Caulking of Concealed Connections at Glulams

I am not sure about this. I assume someone from timber lab or Simpson will know. I believe EU manufacturers also has similar products, and there are some fire tests done on these. Let me pull some people into this thread.

Hi Connor -- The NDS (I'll use 2024 edtion, so Section 16.5) requires the connections of fire rated beams to be fully fire protected to match the fire rating of the beam.  Gaps and abutted intersections at the intersection of the beam to column face.  This is all in the Code and the structural engineer needs to take on the fire protection specification of structural members (architects don't do NDS engineering).  Simpson strong tie has performed fire testing and uses an intumescent strip in the gap as the fire seal.  Simpson has posted the testing summary that provides all of the backgrounds I've summarized here and is required reading.  This is an essentail topic!  So thanks for asking the question.  https://ssttoolbox.widen.net/view/pdf/lwpfsu3u5h/CBH_SimpsonStrong-Tie_FireTestingSummaryReport_Issue2.0.pdf?t.download=true&u=cjmyin

There are some supplement information on this:

Simpson connection details are very specific about what is required to make a qualified connection assembly that replicates our fire testing.  The detailing requirements on how a product is to be used should be very clear.  If caulk is not called out in the assemblies then it is not needed. But if it is called out, then they should be used. The example of this connection including the test information can be found here:

https://ssttoolbox.widen.net/view/pdf/fuztymvarp/L-C-HSKPFIRE25.pdf?t.download=true&u=cjmyin

Based on another expert friend (Laura from FPL) I have:

Below are two more references on the topic. I have also included Erica Fisher on this email chain in case she has more to share since OSU + REACTS + FPL are collaborating on a project that is digging into the fire dynamics within the gaps that occur at connections to evaluate the effect of various gap sizes (not caulked) on the charring/temperatures at those connections.

• IBC Section 703.7 (2021 and 2024 Editions) requires sealant or adhesive, unless not a required component of a tested assembly, be provided to resist the passage of air and they must meet either ASTM C920 or ASTM D3498, respectively.
• AWC’s Fire Design Specification (Section 3.10 from the 2024 Edition) details requirements for the design of protected connections.

This information is from Prof. Erica Fischer of OSU

The IBC states that all gaps must have caulk within them; however, the Fire Design Specification (FDS) provides a bit more information on varying char depths with different sized gaps (Section 3.2.3). These prescribed char depths can be used to evaluate if screw embedment depth and shear capacity of the beam are sufficient. However, most of the research shows that when the gaps are 3 mm and less, the char within the connections is quite small and the presence of caulk doesn’t influence the temperatures within the connections as much as when the gaps are greater than 3 mm (1/8 inch), and particularly greater than 6 mm (1/4 inch).

We did some fire tests on glulam beam-to-column connections, the goal was not to explore the influence of the gap though we did quantify it:

Fischer EC, Bhandari S, Sinha A (2025) Fire Testing of Glue-Laminated Beam-to-Column Connections. Journal of Structural Engineering 151:. https://doi.org/10.1061/jsendh.steng-13888

Recently we led efforts to study gap influences for CLT panel-to-panel connections:

Garrett W, Fischer E, Hasburgh L (2025) QUANTIFYING FIRE BEHAVIOUR IN GAPS IN CLT PANEL-TO-PANEL CONNECTIONS. In: Proceedings from the 14th World Conference on Timber Engineering: Advancing Timber for the Future Built Environment, WCTE 2025. World Conference on Timber Engineering (WCTE), pp 3684–3692

Rothoblaas did a study on gaps in CLT panel-to-panel connections all including fire protection and recently presented it at WCTE this year:

Brugnara P, Alice S, Sestigiani L (2025) FIRE AND CLT: LINEAR JOINTS. In: Proceedings from the 14th World Conference on Timber Engineering: Advancing Timber for the Future Built Environment, WCTE 2025. World Conference on Timber Engineering (WCTE), pp 4235–4241

Another study that is relevant: (more beam-to-column connections)

du Plessis M, Sulon D, Streicher D, Walls R (2024) Experimental testing on timber connections considering the influence of gap size and intumescent sealants. Fire Mater 48:39–61. https://doi.org/10.1002/fam.3164

The thesis/dissertations along with this work is here:

https://scholar.sun.ac.za/handle/10019.1/130182

https://scholar.sun.ac.za/handle/10019.1/130182

MTC has performed some research with this (beam-to-column connections) along with Christian Dagenais – he is presenting it at Woodrise next week, but they don’t seem to have a publication out yet. We just tested glulam beam-to-column connections this summer – and probably won’t have a paper out for a while. We are analyzing the data right now.

I hope that this is helpful!

Erica

Erica C. Fischer, PhD, PE

more information from Scott @ Woorworks:

The slides from David Barber's presentation last week are available at https://www.woodworks.org/wp-content/uploads/presentation_slides_Barber_Fire_Design_Mass_Timber_Connections_09.2025.pdf

We will have a recording of the webinar available on https://www.woodinstitute.org/ in a few weeks.

Scott Breneman, PhD, PE (WA, CA), SE (CA)

WoodWorks – Wood Products Council 

@ling @clenen Here is more relevant testing.  It can certainly be a challenge to pass a fire connection test.   FPInnovations has published these reports from Phase 1 and 2 testing of connections with gaps of various sizes and with and without a fire seal.  It is of note that both of the loaded tests in Phase 2 failed.  You learn from failure, so it is great to have these tests published.

Part 1:

https://library.fpinnovations.ca/link/fpipub11016

Part 2:

https://library.fpinnovations.ca/link/fpipub11128

@heblomgren Thanks a lot! Yes, I am always a proponent for testing your system to failure 😎  Unless it is not allowed by the testing site...

thanks for the help to pull together these data.

Thank you for all the info! It's extremely helpful.