Their work is supported by the National Science Foundation MRSEC Program and a Grant-in-Aid of Research, Artistry and Scholarship from UMN.
AFM under the microscope
The atomic force microscope (AFM) was originally developed to image surfaces at an atomic level inside high vacuum systems, completely devoid of liquid or gas. The AFM’s tiny oscillating stylus is attached to the tip of a cantilever, which is deflected by artifacts on the sample surface. A laser beam reflected off of the back side of the cantilever onto a photodetector records the movement of the cantilever. A computer then converts these movements into an image of the sample surface.
Recently, crystal growers have begun implementing the AFM to view growing crystals in situ. To do so, the system is immersed in a liquid, called a fluid cell. Solution is fed through one opening of the cell and exits through another. As the solute is depleted by precipitating to form the crystal, it must be continuously refreshed by flowing new solution through the cell.
“With the AFM fluid cell, crystal growers can observe crystal growth in real-time” Gasperino says. “We are interested in the broader implication of how the measurement system, the AFM fluid cell itself, can potentially impact the measurements.”
Fluid flow could influence deflection of the cantilever and affect the measurement results. Gasperino and his colleagues set out to study this hypothesis, test whether the configuration limited mass transport (the transfer of mass from an area of high concentration to one of low concentration), and quantify the flow and transport processes within the cell.
‘EnSight’ into AFM Flow Effects
The complex 3D nature of the flows in the AFM fluid cell makes this inherently difficult. However, Gasperino and his team have employed finite element analysis techniques using EnSight extreme visualization software to render CFD simulations of the flow within the cell and examine its effect on the growing crystal. By combining EnSight’s clip plane feature with the isoline and pathline features, the team can quickly extract useful observations about the results.
“Recently, researchers have been pushing the limits of AFM scanning frequencies to visualize characteristics of crystal growth and biological systems that occur on very small time scales. We are looking at the implications that these high-speed oscillations might have on the flow and mass transport in the fluid cell,” Gasperino says. “With EnSight, we can rapidly load the enormous amounts of transient data necessary to study this system. Using the vortex core extraction tool, we can observe vortex shedding in the cell and create a movie using the flip book feature to visualize the transient behavior of the phenomena.”
It turns out that the effects of flow and oscillation frequency on the AFM can be significant in some fluid cell systems. Based on the visualizations from EnSight, researchers can now assess these effects and understand how their results may be impacted by the technique.
Gasperino’s research demonstrates EnSight’s efficiency and flexibility. The group uses in-house code to generate the numerical models; meanwhile, all of the heavy data lifting is done on high-performance PC-based servers at the UMN Supercomputing Institute. The data is then exported into EnSight format for post processing and visualization on a Mac platform. This system allows for the ideal division of labor, with the data-centric PC system working in tandem with the graphics-friendly Mac.
“It’s really convenient if you’re using a Mac, and you have a fast network connection, to port over to these machines and run EnSight off of them,” says Gasperino, who has been with EnSight through two upgrades over the past several years. “Most of the large commercial codes are also now exporting in EnSight format. It really is much quicker than any other viewer. For what we do, EnSight is a natural fit; it allows you to quickly elucidate interesting phenomenon within a complex 3D system.”
EnSight also enables the team to render and annotate still images appropriate for professional journals.
“About 95 percent of our time is spent getting our models and codes to work – you get all the way to the end and you haven’t seen anything yet,” Yeckel says. “With EnSight, it’s like you go from zero results to all of these pictures relatively quickly. I’m interested in the science, not in taking a year to write something just to turn my data into pictures when there are already very mature products that can do that for me.”
The theoretical microscope
EnSight has proven to be a critical component of the UMN team’s work in two very distinct ways. As a research tool, EnSight allows them to view images of phenomenon that they otherwise could not see.
“It’s our version of a microscope,” Yeckel says. “It also allows us to develop compelling presentation materials to explain this to other people, and these are distinctly different processes.”
EnSight’s speed, efficiency and accuracy play a significant role in visualizing the data at any point throughout the research process. Yeckel often turns to EnSight briefly during the analytical process to visualize a bit of code. Even though these images may never be seen by another viewer, EnSight allows him to quickly verify the data before proceeding. Then, during the final stages of a project, the team enlists EnSight to generate high-impact images for publication. With limited real estate in professional journals, and the high cost of including color figures, Yeckel says “marquee” images are a must and EnSight delivers.
“You have to convey as much information in as little space as possible and still make it easy on the reader,” Yeckel says. “If you don’t, they will not read it. When people don’t read your science, it is never going to be important.”
Selling the science
High-impact is also the name of the game for EnSight’s animation generating capabilities. The key frame animation feature makes the process of compiling an animation very simple, even for those not skilled in the finer points of movie production.
“At a big conference, you might only have 15 minutes to present a year’s worth of work that you could easily devote hours to,” Yeckel says. “With a compelling EnSight animation, you can accomplish more in a 30 or 60 second video than you can in 15 minutes of talking.”
If this makes the presentation sound more like a TV commercial than a scientific endeavor, Yeckel says that is no accident. While the scientific community may be reluctant to acknowledge the role of sales and marketing in their work, Yeckel argues that scientists do have a product to sell: their ideas. And, it helps to have tools like EnSight to sell them, particularly in the competitive hunt for funding.
“The things that we own are our reputation, our research accomplishments,” Yeckel says. “We’re out there selling our stuff, and it helps to have good supporting materials to really make the case.”
As competition for funding grows increasingly fierce, scientists must leverage every aspect of previous research to gain additional funding. Using EnSight to generate compelling visuals for funding proposals can give researchers a leg up on the competition. Yeckel says that when bootstrapping preliminary, yet promising, results to gain full funding, high impact visuals can tip the scales.
“You’ve got a bunch of tired reviewers, and maybe this is the 20th proposal they’ve read,” Yeckel says. “If only one out of five is going to get picked for funding, presentation really matters.”
Shear stresses on the AFM cantilever overlayed with flow streamlines (background), the computational fluid cell domain with streamlines showing the path of flow through the cell (top right), and the entire fluid cell (bottom right).
A 3D view of vortex shedding at the edges of the AFM cantilever; the red lines are vortex cores calculated by EnSight. Streamlines are seeded from these cores to show the nature of the vortex at this particular time step.