172 - Lessons from mass timber experiments with Danny Hopkin

Transcript

WEBVTT

00:00:00.040 --> 00:00:01.824
Hello everybody, welcome to the Fire Science Show.

00:00:01.824 --> 00:00:05.873
I'm really excited today because I got a really special guest for you.

00:00:05.873 --> 00:00:14.894
That is Dr Danny Hopkins, director at the OFR, and the last time Danny was in the podcast like 150 episodes ago, that was episode 18.

00:00:14.894 --> 00:00:17.342
That was quite something.

00:00:17.342 --> 00:00:27.082
We broke all the listenership records for that podcast episode and it has been the most downloaded episode of the show and stayed like that for like two years.

00:00:27.082 --> 00:00:34.442
That was insane and just showing how much people were interested in timber and fire, mass timber and fire.

00:00:34.442 --> 00:00:41.764
That was what we've been discussing in episode 18 and that's what we are discussing in the episode today.

00:00:41.764 --> 00:00:57.868
Back then we were just after first round of our big experiments, or perhaps in the middle of it, and today, sometime later, we finished our biggest research program together and we have some findings.

00:00:57.868 --> 00:01:05.712
We have some findings to share with you, and in this episode we of course delve deep into the experiments that we have performed together.

00:01:05.712 --> 00:01:10.512
But it's not just an experimental overview episode.

00:01:10.512 --> 00:01:17.251
Actually, it's pretty far from that, because the experiments were done in a full scale.

00:01:17.251 --> 00:01:30.846
The observations carried in the experiments that we are discussing are pretty general and pretty relevant to all mass timber structures, especially heavily compartmentalized timber structures like residential buildings.

00:01:30.846 --> 00:01:49.412
The observations we've carried are like observations of fires in real buildings and I think that's how we try to discuss it with Danny, that's how we try to view it within the research group and with our industrial partners in the Structural Timber Alliance, that's Storenzo, klh, binderholz, henkel and Fermasol.

00:01:49.412 --> 00:01:59.390
That was the point of the research we've carried to create some answers to the real world problems that we're facing when designing mass timber structures.

00:01:59.390 --> 00:02:05.084
And in this episode, answers is what you will find and in this episode, answers is what you will find.

00:02:05.084 --> 00:02:05.945
So, hopefully, a lightning.

00:02:05.945 --> 00:02:24.955
Very interesting episode for you all, for me a source of great pride because those experiments were carried at the ITB laboratory, and it's just so much fun when you carry large research programs with people like Danny and all the team that has been involved in doing this.

00:02:24.955 --> 00:02:29.651
So, without further ado, let's spin the intro and jump into the episode.

00:02:29.651 --> 00:02:36.568
Welcome to the Firesize Show.

00:02:36.568 --> 00:02:40.082
My name is Wojciech Wigrzynski and I will be your host.

00:02:53.092 --> 00:03:09.930
This podcast is produced together with ofr, uk's leading fire and risk consultancy, but you kind of know that already and this time, because it's a special episode, I'm going to step away from the copy and just say a few words from myself about my sponsor of RFR.

00:03:09.930 --> 00:03:19.766
I've been working with them for more than two years on this project and for many, many years in my laboratory doing research like one that we're discussing in this episode.

00:03:19.766 --> 00:03:24.924
If I could describe them with one word, that would be eminence, or maybe excellence.

00:03:24.924 --> 00:03:30.675
When we've signed a contract for the 5 Cents Show, that's what they told me they are looking for.

00:03:30.675 --> 00:03:34.670
They want excellence, and they found that in the podcast, and they want to be a part of it.

00:03:34.670 --> 00:03:40.451
They want to support it, and they've done for so much time more than 100 episodes produced together.

00:03:40.451 --> 00:03:43.108
They've never interfered with the content.

00:03:43.108 --> 00:03:46.782
They've always supported me and I can be just thankful.

00:03:47.164 --> 00:04:26.420
But having worked with them professionally, not just in this gig, I can also tell you these guys are looking for science-based engineering and they're very serious about it, and everything we've done together in the research area was to create knowledge that's missing in the market, that's so much needed for designing a next generation of structures, including mass timber, and I think this is an amazing feat for an engineering company to invest so much in research, in understanding and then using this knowledge to design better and provide better services to their clients.

00:04:26.420 --> 00:04:38.062
So just because of that, I'm very amazed with OFR and I think we're a great match for the podcast, and I'm super happy that today the podcast can be a vessel to translate some of this new knowledge to you.

00:04:38.062 --> 00:04:42.033
So, yeah, let's hear what Danny can tell us about mass timber and fire.

00:04:42.033 --> 00:04:44.504
Hello everybody, welcome to the Fireside Show.

00:04:44.504 --> 00:04:47.031
I am here joined by Dr Danny Hopkins.

00:04:47.031 --> 00:04:48.406
Hey, danny, good to have you back in the show.

00:04:48.560 --> 00:04:49.427
Very good to be back.

00:04:49.427 --> 00:04:50.091
Thank you for having me.

00:04:50.480 --> 00:04:54.687
150 episodes have passed even more than that since the last time we've been here.

00:04:54.687 --> 00:05:00.081
We've broke some records back then, so my expectations towards this episode are quite high, danny.

