FPInnovations research leader for advanced building systems Conroy Lum was the presenter at the "CLT Firewalls" session at the Wood Design and Construction Solutions Conference held March 1 in downtown Vancouver.
Lum began by explaining FPINnovations has done two service penetration tests and two closure penetration tests will CLT as a firewall or fire separation.
The tests used were a test of firewalls, door assemblies and wall systems.
Mass timber chars when heated, which produces a gas that burns when heated.
"It's not the wood that's heated, it's the gas," Lum said.
Char does not burn and protects uncharred wood. But Char, he said, has no strength and may fall off.
CLT charring in one direction is observed to be consistent at about 0.6 mm per minute, he said.
Possible ways heat may be transferred to CLT include conduction, where a heated object touches wood, convection, when hot gasses move past the wood, radiation (heat through glass), and advection.
For the closure test, it isn't the door being tested but the interface. The service penetration test is from a horizontal floor, and is more onerous than a vertical test.
Generic firestop principles, Lum said, is that penetration such as pipe does not melt but will conduct heat.
"So you use mineral wool to minimize this," Lum said. Caulking is used to hold the mineral wool.
Many temperature measurements are taken with thermal coupling, and after one hour the pipes change color, which means the temperature is starting to affect them.
Once the burning portion of testing is done, the sample undergoes a hose stream test.
"For the most part, the CLT is still there, so what do we see around the penetrations?" Lum asked.
After two hours, Lum said they observed the temperature difference between different firestopping pipes. On the unexposed side, uninsulated copper and cast iron pipes were one inch from the CLT surface. Copper was much hotter after an hour than iron, Lum said, but after two hours both materials are near the same temperature.
Uninsulated copper shows no surface charring, while with insulted copper did show some charring.
Closure penetration tests consisted of a CLT wall assembly with one service penetration. It was assembled from four pieces, and double spline joints were used. There are also gypsum wildboard caps on the side and top.
Door frame anchors were filled with mineral wool, then mounted on the frame. Instead of relying on caulking for fire resistance, Lum said, they opted to make the door frame anchor fit airtight. Mineral wool was used behind the fame.
"You don't want the anchor screw to go into the char layer, because the moment it does, it conducts heat," Lum noted.
After testing and hose stream, the team found service penetration on the side wall did "reasonably well," but it was discovered that one side of the frame shifted under the hose stream while the other side did not.
"It had something to do with the anchor arrangement," Lum said.
Concepts confirmed in the test were that char is isolated to areas around areas of penetration. Copper will need extra insulation compared to steel. They also found the tested annular space of 05. Inches is a minimum.
Guiding design principles for CLT penetrations are "manage heat transfer and protect the flanks," Lum said. "Shield, seperate and insulate."