What are the key steps in FEA modeling and meshing for Structural Engineering?

The first step in FEA modeling for structural engineering is defining the problem, which involves understanding the structure's purpose, the types of loads it will encounter, and environmental factors like seismic activity. This step is crucial as it guides the selection of the appropriate model type, materials, and boundary conditions. It's essential to have a clear, detailed understanding of the structural system and its operational environment to ensure the model accurately reflects real-world conditions and provides reliable results.

Often, FEA models are used to find an answer to a specific problem, things like peak displacements, anticipated settlements, ductility limits, etc. Understanding where the goalposts are before you begin can save some serious headaches. A problem well stated is a problem half solved. Specify the context and get agreement on the range of accuracy you need, these models become fractal wormholes when you are looking for extreme precision.

En el análisis por elementos finitos en ingeniería estructural, comprende a fondo el problema, usa software especializado, modela la estructura tridimensional, divide en elementos finitos, asigna propiedades y define condiciones de contorno. Aplica cargas con precisión, ajusta la densidad del mallado equilibrando precisión y eficiencia. Revisa la coherencia del modelo y su conformidad con leyes físicas. Ejecuta el análisis por elementos finitos, revisa e interpreta los resultados, asegurándote de la convergencia coherente. Si es necesario, optimiza el modelo mediante ajustes para mejorar la eficiencia o precisión del análisis. Este enfoque sistemático garantiza una evaluación efectiva de la estructura.

Boundary conditions and validation stand out as the most critical components in any FEA modeling, while many other essential components exist. Refinement of the original analysis is often necessary. This might involve updating the geometric model, modifying the material characteristics, fine-tuning or coarsening the mesh, or any combination thereof. The results must be validated against experimental data or theoretical predictions to ensure accuracy.

Finite Element Analysis (FEA) is a powerful tool in structural engineering used to simulate and analyze complex structures - Problem Definition: Clearly define the problem and establish the objectives of the analysis. Understand the structural behavior, loading conditions, material properties, and boundary conditions. - Geometry Creation: Create or import the geometric model of the structure in software capable of conducting FEA. This includes defining dimensions, shapes, and components accurately. - Idealization and Simplification: Divide the complex structure into smaller, simpler elements such as beams, plates, or solids. Simplify the geometry to model elements that can be analyzed effectively.

....applying loads and boundary conditions, preparing the deck for running the FEA (selecting the type of analysis, expected results, selection of solving methodology etc). 2) Processing: This is the stage where the FEA software runs the numerical analysis for obtaining the expected results. This is a back end process of the FEA software where different solvers are using for obtaining the required results. 3) Post-Processing: This is the last stage of the FEA where in the obtained results from the second stage (processing) are decorated and presented for a clear representation. These results could be stresses, strains, deformation which are decorated with different colored plots, animations, amplification of the results etc.

It’s crucial to simplify the FEA model while maintaining its integrity: 1. Remove minor details (small features, fillets, etc.) ditch those tiny details that don’t really change the game. 2. Utilize symmetry and submodeling to make the model less of a beast to handle. 3. Aim for a balance between detail and computational efficiency.

A simplifica??o do modelo pode ser necessária em algumas situa??es para tornar a análise mais eficiente, especialmente quando se lida com grandes modelos ou recursos computacionais limitados. No entanto, isso n?o significa que a simplifica??o extrema seja sempre a melhor escolha. Algumas nuances devem ser consideradas como precis?o, características críticas, condi??es de contorno, recursos computacionais disponíveis, valida??o experimental caso possível... dentre outros aspectos. Sugiro uma abordagem mais equilibrada nesse item visto que muitas vezes sequer poderemos quantificar o comprometimento dessas simplifica??es à precis?o e relevancia da análise sem fazer um comparativo com modelagens mais refinadas.

Simplify the model it is always a good idea in order to do not construct heavy model to analyze and difficult model to understand in term of result. To understand the first order of magnitude of the solution always run a linear elastic analysis first, then if necessary add the non linearity

Simplifying a FEA model involves reducing the complexity of the real-world structure into a manageable and computationally efficient model while retaining essential characteristics. This process includes omitting minor features that don't significantly affect the overall behavior under loads, such as small notches or fillets, and using symmetrical halves or quarters of a structure if applicable. The goal is to create a model that accurately represents the structural behavior under expected conditions without unnecessary details that could complicate the analysis and increase computational time.

Mesh generation is a pivotal step in FEA, where the model is divided into smaller, discrete elements. The quality and size of these elements directly impact the accuracy and computational efficiency of the analysis. A finer mesh provides more detailed results but requires more computational resources, while a coarser mesh reduces computational demands but may lead to less precise outcomes. The key is to balance mesh density, ensuring finer mesh in areas with high stress gradients or complex geometry, and coarser mesh in less critical regions. This step requires careful planning and expertise to optimize the mesh for accurate and efficient analysis.