00:05:00.322 --> 00:05:02.923
Well, I mean that's a lot of pressure.

00:05:02.923 --> 00:05:08.204
I think you've had a lot of good speakers since the early days when I managed to grab the top of the pops.

00:05:08.985 --> 00:05:14.187
And I must start by appreciating the role of OFR in running this project.

00:05:14.187 --> 00:05:27.435
It's been extreme help and makes me super happy to produce the show together with you guys, and it has been more episodes we've produced together with OFR than episodes I've made myself before you joined me.

00:05:27.435 --> 00:05:28.175
That's amazing.

00:05:28.175 --> 00:05:29.255
Thanks for that man.

00:05:29.295 --> 00:05:30.055
You're very welcome.

00:05:30.055 --> 00:05:41.209
The show is a really important resource for our engineers as a CPD tool, and I suspect the same holds for many other organizations, so we're very happy to support it and see it flourishing.

00:05:41.711 --> 00:05:46.391
Yeah, fantastic, let's make it flourish with some hardcore science today.

00:05:46.391 --> 00:05:57.608
And I've invited you to talk over a very interesting experimental program that we've kind of did together in ITB Labs and you were supervising that.

00:05:57.608 --> 00:06:00.788
We've done it for the Structural Timber Alliance.

00:06:00.788 --> 00:06:04.165
I hope I did not misspoke the name of the sponsors.

00:06:04.439 --> 00:06:16.182
It's an association association, but yeah, you got the acronym correct hi, close enough, close enough, um, and it was encapsulated timber and there's so much we we've run through.

00:06:16.182 --> 00:06:20.951
Perhaps let's briefly talk over the idea behind the experiment.

00:06:20.951 --> 00:06:22.194
So we were.

00:06:22.194 --> 00:06:30.168
You approached me that you want to burn multiple cubicles of of timber, with different strategies of how the walls and ceilings are protected.

00:06:30.168 --> 00:06:40.161
Tell me, like, how it made its way towards you and what were the initial assumptions you wanted to to confirm or discredit with this experiment sure.

00:06:40.182 --> 00:06:43.108
well, this goes back to 2019 actually.

00:06:43.108 --> 00:07:05.826
Actually, I think it was when the CLT industry were getting an awful lot of sort of obstacles that were preventing in particular, clt buildings from being constructed, because, at that point, sort of this consensus developed that mass timber structures perform different in fire to non-combustible structures and they present some unique hazards that needed to be addressed.

00:07:05.826 --> 00:07:16.382
And so, at the point of accepting that those hazards exist, the next sort of logical step is what evidence, what information can we develop that allows us to design in a more responsible way?

00:07:16.382 --> 00:07:28.130
And part of that is sort of designing an envelope within which structures might be able to survive burnout, which I think we've sort of spoken about ad nauseam across various podcasts in in this series.

00:07:28.130 --> 00:07:34.473
And so, um, before I go any further, I guess first thing I should do is acknowledge the great support we had from the industry.

00:07:34.473 --> 00:07:52.903
So, whether it was store renzo, klh, binderholz, um, henkel and firmacel, under the remit of the structural timber association, that allowed us to produce this work, and so this is, this particular paper is part of a lot of work packages that we've delivered.

00:07:53.444 --> 00:08:02.134
In terms of the specific motivation for this one, one of the very early applications across laminated timber was actually in in residential buildings.

00:08:02.134 --> 00:08:14.334
It was to build highly cellular buildings, sort buildings that apartment buildings tend to lend themselves to, and actually some of the tallest buildings in the world that were constructed out of mass timber were residential buildings.

00:08:14.334 --> 00:08:48.000
So things like Murray Grove in London in the early 2000s and Dalston Lane and obviously Greenfield happened and the implications for that at a regulatory level in England in particular was to introduce what I kind of refer to as the in-effect ban on combustible materials in the external wall zone, which is to say that for relevant buildings, which is mainly areas where you have a sleeping risk over 18 meters, that you can't have materials in the external wall that are combustible and we use Euroclass A1 and A2 to define that in the external wall that are combustible and we use Euroclass A1 and A2 to define that.

00:08:48.000 --> 00:08:52.025
So that created a bit of a challenge for combustible structures because you could no longer have combustible structure in that external wall zone.

00:08:53.460 --> 00:08:54.147
A clarification.

00:08:54.147 --> 00:09:04.812
But is that understood as you cannot have any combustible material, or are you referring to the surface of the material, Because in Poland we had like a total combustible ban nothing, nowhere.

00:09:04.745 --> 00:09:06.634
No, so this is in the external wall zone, specifically of the material, because in poland we had like a total combustible ban nothing, nowhere, uh no.

00:09:06.567 --> 00:09:19.469
So this is in the external wall zone specifically, and if your structure happens to sit within that external wall zone, so that's from the sort of outer surface of the external wall to the inner sort of warm face that you can touch, um, nothing within that zone.

00:09:19.929 --> 00:09:28.072
There is a few exceptions that are listed in regulation 7-2, things like membranes and such like, but sort of non-substantive components are exceptions.

