Engineers hope study of fatal Bullard fire reveals early warning signs for firefighters

Engineering professor Mike McGinnis keeps a newspaper from 2007 in a file cabinet close by his desk.

Yellow marks highlight information he doesn’t want to forget, but he hopes he never sees again.

The front-page clip from The Tyler Morning Telegraph describes the results of an investigation into the deaths of Noonday firefighters Austin Cheek and Captain Kevin Williams.

The firefighters died within minutes of entering the burning house near Bullard in the early morning hours of Aug. 3, 2007.

McGinnis highlighted sentences in the article about how the home’s reconstructed roof hindered firefighter’s efforts as determined by a Firefighter Fatality Investigation Report issued by the State Fire Marshal’s office.

Unaware a new roof had been added over the old one, firefighters attempted to ventilate the home by cutting a hole in the roof, but could only puncture one section.

Ventilating a structure removes smoke, heat and particles of combustion, increasing firefighter visibility and reducing chances of flashover, a potentially deadly event involving super heated air, according to the report.

It was a problem McGinnis and research partner Torey Nalbone, both associate professor of civil engineering, believed they could help solve.

“Flashover kills firefighters. It happens so fast that firefighters die because they have literally 10 seconds, even in full gear, to get out of the room or they die,” McGinnis said.

“It’s horrifying to read and yet very powerful, but a firefighter will describe flashover as ‘heat pouring down on them like rain,’ he said. “That’s a powerful image.”

Nalbone said standard procedure is to at least visually evaluate the course of the fire and pull that fire to a location that is the furthest remote in the building, clearing a path for the firefighter with a water hose.

“We’ve looked at it, and there are five or six fires in just the last three years alone where firefighters have either died or been injured in this kind of fire,” McGinnis said.

Since reading that first article, the researchers have gathered official reports on double-roof fires from all over the country, often communicating directly with the fire chiefs involved. McGinnis and Nalbone request copies of the official investigation reports for use in their research.

“It’s horrifying. You can read all the radio transmissions from the whole thing. You know what’s happening when you read, ‘Where are they? Where are they?’” McGinnis said.

The researchers say they are focusing on providing firefighters with external cues that may indicate what stage the fire is in. They hope this prevents deaths caused by firefighters unknowingly walking into flashover situations.

In July, McGinnis and Nalbone received a $4,798 internal grant from the University’s Office of Sponsored Research to further their study of double-roof fires.

They presented their preliminary findings at a Faculty Research Day conducted Friday at the David G. and Jacqueline M. Braithwaite Building.

“The grant program is competitive, so only the best ones are selected for funding by the Research Council. This program is one of the ways that we try to help faculty get prepared for submitting a research proposal for external funding, like from the National Science Foundation,” said Arlene Horne, associate vice president for research and federal relations.

Though the researchers say their project has been well received among local fire departments, finding funding has been more difficult.

“We’ve been trying to build a coalition locally to look at this and get national exposure for it. The Smith County fire marshal and his staff have been very supportive, and we’ve talked with the Flint fire chief who lost the firefighters. Three or four different fire chiefs in the area wrote letters of support for us saying this is good research,” McGinnis said.

Nalbone said they submitted their idea to the National Institute of Occupational Safety and Health, but did not receive funding.

“We are going to go again, because now they have better defined the objectives for firefighter research and we believe now we can use those objectives to better craft our grant proposal,” Nalbone said.

The research project is unique in that McGinnis and Nalbone have divided the tasks equally.

“I am really focusing on the structure of the building and Dr. Nalbone is really focusing on all the particles in the fire and all the gases and all the chemistry. That is not a collaboration that normally happens,” McGinnis said.

The grant provides the researchers a means of moving forward in their research.

“We are looking at trying to build a platform to actually perform our research on, so we’ve got a new kiln and fire blocks. We are putting instruments on that so we can measure the temperatures,” McGinnis said.

Nalbone and McGinnis are using a Forward Looking Infrared camera, called a FLIR camera, to read heat levels in a structure’s window. With that information, firefighters can better estimate how much time is available before a flashover occurs.

“By looking at the exterior of the building, and maybe some high rates of heat accumulations in the windows, we hope we can say the amount of heat that the window is telling them is not consistent with the assumed time of burn. There’s something going on,” McGinnis said.

The researchers said the thermal imaging camera they are using is the same type used by fire departments around the country.

“We are using the FLIR camera for a couple reasons. One is that it’s great for heat. The second thing is that it’s practically a mainstay in fire departments. They usually use a black and white FLIR to look for survivors because it can penetrate smoke and pick out body heat, or look behind a wall and see if they have fire in the wall,” Nalbone said.

He added the cameras are not cheap, costing between $10,000 and $50,000.

“What we are going to attempt to do is give them another use for that technology. In most cases they have made significant investments in those cameras,” he said.

McGinnis and Nalbone recently conducted an experiment using a torch to heat a pane of safety glass. They attached tiny dots to the glass and used a digital image correlation system to analyze how those dots moved in relation to the heat application.

“We take photographs of that dot pattern, two at once, and essentially trace the light rays from each camera. Each of those little dots then becomes a place we can measure,” McGinnis said.

After being photographed later, the researchers use a computer to take those two patterns, overlay them and figure out where the dots all moved.

“We get really accurate moves that way, down to microns,” McGinnis said. “I can breathe on something and it moves 100 microns. It’s really accurate stuff.”

“If we can tie what we find with the digital correlation to the patterns we see in the FLIR, then we can give them [firefighters] that information and say, ‘When you see this, this is what you need to be concerned with,’” Nalbone said.

McGinnis also said they are using this opportunity to involve students in their research.

Malick Bittaye, a senior civil engineer major, assisted with some of the preliminary testing during the summer. Nalbone said another student who is a mechanical engineering major is scheduled to assist when actual particle testing begins.

“If they [students] don’t get the opportunity to create new knowledge, explore the boundaries, then they are at a disadvantage when they leave here,” McGinnis said.

“As we do research, we can actually bring it into the classroom and demonstrate the physical effects of something that we only talk about theoretically out of the textbook,” Nalbone said.

Though inspired by Cheek and Williams, the researchers said they have not had direct contact with the victim’s families.

“It’s a sensitive issue, but they may have heard about the research through the grapevine,” McGinnis said.

For now the only link is that well-worn newspaper with words highlighted in yellow.