論文の概要: Deep learning for the rare-event rational design of 3D printed
multi-material mechanical metamaterials
- arxiv url: http://arxiv.org/abs/2204.01769v1
- Date: Mon, 4 Apr 2022 18:04:23 GMT
- ステータス: 処理完了
- システム内更新日: 2022-04-06 14:44:25.768383
- Title: Deep learning for the rare-event rational design of 3D printed
multi-material mechanical metamaterials
- Title(参考訳): 3Dプリントメカニカルメタマテリアルの希少合理的設計のための深層学習
- Authors: H. Pahlavani, M. Amani, M. Cruz Sald\'ivar, J. Zhoua, M. J. Mirzaali,
A. A. Zadpoor
- Abstract要約: マルチマテリアル3Dプリンティング技術は、メタマテリアルの合理的設計の道を開いた。
ネットワークの結果として生じる異方性力学特性と,特に稀な設計について検討する。
ディープラーニングに基づくアルゴリズムは、異なる設計の機械的特性を正確に予測することができる。
- 参考スコア(独自算出の注目度): 0.0
- License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
- Abstract: Emerging multi-material 3D printing techniques have paved the way for the
rational design of metamaterials with not only complex geometries but also
arbitrary distributions of multiple materials within those geometries. Varying
the spatial distribution of multiple materials gives rise to many interesting
and potentially unique combinations of anisotropic elastic properties. While
the availability of a design approach to cover a large portion of all possible
combinations of elastic properties is interesting in itself, it is even more
important to find the extremely rare designs that lead to highly unusual
combinations of material properties (e.g., double-auxeticity and high elastic
moduli). Here, we used a random distribution of a hard phase and a soft phase
within a regular lattice to study the resulting anisotropic mechanical
properties of the network in general and the abovementioned rare designs in
particular. The primary challenge to take up concerns the huge number of design
parameters and the extreme rarity of such designs. We, therefore, used
computational models and deep learning algorithms to create a mapping from the
space of design parameters to the space of mechanical properties, thereby (i)
reducing the computational time required for evaluating each designand (ii)
making the process of evaluating the different designs highly parallelizable.
Furthermore, we selected ten designs to be fabricated using polyjet
multi-material 3D printing techniques, mechanically tested them, and
characterized their behavior using digital image correlation (DIC, 3 designs)
to validate the accuracy of our computational models. The results of our
simulations show that deep learning-based algorithms can accurately predict the
mechanical properties of the different designs, which match the various
deformation mechanisms observed in the experiments.
- Abstract(参考訳): マルチマテリアル3Dプリンティング技術は、複雑なジオメトリだけでなく、それらのジオメトリ内の複数のマテリアルの任意の分布を持つメタマテリアルの合理的設計の道を開いた。
複数の材料の空間分布の変化は、異方性弾性特性の興味深い、潜在的に独特な組み合わせをもたらす。
弾性特性の可能な組み合わせの大部分をカバーする設計手法が利用可能であることは、それ自体が興味深いが、非常に稀な材料特性(二重運動性や高弾性変調など)の組み合わせにつながる設計を見つけることがより重要である。
そこで本研究では,規則格子内の硬相と軟相のランダム分布を用いて,ネットワーク全体の異方性力学特性と,特に上述した稀な設計について検討した。
主要な課題は、膨大な数の設計パラメータと、そのような設計の極めて希少性に関するものである。
