University of Twente Student Theses

Login

Internship Report Jasper Reichardt

Reichardt, J.M. (2015) Internship Report Jasper Reichardt.

Full text not available from this repository.

Full Text Status:Access to this publication is restricted
Abstract:At the most generic level, a recent EU project DESICOS (new robust DESIgn guideline for imperfection sensitive COmposite launcher Structures) is an effort to devise new design guidelines for imperfection-sensitive composite structures. One of the types of structures under investigation is the unstiffened thin-walled truncated cone. This type of structure can for example be found in the Ariane 5 rocket. These cones are loaded in axial compression, such that the main failure mode is buckling. ir. Regina Khakimova is currently doing a PhD research project at DLR in order to study this type of structure in detail. Her research includes the design and manufacturing of the cones, as well as subsequent experimental tests. The first batch of tests took place during my internship. My main task was to validate these experiments using finite element simulations, using the commercial software package Abaqus. An important aspect was to thoroughly document the parameters and results of these simulations, such that similar FEM simulations may be done for subsequent test batches. In many cases, new and/or modified software tools and/or scripts were required for the FEM simulations. These were documented as well and made available for re-use. The first sub-task was to do a detailed analysis of the possible boundary conditions. There were twelve possible combinations of composite lay-up and boundary conditions to be analysed. In the end it was concluded that restricting all degrees of freedom of the cone edges (“clamping”) provided sufficiently accurate results, while also being the most straightforward option. Hence the decision was made to use these boundary conditions instead of more realistic (albeit also more complex) ones. The second step was to make finite element model match the experimental cone as well as possible. This means that the imperfections in the actual cone should be included in the model as well. These imperfections include:  Deviations in cone geometry, based on measurements of the experimental cone  Deviations in material thickness, based on measurements of the experimental cone  Deviations in fibre orientation, based on the pieces of material that were used to construct the cone  Deviations in fibre volume fraction, related to the variations in thickness  Circular cut-outs, i.e. purposely added circular holes with a diameter of 10-50 mm A special Abaqus plug-in, written in Python, had already been created previously for the DESICOS project. This plug-in allows (amongst other things) to easily create FEM models of cones and to include deviations in geometry and thickness. Quite some processing (i.e. writing scripts) was however required to put the measured deviations into a format that could be read by this plug-in. For the fibre volume fraction deviations, I extended the plug-in allow both myself and (subsequent) 2 others to easily include this type of imperfection. The creation of cut-outs also required an extension of the plug-in. This extension was mainly written by someone else, with some testing, fixes and improvements by me. For the variations in fibre orientation, a previous student had written a separate tool (labelled “CPPOT”) to calculate the fibre orientation at each location on a cone, based on a set of design parameters. This tool could then be used to compare and evaluate different designs. I re-implemented the core algorithms behind this tool as a separate code module. Subsequently I extended to Abaqus plug-in to use this module to assign material orientations to each individual element. I updated the CPPOT-tool re-use this same code module, fixing some bugs in the process. Also I included in the DESICOS software package, such that it is available for use by others. Before combining all imperfections into a single finite element model, many simulations were done with only one of them, or just a few. These FEM simulations were necessary for two reasons. Firstly, they provide a lot of insight into how each individual imperfections affect the behaviour of the cone. Secondly, in many cases there were parameters that could be varied or some other choice that had to be made, these alternatives were evaluated and compared before deciding on a “final” FEM model. In the meantime, experiments at the DLR buckling test facility started. Each experiment resulted in an Excel sheet with data, requiring a bit of processing to e.g. obtain a proper-load displacement curve. Based on the “final” FEM model I ran simulations corresponding to each of these experiments, so the experimental and numerical results could be compared. In general these results agreed pretty well, which is of course a very positive result. My last weeks were spent mostly spent on writing a detailed (approximately 100 pages) internal report on all my activities and results. A second series of experiments is scheduled to be carried out in June. The detailed report was written with the intent that based on its contents, someone else should be able to do the accompanying FEM simulations in the same manner as I did. Both the experimental and FEM results are destined to appear in academic publications later this year.
Item Type:Internship Report (Master)
Clients:
Deutsches Zentrum für Luft- und Raumfahrt (DLR), Germany
Faculty:ET: Engineering Technology
Subject:52 mechanical engineering
Programme:Mechanical Engineering MSc (60439)
Keywords:Composite, buckling, truncated conical shell, Single pertubation load-approach, imperfections
Link to this item:http://purl.utwente.nl/essays/70256
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page