| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 5.15 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
As juntas adesivas ligadas por adesivos estruturais são frequentemente analisadas usando modelos
de materiais elásticos e elástico-plásticos. Atualmente, os adesivos flexíveis são utilizados em
aplicações estruturais. Os modelos de materiais utilizados para adesivos estruturais nem sempre
são adequados para adesivos flexíveis. Modelos de materiais hiperelásticos têm se mostrado
adequados para modelar adesivos flexíveis. O Método dos Elementos Finitos (MEF) é uma técnica
bem aceita para modelar e analisar juntas adesivas. Softwares comerciais de MEF já possuem
pacotes hiperelásticos, mas as aplicações, deste modelo material, a juntas adesivas é ainda pouco
estudado. Recentemente, métodos sem malha foram introduzidos nesse campo. Embora, a
natureza dos métodos sem malha os torne atrativos para a análise de adesivos flexíveis e das juntas
adesivas ligadas por eles, não existem este tipo de análises na literatura. Vários modelos
hiperelásticos estão disponíveis. Um deles, o modelo de Ogden, pode representar vários modelos
por meio da conversão de seus parâmetros. Portanto, a implementação do modelo de Ogden é
mais versátil.
Neste trabalho, os adesivos estudados foram modelados pelo modelo hiperelástico de Ogden, e
comparado com Natural Neighbour Radial Point Interpolation Method (NNRPIM), para analisar
juntas de sobreposição simples (JSS) com diferentes comprimentos de sobreposição (𝐿0). A
caracterização das constantes hiperelásticas foi feita através de dados experimentais de provetes
maciços dos adesivos SikaForce® 7752 e Araldite® 2015. Realizaram-se análises numéricas de JSS
ligadas com adesivos flexíveis através MEF e o NNRPIM, ambos considerando elasticidade linear.
Como ambos os casos de elasticidade parecem concordar com a forma da distribuição de tensões,
eles foram comparados.
Com este trabalho, foi possível concluir que o modelo de material hiperelástico é mais adequado
para analisar adesivos flexíveis. O modelo de material proposto prevê corretamente o
comportamento da junta adesiva e mostrou aumentar a capacidade de a junta suportar carga. Além
disso, ele representa melhor o efeito da ductilidade do material e o aumento do 𝐿0 corresponde a
um aumento no pico de tensão, mas, ao contrário do modelo elástico linear, não é linearmente
proporcional.
Structural adhesives and the joints involving them are often analyzed using elastic and elastic plastic material models. Nowadays, flexible adhesives are being used for structural applications. The material models used for structural adhesives are not always suitable for flexible adhesives. Hyper-elastic material models had been proven more suitable to model flexible adhesives. The Finite Element Method (FEM) is a well-accepted technique to model and analyze adhesive joints. Commercial FEM packages already have hyper-elastic capabilities, but not many applications to adhesive joints had been studied. Recently, Meshless Methods (MM) have also been brought into this field. Although the nature of MM made them attractive for the analysis of flexible adhesives and the joints bonded with them, no previous analyses had been reported. Several hyper-elastic models are available. One of them, the Ogden model, can represent several models through converting their parameters. Therefore, the Ogden model implementation is more versatile. In this work, the studied adhesives were modeled using the hyperelastic Ogden model and compared to the Natural Neighbor Radial Point Interpolation Method (NNRPIM) to analyze single lap joints (SLJ) with different overlap lengths (L0). The characterization of hyperelastic constants was carried out using experimental data from bulk specimens of SikaForce® 7752 and Araldite® 2015 adhesives. Numerical analyses of SLJ bonded with flexible adhesives were conducted using both the Finite Element Method (FEM) and NNRPIM, considering linear elasticity in both cases. Since both elasticity scenarios seem to agree with the stress distribution shape, they were compared. With this work it was possible to conclude that the hyperelastic material model is better suited to analyze flexible adhesive. The proposed material model correctly predicts the behavior of the adhesive joint, and it shown to increase the capacity of the joint to sustain load. Also, it represents better the effect of the ductility of the material and the increasing of 𝐿0 corresponds to and increased peak stress but contrary to the linear elastic model, it is nonlinear proportional.
Structural adhesives and the joints involving them are often analyzed using elastic and elastic plastic material models. Nowadays, flexible adhesives are being used for structural applications. The material models used for structural adhesives are not always suitable for flexible adhesives. Hyper-elastic material models had been proven more suitable to model flexible adhesives. The Finite Element Method (FEM) is a well-accepted technique to model and analyze adhesive joints. Commercial FEM packages already have hyper-elastic capabilities, but not many applications to adhesive joints had been studied. Recently, Meshless Methods (MM) have also been brought into this field. Although the nature of MM made them attractive for the analysis of flexible adhesives and the joints bonded with them, no previous analyses had been reported. Several hyper-elastic models are available. One of them, the Ogden model, can represent several models through converting their parameters. Therefore, the Ogden model implementation is more versatile. In this work, the studied adhesives were modeled using the hyperelastic Ogden model and compared to the Natural Neighbor Radial Point Interpolation Method (NNRPIM) to analyze single lap joints (SLJ) with different overlap lengths (L0). The characterization of hyperelastic constants was carried out using experimental data from bulk specimens of SikaForce® 7752 and Araldite® 2015 adhesives. Numerical analyses of SLJ bonded with flexible adhesives were conducted using both the Finite Element Method (FEM) and NNRPIM, considering linear elasticity in both cases. Since both elasticity scenarios seem to agree with the stress distribution shape, they were compared. With this work it was possible to conclude that the hyperelastic material model is better suited to analyze flexible adhesive. The proposed material model correctly predicts the behavior of the adhesive joint, and it shown to increase the capacity of the joint to sustain load. Also, it represents better the effect of the ductility of the material and the increasing of 𝐿0 corresponds to and increased peak stress but contrary to the linear elastic model, it is nonlinear proportional.
Description
Keywords
Meshless methods Hyperelastic Flexible adhesives NNRPIM FEM