Computational mechanics and applications


Here you can find information about the research projects conducted by Prof. Alfredo Gay Neto, co-workers and supervised students at the University of São Paulo, Brazil.

Please, scroll down to see details of research areas and specific projects.

You can contact Prof. Alfredo Gay Neto directly by e-mail at:

Research areas

New formulations on computational contact mechanics


The objective is to develop new mathematical formulations to consider contact interaction between bodies. On the finite element modeling context, we have already developed models for beam-to-beam and beam-to-shell interactions.

We also develop models for contact interaction of rigid-rigid bodies and rigid-flexible bodies. An in-house finite element platform (Giraffe) comprises all our contact models.

Railway engineering modeling


Here one ot the targets is to develop our own wheel-rail contact models. The interest is in studying the influence of geometric profiles of wheel-rail on contact bahavior and its consequences, such as crack initiation. Computational tools are continuously developed on this study.

We are also studying mechanical behavior of railway ballast. For that, a discrete element model is being used.

Offshore engineering modeling


We investigate the global mechanical behavior of offshore risers (for oil exploitation). Static and dynamic analysis are performed, considering the floating unit movements, sea current loads and the contact between the risers and the seabed. Moreover, we have developed a tool for analysis of an entire mooring system, which is under continuous updating.

Another topic of interest is the local mechanical behavior of flexible pipes. For that, we have experience on collapse and burst models, using ANSYS software.

Wind turbine modeling


We are investigating the mechanical behavior of wind turbines. The wind turbine blades and the tower are analyzed employing quasi-static or dynamic time domain models. We also make models for the whole system considering finite element beams together with rigid bodies, composing a rigid-flexible multibody model. Simulations are done using Giraffe platform.

We are interested on onshore and offshore (floating) wind turbines. Ongoing works are aeroelasticity studies of wind turbine blades and study of composite material failures.

Giraffe platform

Giraffe is the acronym of “Generic Interface Readily Accessible for Finite Elements”. It is a platform coded using C++ language, with the objective of generating a base-interface to be used by researchers to implement their own finite element formulations. “Giraffe Project” was started on 2014 by Prof. Alfredo Gay Neto, at the University of São Paulo, Brazil.

Giraffe platform was thought for geometrically-nonlinear applications. We have already implemented, among others, the following resources:

  • Geometrically-exact beam element
  • Geometrically-exact shell element
  • Geometrically-exact truss and spring elements
  • Rigid body element
  • Hinge, universal and spherical joints
  • Node-surface contact formulation
  • Master-master contact formulation (between surfaces and their degenerations)

Useful links for Giraffe software:

About me

I am a Mechanical Engineer at Polytechnic School at the University of São paulo (2006) and Doctor of Sciences at the same institution (2012). I developed a post-doc at the same institution (2012-2014), with a part-time at Leibniz Universität-Hannover, Germany (2013). My research is based on developing computational mechanics models and applying them to practical problems. My main background is related to contact problems modeling.

Since 2014 I work at the University of São Paulo, as Assistant Professor. On 2018 I became Associate Professor at the same institution, when I defended my Habilitation Thesis.

In this website I show an overview of my areas of interest on research, such as some projects. In case of interest in some topic, please contact me by e-mail:

Useful links for academic information:

Supervised students

Celso Jaco Faccio Júnior

PhD Student

Mesoscale modeling of textile composite mechanics

The modeling of textile composite mechanics is not a trivial task. The nonlinear mechanical behavior of this kind of material comes not only from the intricate geometry (weaving pattern, yarns cross-section and crimp), but also from the complex contact interaction between yarns. To model these materials a strategy that combines the geometrically-exact beam theory and the master-master contact approach is adopted. Preliminary results show that this strategy can reasonable reproduce experimental results and give valuable information regarding textile mechanics.


Gabriel Vicentin Pereira Lapa

MSc Student

Global Modeling and Dynamic Analysis of Wind Turbines

Offshore wind turbines have a complex dynamic behavior. In order to analyze them, we are creating a computational implementation in the Giraffe software, such that it is possible to calculate the aerodynamic loads on the turbine blades. This will make possible to carry out aeroelastic simulations. In the end, it is desired to perform multi-body dynamics simulations to investigate the overall turbine behavior in different scenarios.


Guilherme Rocha Martins

Pre-MSc Student

Numerical analysis of the mooring system of a floating offshore wind turbine (FOWT)

The effects of the mooring system on floating offshore wind turbine (FOWT) are still under investigation, especially in deep waters. This research aims to develop a tool to model, simulate and analyze these systems. It is a preprocessor to Giraffe solver, named "GiraffeMoor". This software generates an input file for Giraffe with a reduced set of geometric characteristics, line materials and environment characteristics. GiraffeMoor can also run Giraffe directly in the same window and provides the user a guide for interpreting the generated monitors and post files. The current goal is to couple another code with Giraffe+GiraffeMoor in order to compute the hydrodynamic effects of the floating unit, to gradually move to a coupled analysis.


