Mechanics of Extreme Thin Composite Layers for Aerospace Applications – Update 10-2016 – Introduction

View Github Repository Open presentation in a new window

LucaDiStasio

See all presentation from LucaDiStasio

Mechanics of Extreme Thin Composite Layers for Aerospace Applications – Update 10-2016 – Introduction

0 0

tpm-update_2016-10

Slides made for project update 2016-10

On Github LucaDiStasio / tpm-update_2016-10

Mechanics of Extreme Thin Composite Layers for Aerospace Applications

Update 10-2016

Luca Di Stasio

How to Navigate the Slides

Press ESC to enter the slide overview.

Press ESC again to exit the slide overview.

Hold down Alt and click on any element to zoom in.

Alt + click anywhere to zoom back out.

Use keyboard arrows or click on the displayed arrows to move through the slides.

Use the spacebar to go to the next slide.

Goals

... or, where are we heading to?

Main objectives

  • Develop versatile tools to facilitate the geometric modeling and mesh generation for the creation of FEM models
  • Specify details of micromechanical models for the study of thin ply mechanics

Requirements

Developments

... or, what has been done?

Let's

  • visualize the status of the software and the code in June 2016
  • visualize their current status (October 2016)
  • compare them

The software in June 2016

The software in June 2016

  • The front-end and its code are in the same notebook

The code in June 2016

The code in June 2016

  • Very few functions
  • Functions are long, complex and highly specialized

The software in October 2016

The software in October 2016

  • The front-end and its code are in separate notebooks, i.e. the front-end is independent
  • The connections between the Mathematica front-end and the Matlab back-end are working correctly, on different computers
  • The overall program works correctly on different computers, given that its dependencies are satisfied (Matlab, Mathematica and Matlink installed)

The code in October 2016

The code in October 2016

  • Significant increase in the number of functions
  • Functions are shorter and less complex
  • Many functions are much less specialized
  • Specialized functions differ by only a few parameters or lines of code

Side-by-side comparison

Side-by-side comparison

Issues

... or, what problems have arised?

One main problem

How to correctly estimate the time needed for the development

Tendency to underestimate

How to tackle it?

One minor problem

One minor problem

Laptop's fan broke down in August during EEIGM's summer closure

Thus,

no access to office's computer

no laptop to work from home

Solution

Old fan substituted with new one builted out of scrap parts

However, it took around 4/5 days to be fully operational again

TO-DO's

... or, what's next?

In the next 4-6 weeks

  • Create transpiler to abaqus, for FEM analysis: est. 5/6 days
  • Test transpiler to abaqus: from 0 up to 4 days
  • Test and correct any unobserved bug: from 0 up to 7 days (which may neither be consecutive nor at this stage)
  • Start FEM analysis: in 5 to 17 days (depending on debug phase length)

In the next 4-6 weeks

  • Create and test data writer to json format, for portability: est. 1/2 days
  • Create and test data writer to csv format, for analysis: est. 1/2 days
  • Create and test transpiler to latex for geometry and mesh, for automatic reporting: est. 2/3 days
  • Create and test transpiler to latex for mesh quality, for automatic reporting: est. 2/3 days
  • Fully working prototype in 2 to 4 weeks, plus 2 weeks to solve any unforeseen problem

Gantt Chart

Mechanics of Extreme Thin Composite Layers for Aerospace Applications Update 10-2016 Luca Di Stasio