A. hexahedron element formulation with a new multi-resolution analysis

This paper presents a new multi-resolution hexahedron element formulation based on a novel framework of mutually nesting displacement subspaces, which allows for adjustable resolution levels to modulate structural analysis accuracy more efficiently and rationally than traditional mono-resolution methods.

The Gist

Imagine you are trying to take a photograph of a complex 3D object, like a bridge or a building, to see how it bends or vibrates under stress. In the world of engineering, this is done using a method called the Finite Element Method (FEM).

The Old Way: The Pixelated Puzzle

Traditionally, engineers break a structure down into a grid of small 3D blocks (like tiny dice). These are called "hexahedron elements."

• The Problem: Each block is "mono-resolution," meaning it has a fixed number of corners (nodes) where calculations happen. Usually, it's just 8 corners.
• The Limitation: If you want to see a tiny crack or a sharp bend in the material, the old method forces you to cut the whole structure into more and more tiny blocks. It's like trying to get a clearer photo by cutting your entire picture into millions of smaller, separate puzzle pieces and gluing them back together. This is slow, messy, and requires a lot of re-doing (re-meshing) every time you want more detail.

The New Way: The "Smart Zoom" Lens

The authors of this paper (Xia YiMing and Chen ShaoLin) have invented a new type of 3D block that acts like a smart camera lens with a zoom function. They call this a Multi-Resolution Hexahedron Element.

Here is how it works, using simple analogies:

1. The "Basic Shape" (The Master Template)
Instead of using a standard 8-corner block, they created a special "Basic Node Shape Function." Imagine taking a standard 8-corner block and stretching its influence out to cover a larger area, effectively creating a "master template" that knows about its neighbors. This template is built by extending the shape of one corner into the surrounding space, covering 27 potential spots (like a 3x3x3 cube of dots).

2. The "Resolution Level" (The Zoom Knob)
This is the magic part. The new element has a dial called the Resolution Level (RL).

• Low RL (Zoomed Out): The element acts like a traditional 8-node block. It's simple and fast.
• High RL (Zoomed In): You turn the dial, and the element automatically sprouts more internal "nodes" (calculation points) without changing the size of the block itself.
• The Analogy: Think of a digital photo. A low-resolution photo is blurry. A high-resolution photo is sharp. In this new method, you don't need to cut the photo into smaller pieces to make it sharper; you just turn up the "Resolution Level" on the single block, and it reveals more detail (more nodes) instantly.

3. The "Nested" Structure
The paper explains that these different levels of detail fit inside each other perfectly, like Russian nesting dolls. A high-detail version of the block contains the low-detail version inside it. This mathematical "nesting" ensures that the calculations remain stable and accurate as you zoom in.

Why is this better? (The Paper's Claims)

• No More Re-Meshing: In the old way, to get better accuracy, you had to tear the structure apart and rebuild the grid (re-meshing). In this new way, you just adjust the Resolution Level. It's like changing the focus on a camera rather than rebuilding the camera.
• Simplicity: The math behind the new shape is surprisingly simple. It keeps a special property called the "Kronecker delta" (which basically means "I know exactly where my corners are"), making it easy to calculate boundary conditions (like where a wall is fixed).
• Efficiency: Because you can get high accuracy with fewer blocks, the computer does less work. The paper claims this method is faster and more rational than traditional methods or other "multi-resolution" methods that use complex wavelets (which the authors say are too messy and hard to apply).

Real-World Tests (From the Paper)

The authors tested their "Smart Block" on three scenarios:

1. A Cantilever Beam: A beam sticking out from a wall. They showed that one of their "zoomable" blocks could match the accuracy of dozens of traditional blocks.
2. A Square Plate: A flat slab. They compared it to a popular "Wavelet" method and found their method was easier to use and just as accurate.
3. A Folded Plate: A complex, bent structure. They showed that they could adjust the detail level in different parts of the structure easily, whereas the old method would require a massive, complex grid.

The Bottom Line

The paper argues that this new Multi-Resolution Hexahedron Element is a superior tool for structural analysis. It treats the "Resolution Level" as the key to accuracy, not the number of mesh pieces. It claims to be more rational, easier to implement, and more efficient than the traditional 8-node element or other advanced methods, making it ideal for solving complex engineering problems where details matter (like cracks or sharp stress points).

Note: The paper mentions future work will focus on how to connect blocks with different resolution levels (like connecting a zoomed-in photo to a zoomed-out one), but the current results focus on the formulation and testing of the single element type.

Technical Summary

Technical Summary: A Hexahedron Element Formulation with a New Multi-Resolution Analysis

Problem Statement
The paper addresses limitations in traditional Finite Element Methods (FEM) and existing Multi-Resolution Analysis (MRA) techniques within computational mechanics. Traditional FEM is characterized as "mono-resolution," where a single element contains a fixed number of nodes. Consequently, improving analysis accuracy requires artificial meshing and re-meshing, a process the authors argue is irrational, inefficient, and poorly suited for problems with local steep gradients (e.g., material nonlinearity, cracks, or impacts). Conversely, existing MRA methods, such as the Wavelet Finite Element Method (WFEM), Meshfree Methods (MFM), and Natural Element Method (NEM), suffer from inherent deficiencies. These include the complexity of constructing node shape functions, the absence of the Kronecker delta property (complicating boundary condition treatment), and a lack of a solid mathematical basis for MRA, which often renders these methods empirical and difficult to apply in engineering practice.

