For patients with heart failure, blood pumps can be lifesaving. But the very forces that sustain circulation can also harm it damaging red blood cells through hemolysis and compromising the body’s oxygen supply.
Now, supercomputer simulations are revealing how red blood cells deform under stress, offering new insights that could lead to safer, more effective blood pump designs.
"Supercomputing infrastructure is critical to advancing healthcare and scientific knowledge," said Keefe Manning, Ph.D., who holds joint positions in the Department of Biomedical Engineering, The Pennsylvania State University, and the Department of Surgery, Penn State College of Medicine.
"We’re able to create more complex environments and parameterization now because of supercomputers. Their value in science cannot be understated. As we apply concepts such as a digital twin and as we improve the infrastructure, we can modify the physiological models in a way that we never could before and accelerate our understanding."
Manning is the corresponding author of a study on computational modeling of red blood cells, published January 2026 in the Annals of Biomedical Engineering. In it, he and colleagues lay the foundation for understanding hemolysis, the red blood cell destruction associated with blood pumps, referred to as mechanical circulatory support devices (MCSDs).
The science team was awarded supercomputing allocations on the Stampede3 system at the Texas Advanced Computing Center (TACC) by the U.S. National Science Foundation-funded Advanced Cyberinfrastructure Coordination Ecosystem: Services & Support ( ACCESS ) program, which provides support for thousands of scientists across the nation.
Manning and colleagues’ model advances beyond earlier methods that used simplified stress calculations and lacked mechanistic predictability for complex flow conditions.
Instead, they adapted the governing droplet deformation evolution equation to scale to the complex, multidimensional flow environment characteristic of MCSDs in a continuum approach that uses the open source computational fluid dynamics software OpenFOAM .
The authors calibrated the model's constitutive parameters to reproduce human red blood cell data with reasonable mean absolute error.
On the data side, the scientists collected blood samples from a dozen people. They ran the blood through microfluidic channels, capturing images that were post-processed to reveal the extraordinary shape-changing of red blood cells. While many people are familiar with their ubiquitous biconcave disc shape, red blood cells can also be squished flat like a folded parachute or stretched out like a torpedo as they squeeze through tiny capillaries.
"The novelty of our study lies in the volume of experimental data that we collected internally to calibrate the droplet model," Manning added.
This research marks a step forward in understanding the biophysics of red blood cell behavior in very thin layers, behavior at the flow fields that has been hard to reproduce in simulations.
"A simple laptop computer won't be able to do the computation, "Manning said. "ACCESS resources have been critical for Hannah Palahnuk, my PhD student, to get this work done. Having OpenFOAM supported within ACCESS is important for our research in helping improve healthcare of our nation’s population and create future technology that will be less harmful to red blood cells."
"I utilized the ticketing system when I needed help, and the staff were responsive and helpful," said study co-author and principal investigator Hannah Palahnuk, a Ph.D. candidate in Biomedical Engineering at Penn State University.
"TACC's Stampede3 supercomputer provided a much higher core-hour allocation, allowing high-fidelity, high-resolution simulations in reasonable amounts of time. This was extremely helpful as the resources are free and they are driving forward important research in the biomedical engineering field," she added.
Moving forward, the research will model more realistic physiological conditions for red blood cell concentrations and translate these advances into direct hemolysis measurements.
"Once validated at scale across real-world fluid environments, this work could deepen our understanding of hemolysis in blood pumps, and help save lives in the process," Manning concluded.
The study, "Modeling Red Blood Cell Deformation at Supraphysiological Strain Rates Using a Droplet Framework," was published January 2026 in the Annals of Biomedical Engineering. The study authors are Hannah P. Palahnuk, Nicolas A. Tobin, and Keefe B. Manning of The Pennsylvania State University. Funding was provided by the Walker Assistantship program at the Penn State Applied Research Laboratory and grant U54 TR002014-05A1 from the National Institutes of Health, with computational support from ACCESS allocation MDE 24001.
Annals of Biomedical Engineering
Computational simulation/modeling
Not applicable
Modeling Red Blood Cell Deformation at Supraphysiological Strain Rates Using a Droplet Framework.
29-Jan-2026