In the Center for Rotating Machinery in Patrick F. Taylor Hall, a stainless steel bar is getting bent out of shape.
One end is clamped into a vice. The other is attached to a crank that pumps up and down, bending the bar like a 5-year-old playing with a paper clip.
It suddenly, loudly snaps. But mechanical engineering professor Michael Khonsari and his doctoral students are not caught off guard.
They knew when it was going to snap long before it happened.
Khonsari and his team have pioneered a radical new method to predict the strengths — and lifespans — of materials.
“If the material is new, we can tell you its life,” he said. ”If it’s been in use, we can tell you what its remaining life will be.”
Khonsari, the Dow Chemical Endowed Chair in Rotating Machinery, and two of his Ph.D. students, Mehdi Amiri and Mehdi Naderi, have pioneered a whole new way to look at wear and tear.
Currently, most engineers use a process called “stress-life approach” to estimate how long their machinery will last.
But in the past couple years, the professor and his students have found a way to use much smaller-scale, shorter-term tests to yield the same results using the “thermodynamic entropy approach.”
It all hinges on entropy — the physical phenomenon that represents disorder in a system.
A few years ago, Khonsari collaborated with professors at the University of Texas at Austin, who linked entropy to wear in mechanical components. The LSU team has extended the ideas to fatigue.
“When a material is new, it has very little or no entropy — no real disorder in it,” he said. ”When you cyclically stress it, entropy is generated in the material.”
With that connection established, the team started measuring the heat materials give off when under stress and translating the data to measure entropy in the system.
The team uses heat-sensing infrared cameras to precisely record the temperatures. The cameras were aimed at materials under stress tests, recording the temperatures at the breaking points.
That’s when Khonsari said the team had its first “eureka moment.”
They noticed the material’s temperature steadily rose until it reached a plateau. But right before the material failed, the temperature spiked dramatically.
That’s when one of Khonsari’s graduate students had the idea to chart the temperature in the materials and look for trends.
As they carefully monitored the materials’ heat and entropy, the researchers found the stainless steel bars broke once they accumulated a certain entropy value.
After repeated tests, they discovered that value was the same every time for each individual type of material.
“It looks like we’ve discovered a new material constant,” Khonsari said one of his students told him. “We have measured the value of entropy at which a material fails and use this information to predict how long the material will last.”
To prove it, the students rigged up a machine that monitors the entropy in a specimen as it is stressed. Once the specimen reaches 90 percent of its entropy capacity, the machine stops the test.
Once the machine starts back up again, it breaks — 10 percent later.
In addition to bending stress, the team has tested the materials for torsion stress (twisting), tension and compression stress and even stress that bends and rotates at the same time. The method works for all of them.
The new predictive ability significantly improves upon the traditional stress-life approach, Khonsari said.
First, it allows engineers to test the strength of their materials without having to simply use them until they break. The process requires only a small sample of material and a few cycles — a short amount of time.
Second, the thermodynamic entropy approach works independently of loading sequence.
In the field, a material isn’t under constant stress. The stress changes as the machine performs different tasks. A jumbo-jet-sized windmill blade, for example, is under more stress on a blustery day than it is on a calm one. A car engine is under more stress when accelerating quickly than when it’s idling.
That causes problems for the stress-life approach, because it measures material life in the number of cycles — or time. If a material is under more stress, that time decreases. If it is under less stress, that time increases.
But the thermodynamic entropy approach doesn’t rely on time for its predictions — it relies on accumulation of entropy.
“It’s like you have a cup, and you’re measuring it as it fills up,” Khonsari explained. ”Instead of saying, ‘Water will spill out in so many seconds,’ we’re saying, ‘Water will spill out once it reaches the top.’ It doesn’t matter if you’re doing a drip or a flood.”
Khonsari thanks the College of Engineering and Department of Mechanical Engineering for giving him the resources to make the discovery, which he says could potentially revolutionize the way engineers discuss material fatigue.
“The theory in textbooks hasn’t changed in 40 or 50 years,” he said. “While our work is still in the research stage, the concept is quite profound.”
Khonsari and his graduate students have filed two patent applications for their discoveries.  
“LSU stands to gain greatly from this technology,” he said.
 
—-
Contact Matthew Albright at [email protected]