UNH researchers find better ways to estimate shape of very small, irregular particles

  • age of an ash particle from the Mt. St. Helens volcanic eruption of 1980. A series of closely aligned images are used to construct a 3-D model of the particle. Due to the complexity of reconstruction, imaging alone cannot capture the whole particle. Photo credit: Mark Townley/UNH UIC Mart Townley—UNH UIC

Monitor staff
Monday, August 07, 2017

It’s always fun to learn how science – despite its accumulated knowledge, wisdom and evidence – can sometimes be stumped by the obvious. A case in point: the concept of shape.

Yes, shape – as in: “My nose isn’t broken; it has always been this shape.” Easy to say, hard to analyze.

“You talk to five different scientists, each one will give you a different answer about what shape is,” said Gopal Mulukutla, a research scientist at UNH’s Institute for the Study of Earth, Oceans and Space.

Mulukutla knows whereof he speaks. As a civil engineer by training whose doctorate in ocean engineering makes him “basically a sediment physicist,” Mulukutla is lead author on a new study that tackles the problem of determining the shape of teeny-tiny volcanic ash particles.

This is a problem because when a volcano blows its stack and sends thousands or millions of tons of ash into the air, we want to know where the massive cloud will go. That, however, depends on the extremely irregular shape of the billions to trillions of microscopic bits of rock and lava that get tossed skyward to be carried around by winds.

“They fall, but they also tumble, and that is a function of the shape. Each size and shape will have different tumbling characteristics ... so in order for anyone to better predict the movement of a volcanic ash cloud, we need to develop parameterization of (size and shape),” Mulukutla explained.

Size and shape also determine when and whether particles will nucleate, or gather enough moisture to create precipitation that falls to the ground, another important variable when making predictions.

In other words, a cloud of ash particles shaped like spheres would behave very differently than a cloud of particles shaped like porcupines – hence the need to quickly analyze and categorize the shape of these particles from microscopic pictures of a few of them, the way we can quickly analyze mass.

Without a good science of shape, however, that is hard to do. A paper just published by Mulukutla and three co-authors in the journal Measurement Science and Technology presents a mathematical model that appears to do a much better job than current processes.

“There are methods where you can take a small particle and over many, many hours capture the whole surface, but it is very expensive and not really practical when wanting to capture properties of hundreds of particles,” said Mulukutla, who published with Kimberly Genareau of the University of Alabama, Adam Durant of the University of Oslo in Norway and a UNH EOS colleague Alexander Proussevitch.

The details, as details tend to be in science, are not easy for the layman to follow. The team use a system called Delaunay triangulation to create a mathematized version of the surface of particles from microscopic pictures, a structured mesh that is further analyzed using solid angles, a measure of how large a physical object appears from any given observation point. It’s not something you and I will be using anytime soon, but it appears to be effective.

“The error does not vary more than plus or minus 5 percent. If you don’t do it this way it can vary from 70 to 80 percent,” Mulukutla said.

Testing this method on microscopic particles of known size and shape has given the team enough confidence that UNH Innovation, the university’s department managing and promoting intellectual property, has filed for a patent.

It’s not just vulcanologists who can use this process. Mulukutla said it can help understand how rivers and streams are affected by sediment, which, like ash, consist of a zillion tiny particles being moved around by fluid mechanics. It can also help develop new blood tests, which require assessing the shape and properties of elongated blood droplets, and improve analysis and reaction to algae and the study of ocean fossils.

All well and good. But will it lead to a scientific understanding of the concept of shape, perhaps based on a new SI unit of measure, the Mulukutla?

Probably not, I fear. But if Concord’s Rattlesnake Hill ever surprises us and does a Mount St. Helens routine, we’ll have a better chance of knowing which way to run.

(David Brooks can be reached at 369-3313 or dbrooks@cmonitor.com or on Twitter @GraniteGeek.)