to be able to simulate solid and fluid dynamics at the same time. The idea developed in this thesis is to use the Smoothed Particles Hydrodynamics (SPH) technique, which is very recent, and which models the fluid as a set of particles without mesh. In order to validate the simulation results a first study has been performed with a hydrodynamic pendulum. This study allowed the development of an innovative method for estimating the hydrodynamic parameters (friction forces and added mass) which is more robust than previous existing methods when it is necessary to use numerical derivatives of the measured signal. Then, the use of two types of SPH solver : Weakly Compressible SPH and Incompressible SPH, is validated following the validation approach proposed in this thesis. Firstly, the behaviour of the fluid alone is studied, secondly, a hydrostatic case, and finally a dynamic case. The use of two methods for modelling the fluid-solid interaction : the pressure mirroring method and the extrapolation method is studied. The ability to reach a limit velocity due to friction forces is demonstrated. The results of the hydrodynamic parameters estimation from simulation tests are finally discussed. The simulated added mass of the solid approaches reality, but the friction forces currently seem not to correspond to reality. Possible improvements to overcome this problem are proposed.

Keywords : Hydrodynamic parameters, SPH, Numerical method, Interaction, Fluid-solid, Fluid-structure, Incompressible flow, Underwater robotics.

PhD Defense of Maxime Chalvin, July the 9th

Additive manufacturing of tubes by multi-axis robotized wire deposition :
Trajectory generation and optimization

Additive manufacturing through Directed Energy Deposition (DED) enables small batches
of parts to be rapidly manufactured. However, manufacturing trajectories usually used
for the manufacture of overhanging parts require the use of supports, material which
is not useful for the finished part and time consuming. If multi-axis trajectories can be
used to avoid them, they present generally a heterogeneous local inter-layer distance, thus
requiring a variation of the deposition parameters to adapt the layer height ; variation that
can be harmful to the mechanical characteristics of the final part. This thesis first proposes
a constant local inter-layer trajectory generation method for DED additive manufacturing
of tubular parts defined by parametric curves and which can have profile radius variations.
The proposed trajectories have been validated by robotized manufacturing trials of polymer
parts. Since the rotation about a coaxial deposition tool axis has no impact on the deposit,
the use of 6-axis robots offers a redundancy. Using this redundancy, a layer by layer
optimization of the trajectory in the robot space is then proposed. In a constrained robot
configuration, the trajectory optimization allows the manufacturing of parts that cannot
be manufactured in the usual way, and improves the geometrical quality of the parts with
a better repeatability.