そこで我々は,設計パラメータの空間から機械的特性の空間へのマッピングを作成するために,計算モデルとディープラーニングアルゴリズムを用いた。
一 各指定書の評価に要する計算時間を短縮すること
(ii)異なる意匠を評価する工程を高度に並列化すること。
さらに,ポリジェット多材料3Dプリンティング技術を用いて製造する10種類のデザインを選択し,その特性をデジタル画像相関 (DIC, 3デザイン) を用いて評価し,計算モデルの精度を検証した。
その結果,深層学習に基づくアルゴリズムは,実験で観測された変形機構と一致する異なる設計の力学特性を正確に予測できることがわかった。
関連論文リスト
- Geometry Distributions [51.4061133324376]
本稿では,分布として幾何学をモデル化する新しい幾何学的データ表現を提案する。
提案手法では,新しいネットワークアーキテクチャを用いた拡散モデルを用いて表面点分布の学習を行う。
本研究では,多種多様な対象に対して質的かつ定量的に表現を評価し,その有効性を実証した。
論文 参考訳(メタデータ) (2024-11-25T04:06:48Z) - Data-Driven Nonlinear Deformation Design of 3D-Printable Shells [1.5088726951324294]
本研究では,3次元プリント可能なシェルの設計・性能関係と圧縮力変位挙動を学習するための実験データに基づくニューラルネットワークを提案する。
何千もの物理的実験に基づいて、我々のネットワークは前方および逆設計の両方を補助し、所望の弾塑性および超弾性変形を示す殻を生成する。
論文 参考訳(メタデータ) (2024-08-27T14:30:06Z) - Porosity and topological properties of triply periodic minimal surfaces [0.0]
3つの周期的最小面 (TPMS) は、その構造的効率と制御可能な幾何学のために大きな関心を集めている。
本稿では, ポロシティとエントロピーとTPMSの形状因子との関係について検討する。
論文 参考訳(メタデータ) (2024-06-23T20:46:19Z) - Machine Learning-Guided Design of Non-Reciprocal and Asymmetric Elastic Chiral Metamaterials [0.0]
本稿では, 靭帯接触角, 靭帯形状, 円半径など, キラルなメタマテリアルの設計空間を定義した。
次に、機械学習アプローチ、特にベイズ最適化を活用して、最大非相互性あるいは剛性非対称性を満たす最適な設計を行う。
この機構を解析したところ, 異なる方向の荷重下で複数の異なる接触状態を示すことができるキラルなメタマテリアルは, 高い非相反性と剛性非対称性の両方を同時に示すことができることがわかった。
論文 参考訳(メタデータ) (2024-04-19T23:39:56Z) - Differentiable graph-structured models for inverse design of lattice
materials [0.0]
異なる環境条件に適応可能な物理化学的性質を有する建築材料は、材料科学の破壊的な新しい領域を具現化している。
正規および不規則な格子材料に対するグラフベース表現を用いた新しい計算手法を提案する。
論文 参考訳(メタデータ) (2023-04-11T18:00:21Z) - A Pareto-optimal compositional energy-based model for sampling and
optimization of protein sequences [55.25331349436895]
深層生成モデルは、生命科学における逆問題に対する一般的な機械学習ベースのアプローチとして登場した。
これらの問題は、データ分布の学習に加えて、興味のある複数の特性を満たす新しい設計をサンプリングする必要があることが多い。
論文 参考訳(メタデータ) (2022-10-19T19:04:45Z) - Symmetry Group Equivariant Architectures for Physics [52.784926970374556]
機械学習の分野では、対称性に対する認識が目覚ましいパフォーマンスのブレークスルーを引き起こしている。
物理学のコミュニティと、より広い機械学習のコミュニティの両方に、理解すべきことがたくさんある、と私たちは主張する。
論文 参考訳(メタデータ) (2022-03-11T18:27:04Z) - How to See Hidden Patterns in Metamaterials with Interpretable Machine
Learning [82.67551367327634]
我々は,材料単位セルのパターンを見つけるための,解釈可能な多分解能機械学習フレームワークを開発した。
具体的には、形状周波数特徴と単位セルテンプレートと呼ばれるメタマテリアルの2つの新しい解釈可能な表現を提案する。
論文 参考訳(メタデータ) (2021-11-10T21:19:02Z) - A deep learning driven pseudospectral PCE based FFT homogenization
algorithm for complex microstructures [68.8204255655161]
提案手法は,従来の手法よりも高速に評価できる一方で,興味の中心モーメントを予測できることを示す。
提案手法は,従来の手法よりも高速に評価できると同時に,興味の中心モーメントを予測できることを示す。
論文 参考訳(メタデータ) (2021-10-26T07:02:14Z) - GeoMol: Torsional Geometric Generation of Molecular 3D Conformer
Ensembles [60.12186997181117]
分子グラフからの分子の3Dコンホメーラーアンサンブルの予測は、化学情報学と薬物発見の領域において重要な役割を担っている。
既存の生成モデルは、重要な分子幾何学的要素のモデリングの欠如を含むいくつかの欠点がある。
エンド・ツー・エンド、非自己回帰、SE(3)不変の機械学習手法であるGeoMolを提案し、3Dコンバータを生成する。
論文 参考訳(メタデータ) (2021-06-08T14:17:59Z)
関連論文リストは本サイト内にある論文のタイトル・アブストラクトから自動的に作成しています。
指定された論文の情報です。
本サイトの運営者は本サイト(すべての情報・翻訳含む)の品質を保証せず、本サイト(すべての情報・翻訳含む)を使用して発生したあらゆる結果について一切の責任を負いません。