Citation Link: https://nbn-resolving.org/urn:nbn:de:hbz:467-6816
Modelling and numerical simulation of contact and lubrication
Source Type
Doctoral Thesis
Author
Subjects
SubStokes model
Reynolds model
Signorini-contact
linear finite elements
DDC
510 Mathematik
GHBS-Clases
Issue Date
2012
Abstract
This thesis considers the simulation of a steel forming process including the lubricant between tool and workpiece. The model equations, that describe deformation, contact and hydrodynamic flow, are derived from fundamental physical laws. For the contact-simulation the two models of Signorini- and z-contact are introduced. The 3-dimensional Stokes problem for a thin fluid film is condensed to the 2-dimensional SubStokes model (for velocity and pressure) and further reduced to the Reynolds model, which is an equation for the pressure only. All fluid models are extended to variational inequations to consider cavitation. Basic mathematical concepts for the numerical treatment with the finite element method and the corresponding analysis are presented. Error estimation is done separate for the modelling error and discretisation error. An algorithm for model adaptivity is given and model error estimates are derived out of the physical models for z- and Signorini-contact aswell as SubStokes- and Reynolds-fluids.
Error estimates for the discretisation error of elliptic problems - deformation and Reynolds-flow - are presented. The SubStokes problem is stabilized to apply linear finite elements. To estimate the discretisation error of the SubStokes inequation a Lagrangian multiplier is introduced. An algorithm for a simultanous refinement of model and grid is given. Finally the presented methods and estimates are validated by the application to various prototype examples. The results allow to compare the different models (z- and Signorini-contact, Reynolds- and SubStokes-fluid).
Error estimates for the discretisation error of elliptic problems - deformation and Reynolds-flow - are presented. The SubStokes problem is stabilized to apply linear finite elements. To estimate the discretisation error of the SubStokes inequation a Lagrangian multiplier is introduced. An algorithm for a simultanous refinement of model and grid is given. Finally the presented methods and estimates are validated by the application to various prototype examples. The results allow to compare the different models (z- and Signorini-contact, Reynolds- and SubStokes-fluid).
File(s)![Thumbnail Image]()
Loading...
Name
gimbel.pdf
Size
11.63 MB
Format
Adobe PDF
Checksum
(MD5):7fcf4f76d23b0f32271c758a5f926d08
Owning collection