Mesh generation in some problems can be much more important than some others. For example in a problem of plain strain or plane stress analysis, it doesn't matter so much to use triangular or tetrahedral with some notes and or at Boundary in the middle domain of the problem. But if the plane has a hole in it or we are analyzing growing cracks then we should make fine meshing with specific element types around the hole or around the crack. The importance of mesh generation is to reduce the error and computation expense as much as possible simultaneously.

FEA mesh size is generally governed by the thickness, geometry, and dimensions of the structure. The mesh plates shall be defined such that the structural behavior of the model shall replicate the actual behavior. The mesh shape also depends on the shape of the structure.

Sometimes multiple mesh types may be used to generate better results and yet consume less time and energy. - Finer mesh is required where it is necessary like around the holes where stress trajectory is quite close to each other - Or using smaller mesh for small element such as sensor attached to some beam or column with superstructure in more coarse mesh size

Seleccionar el tipo de elemento finito adecuado para representar la geometría de la estructura. Esto puede incluir elementos unidimensionales (vigas), bidimensionales (placas) o tridimensionales (sólidos), según la naturaleza del problema. Crear un modelo geométrico preciso de la estructura. Este modelo debe capturar con precisión la forma y dimensiones reales de la estructura. Dividir la estructura en elementos finitos. La calidad del mallado es crucial para obtener resultados precisos. Se deben considerar técnicas como la refinación del mallado en áreas críticas.

Applying loads and boundary conditions involves defining how the structure is supported and the forces it will experience. Loads can include weights, pressures, thermal effects, or seismic forces, depending on the structure's purpose and environment. Boundary conditions specify how the model is restrained, such as fixed, pinned, or roller supports, representing the real-world constraints of the structure. Accurately applying these factors is crucial as they significantly influence the model's response, ensuring the analysis reflects the true behavior of the structure under expected conditions. This step requires a deep understanding of both the structure's design and its operational environment.

Application of loads and CORRECT determination of the boundary conditions is most likely a matter of experience. All results are wrong or at least questionable if you do not have the correct loads and the adequate boundary conditions. The loading conditions are in my opinion in most cases simply too high considering that self-weights and gravity based loads underlying the rounding problem (10 instead of 9,81) In detail the boundary conditions are key to the output of the calculations. Fixed or pivoted or spring loaded makes the difference. And finally the possibility of application of the considered parameters. Often there is no possibility to realize a pivoted support or it will be cancelled due to cost reductions and /or restrictions.

While the long-term functioning of any structure involves time-varying applied loads, some, like the dynamic hydraulic loads from pumped hydro turbines acting on concrete structures, may be transient, occurring only once in a lifetime. Despite their brevity, typically lasting mere seconds compared to the 150-year design life, such loads necessitate meticulous attention to boundary conditions due to their substantial influence.

Loads are applied to the model to simulate the forces acting on the structure. These loads can be static or dynamic, and they can include forces due to gravity, wind, earthquakes, temperature changes, or other factors. Boundary conditions are constraints imposed on the structure, such as fixed supports, pinned connections, or loads applied at specific locations. These conditions help define the behavior of the structure and prevent rigid body motions.

Las condiciones de contorno requieren que se restrinjan algunos grados de libertad, representando asi los soportes de las estructuras. Si se restringen grados de libertad traslacionales y rotacionales se tienen un soporte empotrado, si se restringen solo los grados de traslación se tiene un soporte pivotante, usando estas combinaciones se logra modelar la gran mayoría de condiciones de contorno. Es necesario representar de manera adecuada las conexiones y juntas dentro de la estructura, ya que el variar el tipo de uni′on entre elementos puede influir de manera drástica en la rigidez del sistema, aparte de modificar el estado de tensiones y deformaciones en las proximidades del soporte.

Solving the model in FEA is where the computer calculates the structure's response to applied loads and boundary conditions. Using numerical methods, it computes displacements, stresses, and strains for each mesh element. This step's complexity depends on the model's detail and mesh precision, balancing accuracy with computational feasibility. The solver's output reveals key insights into structural behavior, identifying areas of high stress or significant deformation, crucial for evaluating the structure's safety and integrity.

La solution EF d’un problème de mécanique de solides et de structures est basée sur l’équation de rigidité d’un élément. Les étapes pour obtenir cette équation matricielle sont les mêmes pour tous les problèmes. Dans ce paragraphe nous allons détailler ces étapes pour la résolution d’une équation différentielle (problème 1 D) afin de maitriser la procédure.