00:09:28.072 --> 00:09:45.179
But if your structure is in that zone then it would also have to be a fewer class A1 or A2, which is a real headache for a combustible structure that relies on walls and slabs to sort of generate its structural system, to sort of generate its structural system.

00:09:45.179 --> 00:09:51.153
And then you kind of take that and you sort of combine it with a general mistrust around our ability to work with combustible materials.

00:09:51.153 --> 00:10:02.559
You create a real dip justifiably so, I would say in confidence in building with anything combustible, whether it's ACM or whether it's mass timber, as your structural frame.

00:10:02.559 --> 00:10:13.888
And so we were tasked with trying to develop this envelope that we think we can deliver mass timber residential buildings within, and that's where the experiments came in.

00:10:14.480 --> 00:10:24.732
We've done, also experiment we refer to as Work Package 6, where we've burned big slabs of CLT and then tried to capture the effects of the glues.

00:10:24.732 --> 00:10:27.860
In that experiment we had one massive opening.

00:10:27.860 --> 00:10:33.373
Basically we just had a slab hanging and one entire side of the wall was open.

00:10:33.373 --> 00:10:58.109
In the cubicle experiments the work package five as we call it internally, the one that we refer in this podcast episode it was a small compartment with just one opening doors leading to the compartment, a completely different ventilation, let's say, setting of the fire clearly something we wanted to flash over and have a prolonged fire in that compartment.

00:10:58.109 --> 00:11:02.648
What was the reasoning for for this type of exposure in this experiment?

00:11:03.109 --> 00:11:11.587
it's really just a reflection of the difference in the architecture between what was Web Package 6 that you refer to, that we published in the FIRE journal.

00:11:11.980 --> 00:11:29.429
That was very much aimed at commercial buildings, so offices and retail in particular, and actually more so hybridized structures, where actually your mainframe might be a steel and a concrete and you're using your CLT as a substitute for where your concrete slab would be.

00:11:29.429 --> 00:11:35.460
And actually in those situations you're generally talking about much larger compartments with larger opening factors.

00:11:35.460 --> 00:11:38.726
If we use the Eurocode sort of terminology and nomenclature.

00:11:38.726 --> 00:12:03.404
When you have residential enclosures, you are generally talking about much smaller compartments or much smaller rooms within a compartment with smaller window openings, and that kind of lends itself to arguably a different type of fire most of the time, which is one that tends towards sort of flashover, and then that sort of post-flashover, fully developed phase that we tend to think of when we think of a compartment fire.

00:12:03.404 --> 00:12:14.011
So yeah, it's very much a response to the fact that the architecture is different in a residential building, to a, to a commercial building well, why didn't you just put the walls in the furnace, like?

00:12:14.494 --> 00:12:18.993
we could also do the fire resistance test, and I'm happy to do it, by the way, if you want but why?

00:12:18.993 --> 00:12:24.971
Why we had to build a natural compartments and and look at this in this setting it kind of comes back to.

00:12:25.030 --> 00:12:42.631
I think it's it was podcast 18, where I came on and and spoke a bit about what we're trying to achieve with mass timber buildings and when we are building let's call them higher consequence of failure buildings and and I use that um quite loosely, but but historically you might say it's anything over around 18 meters, something like that.

00:12:42.631 --> 00:12:57.066
If you look at the postal building studies in those kind of buildings you are implicitly, through our sort of regulatory history, setting yourself the objective of the structure surviving the full duration of the fire without intervention or burnout as we refer to it.

00:12:57.066 --> 00:13:20.769
And so the only way you can really understand that is to take a proper enclosure experiment, put some kind of fire load or fire exposure within it, heat the structure, take it through its growth, steady and decay phases and then actually observe how that structure responds in terms of its ability to stop flaming, to stop smoldering or otherwise, as was often the case in our experiments.

00:13:21.309 --> 00:13:44.842
It's what allows us to draw some analogies to what we expect by way of outcomes for for concrete buildings and for steel building I'm asking that because you know for, like you said, concrete or steel buildings, the fire resistance would be in a way a proxy of that which you just described, maybe minus the cooling phase, because you of course do not capture that in a furnace as you would in a building, the.

00:13:44.842 --> 00:13:49.573
The cooling process is a little different in a well-insulated furnace environment.

00:13:49.573 --> 00:13:58.279
But for timber and I've also wrote papers about that fire resistance is a really bad proxy of the fire behavior that you would expect from the timber.

00:13:58.279 --> 00:14:06.028
I had this episode with Daniel Brandon where we went deep into that Fire resistance tests are not made for timber.

00:14:06.028 --> 00:14:18.659
Really, a straight secret of the laboratories is that in tests of exposed timber you're using your furnace as a cooler to the fire, not a burner, and that's a pretty bad use of a burner.

00:14:18.659 --> 00:14:20.827
So I resonate with that.

00:14:21.139 --> 00:14:38.120
If you want to demonstrate this throughout the phases, like the furnace is not really giving you those answers well, I think, I think we can be stronger than that and say sort of with great certainty that fire resistance testing was not built to address combustible structures.

00:14:38.120 --> 00:14:49.505
In fact, if you, if you go back to its genesis, after the many great fires in north america, it was there to give some credence to the fireproof credentials of actually alternatives to wood.

