Development of a Testing Protocol for Glass Window Failure under Thermal Loading

Photo of Sarah Mladenka (Team Leader), Walter Romero, Chris Yeldell Students: Sarah Mladenka (Team Leader), Walter Romero, Chris Yeldell

Sponsor: Los Alamos National Laboratory

Date: Spring 2011

Requirements:
Since characterization of the fire is essential, we need to understand the temperature distribution on the surface of the glass, as well as the kinetics of the fire inside the box. Therefore, we attached an array of 20 thermocouples on the surface of the glass to measure glass temperatures, as well as 10 additional thermocouples inside the test setup to measure gas temperatures. Because we are testing glass in a small scale glove box, we are limited by the size of the fume hood in which we run our tests.

Problem:
Glove boxes are used in a variety of applications around the world to process materials in protected environments. Los Alamos National Laboratory (LANL) utilizes glove boxes to process radioactive materials. Glove boxes are built with fire prevention systems integrated into the design. However, fires can occur when these systems are not functioning properly. The weakest point in glove box design is the glass itself. Once the glass fails, air is introduced into the combustion process and radioactive materials are introduced into the surrounding environment. As a result, lives are at risk from the exposure to radiation and toxic gases. LANL requested that we design a test setup that will characterize the conditions of a glove box fire and the effects of these conditions on glass failure.

Solution:
Our overall goal for the project was to provide Los Alamos with a framework for determining when glass will fail. To accomplish this, we first researched glass failure theories and found several models that are used to predict glass failure depending on the conditions of the glass and the stresses acting on the glass. We then ran physical experiments and input this data to a MATLAB program for analysis. To model the kinetics of the fire, we ran simulations in Fire Dynamics Simulator (FDS) in order to match the temperature readings of our actual experiment to these temperatures in the simulation. We used FDS because, if the gas temperatures and glass temperatures match, then FDS models the heat release rate and heat flux of the actual fire. We postulate that the heat flux could be used in a Finite Element Analysis model to determine the stresses in the glass at the time of failure. These stresses could then be used in one of the glass failure models to determine if the glass will fail.

Images related to the project:

Photo related to Development of a Testing Protocol for Glass Window Failure under Thermal Loading project
Photo related to Development of a Testing Protocol for Glass Window Failure under Thermal Loading project
Joomla SEF URLs by Artio