Once the mesh and boundary conditions are defined, the FEA software solves the equations governing the behavior of the structure. This process involves iteratively updating the stresses, strains, and deflections of the elements until a stable solution is reached. The choice of solver depends on the type of problem, the complexity of the model, and the desired accuracy. Common solvers include the direct stiffness method, the iterative methods, and the boundary element method.

Após a fase de pré-processamento (modelo e defini??es), chegamos na fase do processamento, onde o softwa de FEA analisa todo modelo e distribui as cargas em for?as equivalentes nodais nos nós do modelo. Após tudo isto, o software utiliza um método de solu??o numérica, que depende da análise (estática, dinamica, n?o linear, etc), encontrando os deslocamentos nodais da estrutura. estes deslocamento s?o convertidos em esfor?os, considerando a diferen?a da posi??o indeformada para a deformada. Através da interpola??o das fun??es de forma é possível obter deslocamentos e esfor?os dentro do ambito do elemento, ou seja, entre os nós.

Algunas veces los modelos son relativamente peque?os o sencillos, y dependiendo del programa utilizado, puede tardar en resolver el análisis, lo cual puede ser causado por grandes desplazamiento o inestabilidad, la mayoría de las veces esto es por que algún elemento lo está debidamente conectado, de aquí la importancia de generar un modelo y su respectiva malla de manera correcta

The final step involves interpreting the results of the FEA analysis. This includes examining stress distributions, deflections, and other relevant parameters to ensure the structure meets the design requirements. Visualizations of the results, such as stress contours, deflection plots, and deformed shapes, can be helpful in identifying critical areas and understanding the overall behavior of the structure. FEA modeling and meshing are essential tools for structural engineers in designing and analyzing complex structures. By carefully following these steps and choosing appropriate modeling techniques, engineers can obtain accurate and reliable results that inform their design decisions and ensure the safety and integrity of structures.

Here you need the experience again. In principle you cannot trust your results as you knowingly have altered the basic problem to a computational problem. Basically you need to know the results in advance to analyze the computational results correctly.

One thing I've learned to appreciate is that the results need to be verified regardless of whatever software has been used. This helps pick out possible errors in loading/definitions.

Validation is an important step in FEA analysis. As my professor in the FEA class always said :" I can't stress enough the importance of checking your results with hand calcs. There is always a way to check by hand. If you can't find it. It simply means you don't know the problem well enough." I can't agree more with what he said. Having been in this industry for 6 years, "quick and dirty" hand calcs have saved me from numerous embarrassing silly mistakes.

This is the step of validation. A lot of mistakes could have been done unconsciously during the modelisation even if desgin codes limits validate them. The results should be compared to hand calculation models (1D calculations). The idea is to approach FEA results to hand calculation results. If these results don't match then the model has a mistake somewhere. In order to validate your model, a simple load case can prove it by checking its deformation shape. The deformation shape gives you an idea on how your model behave under simple loading. Next step could be the check of results such as stress or displacements, the most common results used for our design. Once this done, the model can be used for further complex loading.

FEA is an iterative process. The user runs a model, not once, but many times, going back into the process and refining all of the input parameters, then rerunning. The analysis is slowly built up over time until the engineer has full confidence in the results. Most important, the engineer should never put anything into a computer unless he or she knows what the answer should be before hitting, "Run."

Always perform visual inspections and a reality checks. Do the deflections make sense for the applied loads? Are the global load sums reasonable? Spin the model to see if there are any unusual visual discrepancies in the layout or stresses and deformations. Perform simple hand calculation estimates of key indicators and investigate any results that deviate significantly from expectations. Always have someone check your work.

In FEM analysis, collaboration among CAD engineers, meshing experts, material database owners, and solver specialists is crucial for success. To streamline the analysis workflow and prevent errors, a well-defined framework and close collaboration are essential. Mistakes, such as providing incorrect files or material properties, can lead to time-consuming setbacks. To address this, a low-code environment can be utilized to create templates that store the expertise of each team member. These templates enable seamless connection and adaptation to changes, fostering efficient collaboration and automation in the FEA process across projects, saving valuable time for iterations and decision-making.

Things that are very difficult to catch but can heavily affect your model and results: 1) Unconnected nodes: for larger models, your sequence of modelling is important to ensure every node is connected and every element is where it should be. 2) Material specifications: make sure you check your material specifications in every model. 3) Element releases and connections: pinned-fixed, fixed-fixed is your everyday decision that can make or break your model. 4) Load combinations: all of us have the load combinations excel sheet that we copy from, but are your loads correctly defined in the model to be captured by your combinations? 4) Rigid links: when used correctly, it can help you imitate real building behaviour very accurately.

Ensure to check that the model is in fact defined as intended by checking that the results are consistent with the boundry conditions and loads defined. For example: Pinned conditions do not have moment on them. Base shear for an input load meets hand calculated base shear