Journal article
Learning to program dimple-encoded strips for morphing architectures
- Abstract:
- Programmable morphing structures that translate local mechanical states into global shape transformations enable deployable, adaptive, and reconfigurable systems. Dimple-encoded elastic strips provide a minimal platform in which arrays of bistable dimples can collectively program large bending deformation. Yet the nonlinear mapping from the encoded geometry to the final configuration makes inverse design challenging. Here, we report a learning-enabled strategy for programming dimple encoded strips with prescribed morphing profiles and using them as building blocks for higher-level morphing architectures. Finite-element simulations are converted into centreline-based tangent-field representations to train a one-dimensional convolutional neural-network surrogate that maps local dimple encoding and strip geometry to global shape. The trained surrogate enables gradient-based inverse design of target morphologies, with optimized designs validated by independent simulations and physical prototypes. Variable-width strip modules expand the accessible shape space and enable width-conditioned curvature programming. Rotational tessellation of programmed strip modules creates dome-like morphing shells, while hierarchical integration of curvature-programmed strips with folding modules yields an ersatz flower exhibiting sequential opening through dimple switching. These results establish dimple-encoded strips as learnable building blocks that connect local bistable encoding, inverse design, and hierarchical shape formation, providing a route towards task-specific morphing components for soft robotics, adaptive interfaces, and deployable structures.
- Publication status:
- Accepted
- Peer review status:
- Peer reviewed
Actions
Authors
- Publisher:
- Wiley
- Journal:
- Advanced Materials More from this journal
- Acceptance date:
- 2026-11-03
- EISSN:
-
1521-4095
- ISSN:
-
0935-9648
- Language:
-
English
- Keywords:
- Pubs id:
-
2493060
- Local pid:
-
pubs:2493060
- Deposit date:
-
2026-11-04
- ARK identifier:
Terms of use
- Notes:
- This article has been accepted for publication in Advanced Materials.
If you are the owner of this record, you can report an update to it here: Report update to this record