Fire Resistance Testing

Heat testing is one of the main tests for fire-rated glazing and fire-resistive assemblies. For glazing, glass panels in frames are subject to heat that gradually approaches approximately 1,600 degrees F. For its rating, the glass must stay fixed in its frame for the length of time: 20 minutes, 45 minutes, and so on.

A second test, the "Hose-Stream Test" (see section below) is used to evaluate the material's structural behavior at high temperature. In this test, the glass panel is subjected to a high-pressure stream of water from a fixed hose. (Years ago, weights were used to impact the specimens.)

Glass may also be tested and receive ratings for other attributes, including impact force.

Fire resistance ratings for building materials and systems are referenced primarily to ASTM E 119, Standard Test Methods for Fire Tests of Building Construction & Materials, a version of which was first drafted in 1918 under a different name. The standard is used to test numerous building systems made of any combination of materials-such as steel, concrete, gypsum, wood and glass-for a variety of building assemblies: columns, beams, roof-ceiling systems, bearing walls and non-load-bearing partitions.

The standard measures the fire-resistive properties as expressed by time: "the period of resistance to standard exposure before the first critical point in behavior is observed."

A controlled fire-usually gas burners-is applied to the system or element. The heat produced follows a fixed time-and-temperature curve given in the standard. At first it rises quickly, and then it gradually increases-simulating the intense early stages of a fire. The test is configured for realistic condition, and the size of the specimen must meet a minimum.

The heat transmitted through the sample assembly is a critical measurement. If the average temperature in any area on the unexposed side rises more than 250 degrees F above ambient conditions on the same side, the test ends. After passing the specified ASTM E119 standard which includes the hose stream test, the specimen is given a fire-resistance classification, or rating based on that test. The ratings are expressed in hours, so that architects can compare building materials and systems and understand their applicability to building codes.

The Hose-Stream Test

One of the testing standards applied to the testing of fire-protective glazing and fire-resistive glazing is the "hose stream test," so named because water under pressure from a hose is used to test the building materials in question. For that reason, an understanding of the test and its significance is important.

The test has been a subject of debate almost since its inception. In the 1950s, the requirement for the hose-stream test was eliminated from the fire testing of floor/roof-ceiling systems. At the time, it was determined that "no record of failure in the hose-stream test" had been found due to holes in floors. In Europe and in other countries, the hose-stream test is not used. British officials eliminated the hose-stream test entirely from fire-testing standards in the early 1960s. But in the United States and Canada it is still used for fire-rated glazing.

The hose-stream test was created back in the late 1800s to test the integrity of cast iron and wrought iron structural systems. Because cast iron becomes brittle when heated by fire and fails as a result of thermal shock after being exposed to even small amounts of water, the first ASTM standard for fire performance-ASTM C19-1917T, now known as standard E119-included a hose stream test. However, according to a recent work summary published by ASTM subcommittee E05.11, no documented reasons were ever given for the inclusion of the hose stream test.

The test was NOT designed to assess how building systems would be affected by firefighting methods. (In fact, weights were originally used to impact the assembly, not water.)

Instead, the test offered a way to apply a consistent exposure to effects that cause failure, applied easily and evenly across the assembly. In this way, the test has been known to indicate two key characteristics:

• an assembly's integrity under fire exposure, and
• the integrity and reliability of materials to perform their intended functions.

For glass, the basic standard involves subjecting a sample of fire-rated glass to a test furnace and then an even hose stream across its surface. To pass the test, the glass should remain intact and not exceed the level of allowable through-openings.

Questions about the Hose Stream- for Fire-Protective Glazing Under 1 Hour

There is considerable debate about the test's value and reliability for predicting the safety of fire glass. First of all, it was designed for evaluating iron and other structurally supporting building members, not glazing. Second, European and British building agencies have not used the hose-stream approach for decades. Third, many codes officials, fire-protection engineers, and glass manufacturers have publicly questioned its efficacy in measuring glass fire safety.

Some observers of the hose-stream test point out that the test helps demonstrate the potentially early failure of glass when in areas with fire sprinklers. While this approach to testing for thermal shock resistance sounds reasonable, there is no evidence or reference in fire-testing standards or other building industry documentation. And the test is applied following a long exposure of the material to a very high temperature-45 minutes at approximately 1,600 degrees F. (Actual temperature at 45 minutes is 1638 degrees F. per NFPA test procedure.) In a real-world building, the sprinklers would activate after only a few minutes of fire.

Moreover, the hose-stream test was simply never intended as a way to rate glazing for thermal shock due to heat and fire sprinklers.

Glazing materials that fail the hose-stream test are not likely to fail during a fire when the sprinklers are activated. In fact, glass generally will resist failure as a result of water exposure from sprinklers during the first few minutes of a fire. Case reports and test data support this conclusion.