Methodology
The authors propose a new multi-resolution hexahedron element formulation based on a novel MRA framework. The core methodology involves the following steps:

1. Construction of the Basic Node Shape Function: Starting with the standard isoparametric shape function of a traditional 8-node hexahedron element defined on the domain $[0,1]^3$, the authors extend this function to the domain $[-1,1]^3$. This is achieved by shifting the isoparametric element around a specific node (located at the origin) into the other seven quadrants. This extension covers 26 adjacent nodes and creates a "basic node shape function" ($\varphi$) that is continuous and possesses the Kronecker delta property.
2. Formation of a Mutually Nesting Displacement Subspace Sequence: The MRA framework is constructed using a sequence of displacement subspaces. The basis functions for these subspaces are generated by scaling and shifting the basic node shape function $\varphi$.
   • Scaling: The function is scaled by integer parameters ($m, n, l$) to control the resolution level (RL).
   • Shifting: The scaled function is shifted to different node positions within the element domain.
   • This process creates a sequence of subspaces ($V_{ijk}$) where $V_{ijk} \subset V_{(i+1)(j+1)(k+1)}$, establishing a mathematically rigorous, mutually nesting structure without relying on wavelet basis functions.
3. Element Formulation: The displacement field within an element is defined by the resolution level (RL), denoted as $(m+1) \times (n+1) \times (l+1)$. The total number of nodes is determined directly by the RL, not by the mesh density. The element stiffness matrix, mass matrix, and loading vectors are derived using the principle of minimum potential energy, utilizing the constructed shape functions.
4. Implementation: The method allows for the automatic generation of element matrices around nodes. A transformation matrix is used to convert local element matrices to the global structural coordinate system.

Key Contributions

• Novel MRA Framework: The paper introduces a solid mathematical basis for MRA in solid mechanics that does not require wavelet basis functions, relying instead on the scaling and shifting of extended isoparametric shape functions.
• Adjustable Resolution Level (RL): The proposed element possesses an adjustable RL that modulates the element node number. This allows for the direct control of analysis accuracy within a single element, contrasting with the traditional requirement of re-meshing.
• Kronecker Delta Property: Unlike many existing MRA methods, the proposed shape functions retain the Kronecker delta property, simplifying the imposition of boundary conditions.
• Unification of Mono- and Multi-Resolution: The authors demonstrate that the traditional 8-node hexahedron element is a special case (mono-resolution) of the proposed multi-resolution element (specifically when $m=n=l=1$).

Results
The paper validates the method through three numerical examples:

1. Cantilever Beam: Under uniform transverse loading, the proposed element with a single element and varying RL achieved tip deflection results comparable to traditional elements with significantly higher mesh densities (e.g., $1 \times 1 \times 30$ mesh). The results converged toward analytical values as the RL increased.
2. Simply Supported Square Plate: The proposed element was compared against traditional 8-node elements and a 2D B-spline Wavelet Interpolation (BSWI) element. The proposed method demonstrated that increasing the RL improved accuracy similarly to increasing the order in BSWI, but with easier implementation. The authors note that 3D BSWI elements are often too complex due to the irrational increase in degrees of freedom (DOF) and lack of isoparametric formulation.
3. Folded Square Plate: The method was applied to a complex geometry with free and simply supported boundaries. The analysis showed that adjusting the RL of specific elements allowed for accurate modeling of displacement responses with fewer total elements and fewer transformation matrix multiplications compared to traditional re-meshing. The results were consistent with literature data.

Significance and Claims
The authors claim that the proposed multi-resolution hexahedron element method is more rational, easier to implement, and computationally more efficient than both traditional mono-resolution FEM and other existing MRA methods.

• Efficiency: By avoiding complex re-meshing and the artificial assembly of mono-resolution elements, the method reduces computational overhead. The automatic acquisition of matrices around nodes further enhances efficiency.
• Accuracy Control: The paper asserts that structural analysis accuracy is determined by the Resolution Level (RL), not the mesh. This allows for convenient adaptive analysis where accuracy is modulated by adjusting the RL rather than rebuilding the mesh.
• Applicability: The method is presented as particularly well-suited for structural problems involving local steep gradients, such as cracks or impacts, where high local resolution is required without globally refining the mesh.
• Future Work: The authors modestly note that future work will focus on treating the interfaces between multi-resolution elements of different RLs, potentially using bridging domains or expanded Serendipity elements.

In conclusion, the paper posits that this new MRA framework bridges the gap between traditional FEM and advanced multi-resolution techniques, offering a robust tool for accurate structural computation that overcomes the limitations of current methods.