00:14:49.505 --> 00:14:57.190
It was the concrete floors, the very early reinforced concrete floors, the masonries, the cast irons and such like.

00:14:57.190 --> 00:15:04.893
So actually the combustible structures were very much not part of the thinking when fire resistance testing was developed.

00:15:07.840 --> 00:15:13.687
Minus, let's say, k to 90 minutes, where you put so much fire protection in front of your timber that it never gets involved right.

00:15:13.687 --> 00:15:21.139
That's also not that interesting case for the mass timber buildings, because it leads to some sort of non-optimal use of fire protection.

00:15:21.139 --> 00:15:32.062
I don't even know if you can claim that, but definitely putting three layers of plasterboard on everything you build is not the most efficient thing you can do if we go back a few steps.

00:15:32.202 --> 00:15:38.302
I mean, if you want a timber structure to have a good chance of surviving the full duration of the fire, you've.

00:15:38.302 --> 00:15:41.269
You've kind of got two paths that you can go down.

00:15:41.269 --> 00:15:43.721
One is just to stop its involvement.

00:15:43.721 --> 00:15:56.802
Its involvement is a source of fuel and what we often use to assure that doesn't happen, or have some confidence that that's not going to happen, is we put enough plasterboard on it to make sure that the substrate doesn't begin to pyrolyse.

00:15:56.802 --> 00:16:09.369
So we try and keep it kind of in and around 200 degrees or less, and we can put enough plasterboard in principle on a piece of wood to to make that happen, subject to appropriate detailing and fixing and all that good stuff.

00:16:09.369 --> 00:16:20.587
The alternative route is you accept that your structure is going to become involved and if it does become involved then it's going to have to stop without intervention.

00:16:20.587 --> 00:16:33.910
And that's when we're often talking about auto extinction or self-extinction, and in that instance you've got the two regimes of burning whether it's flaming and it's often easier to to demonstrate that you'll have the cessation of flaming combustion.

00:16:33.910 --> 00:16:49.020
What tends to follow is smoldering, which can persist somewhat indefinitely, and so you can deliver mass timber buildings that can perform broadly equivalently to non-combustible buildings, just by virtue of putting enough protection on them that the substrate doesn't doesn't become involved.

00:16:49.020 --> 00:16:52.307
And that was, in effect, our reference experiment.

00:16:52.307 --> 00:17:01.721
What we did is we took our clt box which was I've got here 3.4 by 3.4 by 2.5 meters wide it's kind of broadly a door shaped opening in it.

00:17:01.721 --> 00:17:08.807
Um, we put some propane burners in to deliver a heat release rate that just got us on the brink of a ventilation control fire.

00:17:08.807 --> 00:17:10.701
So that's through a bit of trial and error.

00:17:10.701 --> 00:17:30.048
And then we ran that fire for in and around 90 minutes before stepping it down, and what we were able to show is, if you put enough plasterboard on the walls which in this instance was a combination of plasterboard and gypsum fiberboard we can stop the substrate from becoming involved and the compartment fire behaves in a manner that you would expect of a non-combustible structure.

00:17:30.299 --> 00:17:32.680
But to achieve that requires a huge amount of plasterboard.

00:17:32.680 --> 00:17:40.108
I think we had something of the order of 54 millimetres of plasterboard or gypsum fibreboard.

00:17:40.108 --> 00:18:02.292
So three layers, three 18mm layers, which somewhat undermines the carbon and sustainability credentials of using wood, because you're protecting it with something that is gypsum based and you're having to use a lot of it, and it also compromises things like the amount of floor area you get, because you end up with a very thick wall build up to to achieve it.

00:18:02.292 --> 00:18:12.090
So there is a certain attractiveness in being able to optimize that lining and historically, various schools of thoughts have emerged on this.

00:18:12.090 --> 00:18:26.073
One is that you fully encapsulate, and by fully encapsulate what I mean is every single surface is protected in a manner whereby you are explicitly trying to stop it from becoming involved or pyrolyzing.

00:18:26.073 --> 00:18:42.102
So you've got these three layers of plasterboard, four layers, whatever it might be, on every single combustible surface and that allows you to sort of to design within the domain of fire resistance because your structure is not burning, or you do some kind of optimization, and that optimization can be.

00:18:42.102 --> 00:18:46.593
You have strategic surfaces exposed whilst you stop the others from burning.

00:18:46.593 --> 00:18:50.810
So that might be, in our definitions, partial encapsulation.

00:18:50.810 --> 00:18:58.594
So all of the walls are explicitly designed in a way so they should not become involved as a source of fuel whilst the ceiling is visible.

00:18:59.019 --> 00:19:21.026
But actually, in the early 2000s, one of the more poorly considered strategies, dare I say, was was what I refer to as partial protection, and that's this idea that your fire resistance, achieved by your timber element, can be achieved through a combination of a protection lining and some amount of inherent performance from the substrate, in this case a wall or a slab.

00:19:21.768 --> 00:19:42.037
And so what you will see in the literature and in some of the discussions around existing buildings is this idea that our building got its 90 minute fire resistance by the plasterboard, giving us 30 minutes, and then the clt being designed for charring rates for 60 minutes, 60 plus 30 gives us the 90, and therefore we've we've met our fire resistance goals.

