Designing random fiber networks, optimized for strength and toughness

In nature, random fiber networks such as some of the tissues in the human body, are strong and tough with the ability to hold together but also stretch a lot before they fail. Studying this structural randomness - that nature seems to replicate so effortlessly - is extremely difficult in the lab and is even more difficult to accurately reproduce in engineering applications. Recently, researchers in The Grainger College of Engineering, University of Illinois Urbana-Champaign and the Rensselaer Polytechnic Institute devised a method to repeatedly with desired characteristics and use computer simulations to tune the random network characteristics for improved strength and toughness.

“This is a big leap in understanding how nanofiber networks behave,” said Ioannis Chasiotis, a professor in the Department of Aerospace Engineering. “Now, for the first time, we can reproduce randomness with desirable underlying structural parameters in the lab and with the companion computer model we can optimize the network structure to find the network parameters, such as nanofiber density, that produce simultaneously higher network strength, stiffness and toughness.”  

Chasiotis’ group has been studying how nanofibers behave mechanically for over 20 years.

“To give you a perspective of the size we are working with, the diameter of each nanofiber in the network was approximately 300 times smaller than the human hair,” Chasiotis said. “This small diameter gives special mechanical properties to polymer nanofibers that are not possible for polymer fibers with large diameters.”

He said they already understood what makes a single nanofiber stronger, stiffer and tougher but when they were combined to create non-woven materials, those properties don’t carry over in a direct way. That’s what prompted this research focusing on how nanofibers interact when they come together as a network.

His Ph.D. student, HyongJu Lee, worked in the lab to print nanofiber networks of practical sizes of the order of centimeters via a method called near-field electrospinning. Chasiotis said the student did not invent the near-field electrospinning method but built the apparatus and made the method work for large scale samples, also developing a novel methodology to mechanically test the nanofiber networks, while also providing the modeling team at Rensselaer Polytechnic Institute the input data to replicate the exact structure of the printed networks in their computer models.

Now that the computational model has real input data, it can be used to simulate more complex nanofiber networks using parameters which would be difficult to produce in the lab or make predictions for production scale nanofiber networks made by conventional electrospinning.

“So, for example, we may have already tested random networks containing 500 to 5,000 nanofibers but we want to know what happens when we have millions of fibers,” Chasiotis said. “Experimentally in the lab, millions of nanofibers would take a very long time to print by near-field electrospinning, but it is very fast to spin using conventional electrospinning that is already used for mass production in industry. The model can extrapolate what will happen under different parameters, such as the number of nanofibers, how close they are to one another and how often they cross each other.”

To create the nanofiber network or mesh with near-field electrospinning, a polymer solution was used one droplet at time while extruded from a fine needle, practically drawing it onto a gold-coated silicon wafer using very high voltage. Every place where one nanofiber crossed another was bonded using heat treatment.

To study their properties the nanofiber networks had to be lifted from the deposition surface, a task Chasiotis said brought with it challenges.

“We spent probably six months learning how make the test samples remain freestanding,” he said. “We had to lift them from the deposition surface while still wet without damaging them. We treated the deposition surface to become slippery so we could lift the nanofiber networks off. We also had to assess the crimp in our printed networks and translate it into numbers the modelers could account for in the model. That’s when the experiment and the model began to converge.”

This work was supported by the National Science Foundation.

The study, “An integrated experimental-computational investigation of the mechanical behavior of random nanofiber networks,” was written by HyeongJu Lee, Kathiresan Karunakaran, and Ioannis Chasiotis from U. of I. and Mithun K. Dey and Catalin R. Picu from Rensselaer Polytechnic Institute. It is published in and featured on the cover of the journal Soft Matter, Issue 10, 2025.  DOI:10.1039/D4SM01288G