Fire Performance of CLT Wall and Floor Assemblies Fabricated in the Pacific Northwest Region 


Abstract:

Despite a growing body of empirical evidence generated by European, Japanese and Canadian research on the fire endurance of cross-laminated timber (CLT), a lack of full-scale U.S. testing of structural CLT manufactured within the U.S. is often cited as a major barrier to approval of the new building material for use in tall structures.  Past testing in test furnaces suggests that CLT and glue-laminated timber can outperform both light-frame timber assemblies and steel and concrete elements. This is due, in part, to the fact that in thick CLT panels a layer of charred wood forms on the exposed surface that then serves to insulate and protect the wood behind the char layer. This project will document the flammability of Douglas-fir and spruce-pine-fir CLT panel assemblies produced in the United States. Tests are being conducted on wall and floor panel assemblies with standard overlapping connections and produced with two different types of commonly-used adhesives. Sensors placed throughout panels will provide data about how fire affects the interior and exterior of a panel. A thermal imaging camera will provide information on how the structural integrity of panels is affected by fire and fire suppression activities. When this research project was proposed in 2015, no full-scale fire performance testing of U.S.-made, structural CLT had been performed.

 

Project Duration: 2016 – 2020

 

Research Team:

Lech Muszynski, Department of Wood Science & Engineering (PI)

Rakesh Gupta, Department of Wood Science & Engineering (co-PI)

Milo Clauson, Department of Wood Science & Engineering (FRA)

 

Collaborators:

Dr. David Blunck (OSU MIME Propulsion Laboratory)

Brent Pickett (Western Fire Center in Kelso, WA)

 

Graduate Research Assistants:

Neil Osborn – OSU, Wood Science and Engineering (September 2016 – August 2017)

Seung hyun Hong – OSU, Wood Science and Engineering (since September 2018

 

Undergraduate student researchers:

Jacob Newton (2017)

Marilee Hoyle (2017)

Cassie Holloway (2017-2020)

Joshua Curfman (2018-2020)

 

REEU students involved:

Ben Collar, a REEU intern from CC Roseburg, OR, (Summer 2018)

Lauren Prox, a REEU intern from George Mason University in Newport, VA (Summer 2018)

 

International exchange students:

Miloš Pavelek, a PhD student at Technology and Mechanization in Forestry, Faculty of Forestry and Wood Sciences, Czech University of Life Sciences Prague (September-October 2016)

Mara Gruber, a MS student from HFS Wels in Austria (Summer 2017)

Matthias Kleissl, a MS student from HFS Salzburg in Austria (Summer 2018)

 

Research Advisory Group:

Dr. David Barber, ARUP

TallWood Design Institute

 

Facilities:

Western Fire Center in Kelso, WA (Fall 2016, Winter 2017)

Oregon State University (WSE labs, Oak Creek facility)

 

Introduction:

An important area of concern for building code officials is fire safety, and there is very little data in the U.S. that documents the performance of CLT panels exposed to fire. This project will document the flammability of Douglas-fir and spruce-pine-fir CLT panel assemblies produced in the United States. Tests are being conducted on wall and floor panel assemblies with standard overlapping connections and produced with two different types of commonly used adhesives. Sensors placed throughout panels will provide data about how fire affects the interior and exterior of a panel. A thermal imaging camera will provide information on how the structural integrity of panels is affected by fire and fire suppression activities.

 

Research Details

Objectives

    • The objective of the proposed research project is to investigate the fire performance of 5-layer Douglas fir CLT wall and floor assemblies made in Pacific Northwest Region. The specific objectives are:
    • Determine fire performance for unprotected and fire protected wall and floor CLT assemblies according to ASTM E119 standard.Measure the out of plane deformations of wall and floor elements related to the fire and wetting events and evaluate their effect on the building stability.
    • Assess the impact of fire on the local integrity of the CLT layup.
 

Primary Findings:

    • All three unprotected CLT floor assemblies and two of the three wall assemblies produced by American manufacturers passed the 2-hour fire resistance test following ASTM E119 standard procedure. One assembly passed 100 min mark before the diaphragm was breached.
    • Despite the charred layers fall off observed in assemblies bonded with PUR adhesive system the differences in mean char rates for the first two layers in floor assemblies using PUR and MF systems were smaller than the variability of char rates within individual assemblies (as reflected by the standard deviation).
    • The effect of softening of the PUR bonds is most apparent as an accelerated rate of the floor assembly deflection in the final 40 minutes of the tests.
    • Within the 2-hour standard test exposure the unprotected half-lap joints provided adequate barrier against transmission of hot gases and flames through the assemblies.

 

Results:

All three unprotected CLT floor assemblies and two of the three wall assemblies produced by American manufacturers passed the 2-hour fire resistance test following ASTM E119 standard procedure. One assembly passed 100 min mark before the diaphragm was breached.

Despite the charred layers fall off observed in assemblies bonded with PUR adhesive system the differences in mean char rates for the first two layers in floor assemblies using PUR and MF systems were smaller than the variability of char rates within individual assemblies (as reflected by the standard deviation).

The effect of softening of the PUR bonds is most apparent as an accelerated rate of the floor assembly deflection in the final 40 minutes of the tests.

Within the 2-hour standard test exposure the unprotected half-lap joints provided adequate barrier against transmission of hot gases and flames through the floor assemblies.

In wall assemblies, the unprotected half-lap joints retained full integrity over 100-minute exposure. In 2 out of 3 walls tested minor breaches beyond the 100-minute mark did not result in transmission of hot gases or flames through the joint.  

Char fall-off events during the standard fire performance tests of flammable barriers like CLT panels may be accurately detected and timed by analyzing disturbances in the furnace temperature readings recorded during the tests.

The char rate variability during the standard tests can be determined continuously by analyzing temperature readings from the embedded thermocouples.

 

Reports:

  • Pickett B. (2016): Fire Resistance Testing of CLT Floor/Ceiling Assemblies, Western Fire Center Inc., Western Fire Center Report #16046a, 2016, pp. 25 pp.

  • Pickett B (2017): Fire Performance of ASTM E119 Evaluation of CLT Load-Bearing Wall Assemblies, Western Fire Center Inc., Western Fire Center Report #16046b, 2017, pp. 29 pp.

  • Holloway C. (2018): Paper beats rock: does wood beat steel? WSE401 project report. OSU. 46 pp (a undergraduate research project report describing the methodology of measuring char depth across the wall and floor specimens exposed to fire tests, and processing the data for the char maps. Supervised by L. Muszynski)

 

2019 Mass Timber Conference Poster: Fire Resistance of Unprotected Cross-laminated Timber Wall Assemblies Made in the USA (Seung Hyun Hong et al., M.S. Wood Science 2019)

Funding Acknowledgements:
Funding for this project was provided by the USDA ARS 2015 program, via internal competition administered by COF/IWFL/NCAWPMDD