Exceptional Computer Modeling Advances Cabin Fire Research
FAA computer scientists and engineers have devised a faster, more economical means to explore the early detection of hidden fires in aircraft cabins, bringing more insights into fire research and advancing aircraft safety.
The Fire Safety Laboratory at the FAA’s William J. Hughes Technical Center and Atlantic City, N.J., has been investigating the detection of hidden fires igniting within aircraft cabins.
In a perfect world with unlimited resources, researchers would be able to burn out numerous fuselages outfitted like a standard aircraft cabin and then report outcomes, regardless of time, materials or costs. The world is not perfect, however, and resources are not infinite.
Fortunately, the world has become ever more digital. The lab’s senior scientist, Haiqing Guo, has devised a faster, more efficient and economical means to explore this threat, with help from project supervisor Richard Lyon and colleagues Sean Crowley and Paul Scrofani. Guo and his team have adopted state-of-the-art computational fluid dynamics codes (developed by the National Institute of Standards and Technology (NIST)) to simulate the movement of heat and smoke from a hidden cabin fire. The team conducted full-scale fire tests to validate the computer simulations.
“Computational fluid dynamics is a fancy way of saying that we use a computer to calculate how air flows around objects in the airplane as a function of time when heated by a fire,” Guo said.
His role is to perform the computer simulations, consequent data analysis, and validation test plans. He devised means to gain a fuller understanding of how smoke and hot gases emanate from a hidden fire in an aircraft cabin, so an optimum location for a fire detector can be determined.
Any concealed fire within an aircraft cabin jeopardizes in-flight safety and can lead to catastrophe. Although existing smoke detectors alert a flight crew when an on-board fire ignites, more immediate detection would save precious time. Tracking the combustion path of a hidden fire would give researchers optimum locations for detector installation that would offer more immediate alerts.
Guo’s computer-modeling project focuses on the overhead area in the crown of a Boeing 747-SP, long used by Fire Safety for full-scale fire testing. To acquire the 747’s interior geometry for the simulations, team members scanned the overhead cabin space using light detection and ranging technology to generate a 3-D computer-aided design model.
Next, the team used NIST’s Fire Dynamics Simulator to perform numerical simulations of heat and smoke movement from a fire and entered the data into the computer model. The simulator is a computationally efficient code for low-speed, buoyancy driven flows specifically developed to simulate heat and smoke emanation.
The modeling couldn’t accurately simulate smoke movement in cluttered spaces and curved surfaces of aircraft interiors, so Guo and his team collaborated with NIST scientists to evaluate a new, unstructured geometry capability in the simulator that would better replicate the 747 interior geometry.
“Prior to our study,” Guo said, “the shape of the volume elements had to be rectangular, which became problematic at the curved surface of a cylindrical aircraft fuselage.”
Guo combined the updated simulator data with the cabin dimensions data and ran fire simulations on the FAA’s high-performance computing facility. The facility’s massive parallel processing significantly reduced the time needed to crunch the project’s numbers. The computational fluid dynamics simulation generated a time-dependent temperature field spreading from a hidden fire that identified the optimum location for early smoke detection.
To validate the simulation, Crowley and Scrofani performed full-scale fire tests on the 747-SP, essentially outfitting the airplane with an interior and setting it afire. The temperatures recorded in the actual fire tests matched the computer simulations within 5 degrees Celsius, on average.
This level of correlation validates computational fluid dynamics for large-scale fire modeling. The validation will let Guo expand the simulations to cover environmental conditions and locations that cannot be tested at full scale — for example, in-flight fires at cruise altitudes.
“We recently finished two new validations of different fire sizes and different fire locations with the same model,” Guo said.
Developing this capability for numerical simulation of aircraft cabin fires is part of an ongoing effort to certify aircraft for fire safety by analysis and expand the range of scenarios beyond those that can be investigated at full scale using ground-based testing.
In addition, the Fire Safety team’s close collaboration with Marcos Vanella from the NIST Fire Research Division will soon deliver an updated version of the computational fluid dynamics software and fire dynamics simulator to handle complex aircraft geometries, which is currently undergoing substantial development and will significantly benefit the fire-safety community. “Our case is sufficiently complex, and the NIST will use it as a validation case study when it officially releases its software,” Guo said.
Using computational fluid dynamics simulation saves time, reduces costs, and expedites technology transfer to enhance aircraft safety. Simulating cabin fires also minimizes the physical risks to the investigating team members carrying out full-scare fire testing, simplifies root-cause analysis, and preserves the test article.
“However, a careful [computational fire dynamics] research project cannot live without adequate model validations,” Guo said. “By combining both simulation and experiments, we can bring more insight into fire research and benefit aircraft safety.”
