CCM CONNECTS: Research Series

The annual Research Series comprise of monthly overviews of CCM’s research focus areas with presentations on topics ranging from composites process modeling, interface characterization, thermoplastic composites, micro-mechanics of composites, additive manufacturing and many more. The Research Series are free and open to the public, and speakers include graduate students, post-docs, research associates, affiliated faculty members and invited speakers from the government and industry.

 

January 13, 2026

PRESENTER	TITLE	ABSTRACT
Nuwan Diwapriya, PhD Research Associate III	Machine Learning–Enhanced Quantum-to-Continuum Modeling of Semiconductor Wafer Bonding	In this DARPA-funded project, we are developing a predictive, quantum-to-continuum modeling framework that advances the ability to design and control bonding between semiconductor materials. The framework integrates density functional theory (DFT)–trained machine-learning interatomic potentials (MLIPs) with large-scale molecular dynamics (MD) simulations. Traction–separation relations derived from MLIP-accelerated MD directly inform finite element models, linking quantum-scale bonding mechanisms to wafer-scale device performance. This talk will focus on the initial development of MLIPs for modeling interfacial bonding.
Ankita Bisht, PhD Postdoctoral Researcher	Mixed-Mode Fracture Behavior of Woven S-Glass/Epoxy Composites: Role of Matrix, Fiber Sizing, and Weave Architecture	Fiber-reinforced polymer composites are widely used in load bearing aerospace and defense structures, where resistance to delamination is critical for ensuring safety and durability. In most structural applications, these composites are subjected to complex, mixed-mode interlaminar stresses that can lead to delamination. Improving delamination resistance has therefore been a key focus in composite design, and our previous work using thermoplastic interlayers has demonstrated significant improvements in delamination resistance and stiffness retention. This talk presents experimental investigations on the fracture behavior of composite laminates under mixed-mode bending (MMB) loading. By systematically varying resin systems, fiber sizings, and fabric architecture, the study quantifies Mode I and Mode II energy release rates and establishes R-curve behavior across different mode-mix ratios. Master curves and failure envelopes are developed to enable fracture toughness prediction beyond experimentally accessible conditions. The results highlight the influence of resin chemistry, fiber sizing, and weave architecture on crack growth resistance, energy absorption, and damage tolerance, offering new insights for designing composites with enhanced performance in structural applications
Uday Kiran Balaga Graduate Student, Mechanical Engineering	Influence of Part Geometry on the Fiber Orientation Behavior in Additive Manufacturing	Additive manufacturing (AM) using fused filament fabrication (FFF) with short fiber reinforced polymers offers enhanced mechanical performance; however, these gains depend strongly on the local fiber orientation achieved during deposition. Prior studies have shown that part geometry, particularly regions with sharp curvatures, can disrupt fiber alignment and introduce highly random orientations, limiting achievable performance. Despite this known sensitivity, systematic approaches to dynamically control processing parameters based on local geometric features remain limited. This work investigates the influence of curvature radii on fiber orientation behavior and establishes a framework for geometry aware process optimization. A sinusoidal geometry with varying frequency is selected as the test geometry, providing regions of differing curvature radii within a single printed feature. Experimental orientation analysis revealed that the principal orientation component a11 (printing direction) decreases as curvature radius decreases, indicating loss of alignment under constant processing conditions. In contrast, regions with straight paths in the same part maintained consistent orientation. Using the collected experimental data, we demonstrate the potential for a dynamic control strategy in which the printer interprets local geometry and adjusts parameters in real time to preserve fiber alignment in sharp curvature zones. This approach enables more reliable property translation in fiber-reinforced FFF parts.

 

December 9, 2025

PRESENTER	TITLE	ABSTRACT
Swarna Shiv CEO, Unsmudgeable	From Dorm Room to Customers at 24: A Materials Startup Journey	At 24, I’ve taken a materials startup from a dorm room idea to a patent-pending company with five enterprise pilots, initial revenue, and over $250K in funding: all without a science background. This talk shares the unconventional journey of building application-driven science outside academia: the challenges of gaining credibility in a technical field, the pivots that led to enterprise pilots, and the lessons learned bridging research and commercialization. Whether you’re a researcher considering your first startup or an entrepreneur exploring deep tech, you’ll walk away with practical insights on validating ideas and taking the leap from concept to company

 

November 11, 2025

 

PRESENTER	TITLE	ABSTRACT
Troy Wilford Director, Commercialization Programs, Venture Development Center	Unlocking Resources for Innovation	Research translation is a lengthy and risky process. Many have found success navigating this process by augmenting their mastery of scientific research with an entrepreneurial mindset molded by support programs and resources made available through Horn Entrepreneurship. Hear from Troy Wilford, Director of Commercialization Programming, to learn more.
Navid Niknafs Kermani, Ph.D. Research Associate III	Strain Rate Dependent Constitutive Model for Uncured Aligned Discontinuous Fiber (Or TuFF) Composites for Forming: Development, Implementation, And Validation	A strain-rate-dependent constitutive model is developed, implemented in FEA software and validated for uncured aligned discontinuous fiber (ADF) prepregs under uniaxial tension. Experiments spanning three orders of magnitude in strain rate reveal nonlinear stiffening up to a rate-dependent peak stress, followed by pronounced softening and viscous relaxation. A phenomenological model with distinct stiffening, softening, and relaxation regimes is calibrated, reproducing the full stress–strain behavior across all rates. An incremental formulation is developed that enables arbitrary strain-rate histories and finite element implementation in AniForm and Abaqus. Validation studies using both coupon-level specimens and complex part geometries demonstrate excellent agreement between model predictions and experimental measurements. This methodology can be applied to other ADF material systems and supports future extensions to multiaxial and non-isothermal conditions.
Sagar Doshi, Ph.D. Scientist	Carbon Nanocomposite Coated Fabric Based Flexible Pressure Sensors — Uunderstanding Sensing Mechanism and Applications in Gait Analysis	Dr. Doshi will provide recent progress made in understanding the sensing mechanism for the flexible pressure sensors, uncovering the underlying reasons that enable an extremely wide range of sensing, from tactile range to tons. Recent progress made in using these sensors for gait analysis will be discussed, and how machine learning is leveraged to classify gait types.

 

December 9, 2025

PRESENTER	TITLE	ABSTRACT
Swarna Shiv CEO, Unsmudgeable	Coming Soon	Coming Soon.

 
 

---

PREVIOUS SESSIONS

302-831-8149 • info-ccm@udel.edu