00:19:42.798 --> 00:19:56.881
What that fails to address is that you are accepting that your structure is burning and therefore you are very much challenging the idea that it's going to remain stable indefinitely without intervention, because you've not done that check on auto extinction.

00:19:56.881 --> 00:20:08.931
And so lots of ideas went into this testing program, but it was sort of accepting that full encapsulation is a viable solution, albeit economically not very attractive and environmentally not great.

00:20:08.931 --> 00:20:12.502
If we can optimize, where is the right place to do it?

00:20:12.502 --> 00:20:15.692
Is it through strategically exposing surfaces?

00:20:15.692 --> 00:20:21.506
Is it through partially protecting certain surfaces, or is it none of the above?

00:20:21.506 --> 00:20:31.770
And so we somewhat systematically went through a process of exposing different surfaces in different ways and seeing what we had by way of an outcome at the end?

00:20:31.770 --> 00:20:34.366
Did it start flaming without our intervention?

00:20:34.366 --> 00:20:35.269
Did we have?

00:20:35.309 --> 00:20:35.851
to intervene.

00:20:36.420 --> 00:20:41.862
One thing that was new, I think, compared to the concept of this partial protection you've mentioned.

00:20:41.862 --> 00:20:50.250
The partial protection would be to some extent the product of again, furnaces and fire resistance testing, because that's how you would get the charring rays and everything.

00:20:50.250 --> 00:20:56.192
Yet in this setting you somewhat lose the spatial effects.

00:20:56.192 --> 00:21:22.525
That's something I've realized very quickly when we started proceeding through this experimental program, that it doesn't only matter what percent of the area I have exposed or what percent of area becomes involved in the fire after a while when some sort of protection fails, but it's which parts of the walls are or which parts of the compartment become involved.

00:21:22.525 --> 00:21:39.351
You know, there was this immediate difference between the ceilings, between the back wall, between two walls projecting at each other, and those are the effects that you cannot express in minutes of resistance right or in charring rates, and from my perspective they made a hell of a difference in here.

00:21:39.880 --> 00:21:41.319
Yeah, you're absolutely right.

00:21:41.319 --> 00:22:12.920
I mean, a fire resistance test is a test on a single planar element, whether that's a wall or a ceiling, but buildings are this complex collection of surfaces that can interact, and so what you sort of elucidate through compartment testing is that interaction that occurs between surfaces, and one of the main challenges if you were trying to, in particular, as you show that the flaming will stop is yeah, okay, the contents of the room can be consumed, and it generally will be consumed.

00:22:12.920 --> 00:22:38.684
What then drives the continuous burning of that enclosure is the heat transfer that occurs between the exposed wall to another exposed wall or from an exposed wall to an exposed ceiling, and that often, depending upon their proximity, can be sufficient for that burning to continue without any additional fire load, whether it's furniture that might be in the room and that shouldn't come as a surprise.

00:22:38.684 --> 00:22:38.965
That's.

00:22:38.965 --> 00:22:46.151
That's exactly how we design timber cribs, for example, is we rely on that kind of interaction for them to be effective heat sources.

00:22:46.941 --> 00:23:07.491
This is just a much bigger version of that, and so I think using arbitrary percentages of exposed compartments as a rule for design is kind of missing the point, because, fundamentally, you can have a huge amount of exposed surface area and it stop burning if the interactions with other surfaces are limited.

00:23:07.491 --> 00:23:19.365
So we showed that in our ceiling case, in a large compartment you throw in walls, you throw in beams, which are another interesting, somewhat sort of micro sort of compartment within a compartment.

00:23:19.365 --> 00:23:22.769
You have all these other interactions that you need to address.

00:23:22.769 --> 00:23:46.695
So you're right, these experiments showed that actually one of the primary variables we need to control is which surfaces are exposed, where and to what extent, and that is one of the dominant things that influence whether you get the outcome that you want, which is hopefully that it stops burning without intervention, or whether it continues so.

00:23:46.759 --> 00:23:51.451
So I'll briefly walk the the listeners through the experimental setups we've tested.

00:23:51.451 --> 00:23:54.304
So, as danny mentioned, we had a cubicle building.

00:23:54.304 --> 00:24:01.183
The front wall of the building was always built from bricks and it had a door-shaped opening in it.

00:24:01.183 --> 00:24:14.690
The size of the opening and the propane burners that we placed inside were so that we bring the fire to like, let's say, fully developed phase in there, and we've played a bit with the internal surfaces of the compartment.

00:24:14.690 --> 00:24:22.820
So we obviously had the ceiling, and the ceiling could have been protected by one, two, three layers of the board, or perhaps it has.

00:24:22.820 --> 00:24:24.104
It had been exposed.

00:24:24.104 --> 00:24:34.623
We had a back wall, so the wall that's basically in the back facing the, the opening, and we kept that also at different levels of protection on it.

00:24:34.623 --> 00:24:36.105
And we also had the sidewalls.

00:24:36.105 --> 00:24:48.039
So we would either expose or protect the sidewalls, which overall led to like I think it was 12 different combinations of walls and different protection strategies.

