Bio-inspired brick-and-mortar TiNi/Ti2Ni composites: Microstructural evolution mechanisms and conditional generative prediction

Jun Zha; Guoqing Zu; Mingyu Li; Xingang Liu; Pengwei Liu; Xingwang Cheng; Zhiping Xiong

doi:10.1016/j.coco.2026.102749

Bio-inspired brick-and-mortar TiNi/Ti₂Ni composites: Microstructural evolution mechanisms and conditional generative prediction

Jun Zha
, Guoqing Zu^*
, Mingyu Li
, Xingang Liu
, Pengwei Liu
, Xingwang Cheng
, Zhiping Xiong^*

^*Corresponding author for this work

Yanshan University
Changchun University of Technology
Beijing Institute of Technology

Research output: Contribution to journal › Article › peer-review

Abstract

Bio-inspired brick-and-mortar TiNi/Ti₂Ni composites were fabricated via vacuum sintering and hot rolling. Multiscale characterization revealed that deformation coordination was governed by dynamic recrystallization (DRX) of the brittle TiNi/Ti₂Ni phase and strain-induced precipitation of nanoscale Ni₄Ti₃ in the ductile TiNi matrix. Furthermore, a conditional Denoising Diffusion Probabilistic Model (DDPM) was developed to predict microstructural evolution. The model demonstrated high fidelity in capturing the complex nonlinear mapping between processing parameters and microstructural topology. This work establishes a novel framework combining physical metallurgy with generative AI for designing high-performance heterostructured materials.

Original language	English
Article number	102749
Journal	Composites Communications
Volume	63
DOIs	https://doi.org/10.1016/j.coco.2026.102749
Publication status	Published - Apr 2026
Externally published	Yes

Keywords

Brick-and-mortar structure
Heterostructured
Intermetallics
Materials design
Microstructures

Access to Document

10.1016/j.coco.2026.102749

Cite this

@article{bf5183049947481e99621a2cca43d12b,

title = "Bio-inspired brick-and-mortar TiNi/Ti2Ni composites: Microstructural evolution mechanisms and conditional generative prediction",

abstract = "Bio-inspired brick-and-mortar TiNi/Ti2Ni composites were fabricated via vacuum sintering and hot rolling. Multiscale characterization revealed that deformation coordination was governed by dynamic recrystallization (DRX) of the brittle TiNi/Ti2Ni phase and strain-induced precipitation of nanoscale Ni4Ti3 in the ductile TiNi matrix. Furthermore, a conditional Denoising Diffusion Probabilistic Model (DDPM) was developed to predict microstructural evolution. The model demonstrated high fidelity in capturing the complex nonlinear mapping between processing parameters and microstructural topology. This work establishes a novel framework combining physical metallurgy with generative AI for designing high-performance heterostructured materials.",

keywords = "Brick-and-mortar structure, Heterostructured, Intermetallics, Materials design, Microstructures",

author = "Jun Zha and Guoqing Zu and Mingyu Li and Xingang Liu and Pengwei Liu and Xingwang Cheng and Zhiping Xiong",

note = "Publisher Copyright: {\textcopyright} 2026 Elsevier Ltd.",

year = "2026",

month = apr,

doi = "10.1016/j.coco.2026.102749",

language = "English",

volume = "63",

journal = "Composites Communications",

issn = "2452-2139",

publisher = "Elsevier Ltd.",

}

TY - JOUR

T1 - Bio-inspired brick-and-mortar TiNi/Ti2Ni composites

T2 - Microstructural evolution mechanisms and conditional generative prediction

AU - Zha, Jun

AU - Zu, Guoqing

AU - Li, Mingyu

AU - Liu, Xingang

AU - Liu, Pengwei

AU - Cheng, Xingwang

AU - Xiong, Zhiping

PY - 2026/4

Y1 - 2026/4

N2 - Bio-inspired brick-and-mortar TiNi/Ti2Ni composites were fabricated via vacuum sintering and hot rolling. Multiscale characterization revealed that deformation coordination was governed by dynamic recrystallization (DRX) of the brittle TiNi/Ti2Ni phase and strain-induced precipitation of nanoscale Ni4Ti3 in the ductile TiNi matrix. Furthermore, a conditional Denoising Diffusion Probabilistic Model (DDPM) was developed to predict microstructural evolution. The model demonstrated high fidelity in capturing the complex nonlinear mapping between processing parameters and microstructural topology. This work establishes a novel framework combining physical metallurgy with generative AI for designing high-performance heterostructured materials.

AB - Bio-inspired brick-and-mortar TiNi/Ti2Ni composites were fabricated via vacuum sintering and hot rolling. Multiscale characterization revealed that deformation coordination was governed by dynamic recrystallization (DRX) of the brittle TiNi/Ti2Ni phase and strain-induced precipitation of nanoscale Ni4Ti3 in the ductile TiNi matrix. Furthermore, a conditional Denoising Diffusion Probabilistic Model (DDPM) was developed to predict microstructural evolution. The model demonstrated high fidelity in capturing the complex nonlinear mapping between processing parameters and microstructural topology. This work establishes a novel framework combining physical metallurgy with generative AI for designing high-performance heterostructured materials.

KW - Brick-and-mortar structure

KW - Heterostructured

KW - Intermetallics

KW - Materials design

KW - Microstructures

UR - https://www.scopus.com/pages/publications/105029358997

U2 - 10.1016/j.coco.2026.102749

DO - 10.1016/j.coco.2026.102749

M3 - Article

AN - SCOPUS:105029358997

SN - 2452-2139

VL - 63

JO - Composites Communications

JF - Composites Communications

M1 - 102749

ER -

Bio-inspired brick-and-mortar TiNi/Ti₂Ni composites: Microstructural evolution mechanisms and conditional generative prediction

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this