Technique Developed To Determine Life Expectancy Of Fire-Damaged Concrete

of evaluating fire damage to concrete, enabling contractors and building owners to make a considered choice between concrete repairs and demolition.

A research team has developed a method of evaluating fire damage to concrete, enabling contractors and building owners to make a considered choice between concrete repairs and demolition.

A research team led by Dr Youngsun Heo and working for the Korea Institute of Civil Engineering and Building Technology have developed the first forensic tool that can be used to determine the remaining integrity of a concrete structure damaged by fire.

Concrete as a building material does not burn and can sometimes look undamaged in a fire. However, the intense heat that burns in the centre of a blaze can cause evaporation of the water that is part of the cement-making process.

This can, in turn, lead to concrete cracks and a dramatic loss of strength, which can manifest in the form of sudden collapses which can potentially put more lives in danger than the initial fire.

In a laboratory, some tests can be undertaken such as oxygen measurements, compression tests and ultraviolet spectrum scanning, but the problem with these methods is a lack of reliability and an inability to apply these to an actual scene of a fire.

The solution was an analysis tool known as the Fire Forensic Investigation of Structure (F2IS), which takes small samples of each depth of a fire damaged piece of concrete but can use this to predict the temperature the concrete was damaged at.

It then follows an analysis process using deep learning algorithms to work out which key features need to be addressed, as well as simulating scenarios for how and when the fire started.

It does this with a four-step process, involving the use of practical sampling design (PSD), followed by experimental design (ED) which helps determine the right chemical tools to help analyse specific concrete samples.

Next, a chemical profiling algorithm (CPA) helps to isolate the key points from the chemical results before finally a deep learning interface algorithm (DIA) is used to compare the results with standard data found in a huge database.

One of the main ways of telling the potential for a concrete structure to collapse after a fire is the volume of pores found on the concrete surface, which can be a telltale sign that the concrete has been fundamentally compromised.

The process takes two weeks to complete, which is faster than currently existing systems and could potentially help business continuity, as it allows for vital repair or demolition work to be consulted on and started quicker.

The team have received a patent for the F2IS tool and plan on commercialising it to help with fire forensic investigation, criminology and for contractors to determine whether a structure is still sound after a fire.

Because fires are so innately destructive, they can often destroy a lot of evidence of a fire’s cause, which can make it difficult to determine liability or even a positive cause without other forms of evidence, allowing for appropriate courses of action to be taken to rectify a situation.

It is potentially safer as well, as concrete damaged in this way can sometimes dramatically lose strength within two weeks in particularly serious cases, emphasising the need for quick analysis to allow for decisions to be made before further damage is caused by the delay.