00:24:48.520 --> 00:24:53.613
Yeah, I think it's 11 in total because of some repeats we did.

00:24:53.613 --> 00:25:19.230
But yeah, you're right, and it is very much just trying to get a handle on, progressively, how we can expose more structure and then to sort of understand the tipping point where we have what was often a compartment that didn't self-extinguish and kept burning and your team had to intervene to the few that actually where we had an outcome where the flaming stopped, smouldering often still persisted and required some dealing with.

00:25:19.230 --> 00:25:26.528
In terms of going through that process, we started with a fully encapsulated case, so every single surface having three layers of plasterboard.

00:25:26.528 --> 00:25:30.059
Fully encapsulated case, so every single surface having three layers of plasterboard.

00:25:30.059 --> 00:25:37.000
And then, from our past experiences of the commercial testing, we felt pretty confident we could expose the ceiling and that that would probably lead to an outcome where the flaming stopped.

00:25:37.000 --> 00:25:46.272
So the kind of the first variable was we went from three layers of plasterboard on the ceiling to a reduced number of layers, to one layer, to two layer and ultimately exposed.

00:25:47.021 --> 00:25:59.361
And what I would say one of the conclusions we consistently found is, if you only expose the ceiling, that was a pretty optimal design in terms of of achieving an outcome where, where you could be pretty sure the flaming would stop and that was.

00:25:59.361 --> 00:26:00.766
That was almost.

00:26:00.766 --> 00:26:13.335
I don't want to generalize, but for the two fairly extreme cases of opening factor that we looked at, bearing in mind one was a very small opening and one was a very large opening that kind of held true across that range.

00:26:13.335 --> 00:26:21.314
And it held true largely independently of the adhesive that was used to form the CLT, which I think we'll talk about later.

00:26:21.314 --> 00:26:23.406
So that was a useful finding.

00:26:23.406 --> 00:26:28.181
It gave us the start of an envelope that we could work with useful finding.

00:26:28.201 --> 00:26:29.784
it gave us the start of an envelope that we could work with.

00:26:29.784 --> 00:26:34.253
I can confirm that with a very unscientific measure, and that is how stressed the laboratory personnel is carrying the tests.

00:26:34.253 --> 00:26:53.230
You know, when we were doing your experiments, where the ceiling was the thing that we played with, we were pretty confident and the stress levels were low, like we kind of knew what to expect, especially after we've done the exposed ceiling experiment, which we assume, okay, that's the worst it can get right.

00:26:53.230 --> 00:26:55.082
We knew what's going to happen.

00:26:55.082 --> 00:27:08.066
More or less we have expected how the self-extinction process will look like and we've expected that it's going to self-extinguish, at least the flaming combustion and the stress levels were generally low.

00:27:08.220 --> 00:27:26.404
But when we were approaching, you know, experiments where we were starting to expose the walls, and especially experiments in which we had exposed walls and just one layer on the ceiling, those were very stressful experiments, man, because like we knew it's going to be much worse but no one could tell how much worse.

00:27:26.404 --> 00:27:28.479
It's going to be much worse but no one could tell how much worse it's going to be.

00:27:28.479 --> 00:27:33.576
And also from maintaining the site after the experiments.

00:27:33.576 --> 00:27:50.787
You know that was stressful because when we did the sealing we could clean up very efficiently, Like you dosed it with water, you handled the joints between the walls and the ceiling and you're pretty much done.

00:27:50.787 --> 00:27:52.125
Then the cleanup next day.

00:27:52.125 --> 00:28:03.273
But when we were doing those other experiments we had that persistent smoldering and I had to put a lot more people into the building for a very long time to observe.

00:28:03.273 --> 00:28:15.026
So perhaps we should measure the laboratory stress when doing those experiments and you yeah, exactly exactly, you could actually measure that.

00:28:15.046 --> 00:28:23.742
I mean that's, I think that would be a quite robust measure, to be honest, depends how experienced the lab staff are in your case, very experienced, but in my case I would just constantly be stressed, so it would be a very, very poor proxy.

00:28:23.742 --> 00:28:25.685
No, it's it.

00:28:25.685 --> 00:28:40.353
There were a range of outcomes and and actually, as we work through that process of exposing the ceiling and showing actually that more often than not the flaming stopped there, when you start to think about why that is, I think it's for a combination of reasons.

00:28:40.353 --> 00:28:59.432
One is the very obvious benefit of gravity in the whether it's partially protected, so a single layer of plasterboard, whether it's due to delamination, what is happening is the thing, the thing that is attached to the timber is generally falling off and it's falling away from the surface.

00:28:59.432 --> 00:29:06.173
So you can be more confident that your lining is going to detach if it does get breached.

00:29:06.173 --> 00:29:21.801
You can be more confident, if you have delamination, that that smoldering char layer is falling and landing, in this case, 2.4 meters away, and therefore that that feedback you get from the stuff that's falling off back to that ceiling is is inherently limited.

00:29:21.801 --> 00:29:34.799
Where things got quite interesting was when we started to, let's say, partially protect the walls and and by that I mean that we're accepting that through the course of the fire the wall is going to become involved.

00:29:34.799 --> 00:29:35.240
We've got.

