Carlos Felipe Guzmán – CURRENT PROJECTS

As main researcher

2020–2022 Numerical and experimental characterization of the quasi-brittle behavior of timber

  • Project name: Fondecyt de iniciación en investigación Nº11200638
  • Funding: Agencia Nacional de Investigación y Desarrollo (ANID), Chile
  • Sponsoring institution: Universidad de Santiago de Chile
  • Role: Main researcher
  • Collaborators: None
  • Description: In order to increase the competitiveness of timber as a building material, the ductile behavior should be understood and exploited as much as possible. This phenomenon is mathematically complex and demands high accuracy to the numerical method used to solve the equations. Hence, the interest of this project to develop a numerical tool able to describe the non-linearities in the microscale and their effects on the macroscopic quasi-brittle behavior. This problem would also need the implementation of a mixed finite element formulation which, if correctly stabilized, can lead to high stress accuracy for localization problems and locking-free solutions. Together with this, an experimental campaign is proposed to characterize the macroscopic behavior of radiata pine. Considering the current trends on timber applications, most of them based on rules which overestimate certain properties due the complex behavior of timber, the proposed research can have an important impact in developing numerical methods for engineering applications.

2020 – Caracterización numérica para interacción entre madera y uniones en estructuras

Project name:  Dicyt regular 052018GI

  • Funding: Universidad de Santiago de Chile
  • Sponsoring institution: Universidad de Santiago de Chile
  • Role: Main researcher
  • Collaborators: Erick I. Saavedra Flores, Sergio J. Yanez
  • Description: Timber usually has a quasi-brittle behavior characterized by a strong anisotropy according to the grain orientation. In conjunction with steel joints, the timber structure can show a ductile behavior, which is highly desirable when subjected to cyclic loads. It is proposed a numerical framework to simulate a timber structure and to predict the ductile behavior of the structure under different loads, using computational homogenization for timber and phenomenological models for the steel conector.  the main objective is to replace this experimental methodology by a numerical framework, in order to obtain new types of hysteretic curves that will help to implement finite elements. This can be used to design connectors within a building with timber structures.

 

As collaborator

2018-2022 Identification of material parameters by Single Point Incremental Forming process

  • Project name: MATSPIF-ID
  • Funding:  Fond de la Recherche Scientifique FNRS, Belgium
  • Sponsoring institution: Université de Liège, Belgium
  • Main researcher: Anne Marie Habraken
  • Role: Collaborator
  • Description: Industry relies on simulations to avoid trials and errors in the development of new products and to enhance product quality. Finite Element (FE) simulation is now a usual tool to answer industrial need, however a bottle neck is often the lack of accurate material data. Material data base are generally incomplete and often confidential. The production community waits from academic word to be inventive to solve this data problem. Indeed multi-scale simulations offer solution claiming that, once microstructure is identified, one does not need multiple tests to feed phenomenological models. In fact, it shifts the problem to microstructure identification and it brings CPU problems as FE² methods are very greedy. The classical approach relies on multiple types of tests: bulge, bi-axial, cyclic shear, Marciniak or Nakazyma for Forming Limit Diagram, cup drawing, hole expansion … to investigate the stress and strain states and identify constitutive laws such as anisotropic elasto-visco-plastic law coupled with damage. Multiple difficulties arises: these tests require sample cutting in addition to the technician time, one lab does not gather the required equipment which means delays coordination… This project aims to identify all the material data by the use of a single equipment: a Single Point Incremental Forming platform. The number of tests must be optimized: a minimal number of shapes to provide data allowing to accurately get a single material parameter set. It needs a careful computation of the sensitivity matrix to generate accurate data, to avoid local minimum of the functional expressing the difference between simulations and experiments. Both force and displacement fields must be involved in this functional. To validate the development, the project will be focused on a high strength steel with damage problems and on a titanium alloy with anisotropy and strain rate sensitivity issues.

2016-2019 Advanced Modelling of Ductility and Damage in Mass Timber Structures by Computational Homogenisation, FONDECYT regular 1160691

  • Funding: Agencia Nacional de Investigación y Desarrollo, Chile
  • Sponsoring institution: Universidad de Santiago de Chile
  • Role: Collaborator
  • Main researcher: Erick I. Saavedra Flores
  • More information

2016 – 2018 Fortalecimiento de la investigación en ingeniería a través de la adquisición de una mesa vibradora para el estudio del comportamiento sísmico y vibraciones de estructuras de gran escala,

  • Project name: Fondequip EQM160124
  • Funding: Agencia Nacional de Investigación y Desarrollo, Chile
  • Sponsoring institution: Universidad de Santiago de Chile
  • Role: Collaborator
  • Main researcher: Erick I. Saavedra Flores
  • Description: This project funds the acquisition and implementation of a shaking table to study the vibrations and seismic behavior on large scale structures. It consists in a 3m x 3m multiaxial shaking table, with nearly 20ton static capacity, 15 dynamic capacity and 3 D.O.F. in plane (X, Y translations and in plane rotations). With these characteristics, it will be the largest seismic table in Chile and South America.
  • More information