Hélio Marques Neto

Pre-MSc Student

Finite element simulations of wheel squeal phenomenon

In order to analyze the wheel-rail contact problem and obtain relevant information such as the stress field, computational quasi-static models have been developed to represent the twin disc test, which in turn represents well the phenomenon that occurs in the real field. With the acquired knowledge we are developing transient dynamic models of contact between discs in which there is a small angle of attack between them to produce the phenomenon called squeal, a recurring noise problem of railways that affects the health and well-being of everyone involved in transportation or surroundings of railways.


Isadora Toledo de Almeida

MSc Student

Mechanical Analysis of Laminated Composite for Offshore Wind Turbines

There are still gaps in the comprehension of the failure mechanisms of laminated composite materiais due to the lack of structural models on micro and meso-scale, so that phenomenological models are used in their prediction. Particularly, the impact damage is a major issue in the design of these structures, requiring the comprehension of how the resulting damage affects residual strength. This work proposes a preliminary analysis of the loads acting on a wind turbine blade using a low hierarchy numerical model that aims to obtain local stress over post-processing based on the Classical Lamination Theory (CLT), then to characterize the extension of damage in laminated composite structures through a conservative modelling of quasi-static impact, considering inertial loads and laminate degradation.


Lucas da Silva

MSc Student

Modelling the wheel-rail contact with geometric representation by NURBS

In this project we tackle the problem of geometric reconstruction of wheel and rail profiles given a sequence of points sampled from these elements. A curve composed of smoothly joined circular arcs approximates the original geometry of the profiles. This type of curve is called arc spline, it has a well known parametrization using rational B-splines (NURBS). A least-squares fitting procedure was developed to find the approximation curve. The obtained curves are used in curve-to-curve contact algorithms to represent the wheel-rail interaction in 2D simulations, and also in the construction of surfaces for 3D simulations with surface-to-surface contact algorithms.


Marina Vendl Craveiro

PhD Student

Numerical modeling of track ballast and its application to study the mechanical behavior of railroads

The research aims at developing a new algorithm for the contact between particles modeled as Non Uniform Rational Basis Spline (NURBS). The idea is that the new algorithm allows multiple points of contact interaction and therefore can be applied in the study of track ballast. As many particles are necessary for modeling the ballast, besides being robust, the algorithm also needs to be fast and efficient, making numerical simulation feasible. With that in mind, the research intends to explore concepts and methods from computer graphics, which are very efficient, but little used in the scope of computational mechanics for engineering. The resulting algorithm will be applied to develop coupled models with ballast, rails and sleepers, allowing to evaluate the mechanical behavior of railroads.


Ricardo Nunhez da Silva

MSc Student

Numerical modeling of the twin-disc test for elastic stress field analysis in rolling scenarios

In this project the twin-disc test is reproduced in a numerical approach. This test is commonly used by experimentalists to understand contact phenomena in rolling scenarios, such as friction wear and rolling contact fatigue-RCF. The aims of the project are: i) to explore and expose the theorical concepts related to the mechanics of rolling contact such that the twin-disc test is consistently explained, involving concepts of creep, shift and rigid slip; ii) to simulate numerically the twin-disc test and develop a numerical methodology for creep curve plotting by means a Finite Element method-FEM. The analysis of the elastic stress field that is captured by the FEM model helps to predict plastification candidate points in the bodies and also potential crack nucleation spots. The whole work will provide a conceptual bridge between the physical test approach and the simulation of contact mechanics.


Silvio Tumelero de Moraes

MSc Student

Dynamic analysis of railway ballast behavior by computational modeling

Railways are essential for long distance transportation and for carrying a high amount of cargo. Indeed, to ensure safety, periodic maintenance on track superstructure components is required. The tamping machine and the dynamic stabilization equipment are essential to this process. Nonetheless, there is a great amount of variability and uncertainty between the standards adopted and suggested. In this work, it is intended to develop a representative computational model of railway ballast, describing its dynamic behavior, in order to evaluate and optimize the use of these equipments, using the Discrete Element Method by Rocky DEM software.


Tiago Fernandes Moherdaui

MSc Student

The Virtual Element Method for wheel-rail contact problems

The virtual element method was first published in 2012, and has ever since been explored for a wide range of applications. The method consists of a generalization of the finite element method, which allows elements of general shape. This characteristic has been motivating applications in the field of computational contact mechanics. This work seeks to explore this method to evaluate the stress fields arising from the contact between wheel and rail, which is important for railway engineering. This is going to be achieved by means of successive intermediate applications for classical problems such as St. Venant’s torsion theory, plane linear elasticity, plane nonlinear elasticity, and Hertz’s contact theory.

Journal publications (links to full texts)