00:29:35.240 --> 00:29:37.527
We have not protected it enough to prevent that.

00:29:37.527 --> 00:29:41.926
So rather than three layers we went to two layers and into one layer.

00:29:42.268 --> 00:29:58.327
And that presents a real challenge for a timber structure, because what you will often get with a plasterboard lining on a wall or a gypsum fiberboard is a huge amount of heat transfer through it, enough to start pyrolyzing the substrate with the lining still intact.

00:29:58.327 --> 00:30:06.038
You'll then develop some cracks in that lining which allows a certain amount of airflow to develop behind the lining.

00:30:06.038 --> 00:30:16.305
You will then often get some flaming in and around those cracks because you've got the conditions for your flammable mixture at that point and actually you you can develop burning behind the lining.

00:30:16.305 --> 00:30:28.541
That is kind of this perfect trapped little pocket of heat that is sufficient to allow the timber to keep burning, but not enough that the plasterboard ultimately falls off.

00:30:28.541 --> 00:30:37.575
And that in many instances when we had cases where actually the structure just kept burning and burning, it was because the fire had got behind the lining.

00:30:37.976 --> 00:31:01.345
The lining had remained intact and that was enough just to keep the conditions at the surface of the wood high enough that it could keep pyrolyzing at the rate that allowed the flaming to continue, and the only way to address that was to rip the plasterboard off and apply water, which gives you some sort of insight into some of the challenges you might face as a firefighter if you're going into one of these, these buildings.

00:31:01.946 --> 00:31:22.032
You don't see that when, when you've got the, the exposed structure and and interestingly I think it was, think it was sort of differentiated between what we called Experiment 5 and Experiment 6, is actually we had one case where we had an exposed ceiling and an exposed rear wall that stopped flaming without us intervening.

00:31:22.032 --> 00:31:43.140
Yet when we put plasterboard on that rear wall that created this sort of mechanism whereby the heat was trapped behind the lining, it wasn't enough to stop it from burning, and that one just kept burning indefinitely, and so actually sometimes you're better off exposing more structure, and just applying a bit of token protection can actually create more issues than you think it's solving.

00:31:43.761 --> 00:31:44.625
I would rephrase that.

00:31:44.625 --> 00:31:55.242
I would say that if you have a working protection, that's better than the exposed one, but once the protection fails and you create those pockets, it's worse than if you had exposed.

00:31:55.242 --> 00:32:05.282
Yeah, so there's like a point in, like, if you're only interested in the first 20 minutes of fire or something, then yeah sure, then even one layer of gypsum plastic board will give you that.

00:32:05.282 --> 00:32:10.027
Or if you want to prevent the ignition of the timber, yeah sure, that's probably going to work.

00:32:10.027 --> 00:32:31.111
But if you're considering a long-term, this burnout effect, the fact that you perhaps delayed, you know, the onset of fire of your CLT by 20, 30 minutes by putting that one layer, but you've prolonged the entire thing indefinitely versus a self-extinction, well, that's perhaps a net minus to your equation.

00:32:31.111 --> 00:32:38.631
That was really intriguing to see that perhaps we need to develop a more robust failure mechanisms for fire protection, you know.

00:32:38.631 --> 00:32:41.048
So it falls all together.

00:32:42.663 --> 00:32:51.619
Yeah, it's just another uncertainty and I think we've been trying to grasp and understand is one of the fire engineering holy grails is to be able to predict when plasterboard fails.

00:32:51.619 --> 00:33:08.862
And we're still not there after however many decades, and so to rely on it in a manner that you think you can predict when it's going to fail is a bit of a fool's errand to my mind, and so it's, in many respects, I feel like it's a very binary thing you need to do with plasterboard.

00:33:08.862 --> 00:33:26.106
You even need to be very confident that your substrate is not going to become involved, or you're almost better off not having it in some respects, if your objective is for the structure to survive burnout, because it's just going to introduce this additional uncertainty that you can't control that is going to affect ultimately how that surface behaves.

00:33:26.106 --> 00:33:46.643
And on that topic, I think this is where and one of our conclusions relates to the fact that many researchers doing similar experiments to ours have observed, particularly particularly during the cooling phase, this sort of secondary flashover or this tertiary flashover, whereby you think your enclosure is entering into its decay phase.

00:33:46.643 --> 00:34:07.395
You're seeing a reduction in compartment temperatures, a reduction in heat release rate, and then, very suddenly, there is an increase in heat release rate and increase in compartment temperatures and the flaming reoccurs and often that reflashes over and a lot of the schools of thought very early on as to what was causing this was heat induced delamination, which is this Lamea failure mechanism.

00:34:07.455 --> 00:34:16.365
That sort of occurs in proximity to the bond line whereby the adhesive is softening, that Lamea is falling off, and it's often falling off before it's fully charred.

00:34:16.365 --> 00:34:24.983
So what you're seeing is kind of visible wood underneath and that is exposing fresh fuel to the compartment and giving you something extra to burn.

00:34:24.983 --> 00:34:27.751
That wasn't there when the charlay was intact.

00:34:27.751 --> 00:34:39.971
But actually when you look at that as a cause for secondary flashover all of my observations I don't know if you agree from what you've seen, but generally that's very small areas in little bits at a time.

00:34:39.971 --> 00:34:47.331
You've made reference to charcoal rain before in the past, but it's relatively small surface areas that you're exposing instantaneously.

00:34:54.739 --> 00:35:03.304
What we saw with lining failures failures particularly on ceilings is a very large surface area exposed very quickly when that lining did fail and that was the cause of a secondary flashover in all of the instances where we observed it.

00:35:03.304 --> 00:35:03.565
And so it's.

00:35:03.565 --> 00:35:05.612
It's another reason to sort of create that distinction in how you want the plasterboard to perform.

00:35:05.612 --> 00:35:10.327
You either really very much need to where we observed it, and so it's another reason to sort of create that distinction in how you want the plasterboard to perform.

00:35:10.327 --> 00:35:23.907
You either really very much need it to stay intact, or it's almost better to have a visible surface, because you're at least able to see and predict with far more certainty how you think that surface is going to perform.

00:35:24.581 --> 00:35:28.786
But the second flash overboard that you alluded to, it indeed happened in one of the experiments.

00:35:28.940 --> 00:35:39.619
So we were cleaning up and fortunately some of the cameras were still recording and the temperatures were still recorded by the data system.

00:35:39.760 --> 00:35:49.782
But we were very happy to finally go home after a long day of experiments and I remember I went to some other building at the lab and you text me there's a second flashover.

00:35:49.782 --> 00:35:50.626
What the hell?

00:35:50.626 --> 00:35:58.952
Just when I left and I came back and it really was almost a fully grown fire again and what happened is that we've lost the ceiling protection.

00:35:58.952 --> 00:36:12.989
It basically fell as one huge chunk, like all of it at once, and later, analyzing the videos, it took like 15, 20 seconds to reestablish the fire once again after we removed the fire source from the building.

00:36:12.989 --> 00:36:44.047
So there was no movable fire load, let's say, in the compartment anymore, it was just the structure going, though it's important to note that this was a scenario where we had two exposed walls as well in that scenario, because in previous experiments, when we had three layers, we did not lose ceiling ever, and when we had less layers we've lost the ceiling and even though it participated in the fire, it went through the self-extinction phase once we turned off the movable fire load right.

00:36:44.047 --> 00:36:45.208
Yeah, that's right.

00:36:45.710 --> 00:37:13.648
So I think I mean a really important conclusion of our work is, if your objective for your building is to have a good chance of surviving burnout, then actually partial protection is not a strategy that's really going to be able to facilitate that, particularly if it's the walls that are partially protected, just because of the huge uncertainty that comes from exactly how that plus or detaches, whether it will fall off or not, and such like.

00:37:13.648 --> 00:37:47.773
So when considering that sort of envelope, we're very much of a view that you should either expose surfaces and be able to demonstrate that they stop flaming, or you should fully encapsulate, and anything in between is very much a grey area that, as designers, we where we won't be able to predict what's going to occur you think from a designer point of view, just assuming that partially protected surfaces will participate in the fire, so you may just as well treat them as as unprotected and from that point consider the conditions whether they will go to burnout or not.

00:37:48.201 --> 00:37:50.746
Is this a feasible proxy?

00:37:51.228 --> 00:37:51.527
it's.

00:37:51.527 --> 00:38:13.469
It's an assumption you can make about a partial protection strategy, but that shouldn't detract from the fact that if you put a certain amount of protection on your walls and ceiling and you expect the structure to become involved, then the outcome will be heavily dominated by the, the unpredictability of that plasterboard and how it detaches.

00:38:13.469 --> 00:38:13.650
So.

00:38:13.650 --> 00:38:22.664
So as a designer, you're going to want some surety that either it's going to stop burning because it's exposed or it's not going to become involved.

00:38:22.664 --> 00:38:34.090
That it will become involved at some point in time is the very tricky thing, I think, to manage, so I think we should be avoiding it at this point, based on what we've seen.

00:38:34.541 --> 00:38:49.132
Now, going back in my head to what you've said previously, that it's better to have it exposed, and at least you know what's happening, than have it partially protected by a failed protection, perhaps, perhaps, and have a fire pocket going behind it.

00:38:49.132 --> 00:38:52.871
I agree, that's perhaps a more stressful situation.

00:38:52.871 --> 00:39:04.130
One more observation that I've made, being there firsthand, and that's again kind of this obvious unscientific thing, but it has, in my opinion, significance.

00:39:04.130 --> 00:39:11.992
You've said that the lamination, you have those small parts of your CLT falling down.

00:39:11.992 --> 00:39:15.250
If you have a ceiling, it basically rains on your floor.

00:39:15.800 --> 00:39:19.965
But when we had exposed walls, you know, we still had lamination.

00:39:19.965 --> 00:39:28.030
And it's not that charcoals were shooting for 50 meters away from the timber wall, no, they were like dropping just next to the wall.

00:39:28.030 --> 00:39:44.990
And eventually, after like 60 minutes of the fire, you had a very nice pile of charcoal build up against the wall, you know, and that pile again created those conditions in which self-extinction was very difficult because it would just keep going from that pile alone.

00:39:44.990 --> 00:39:50.402
And when you think about it, it's obvious, you start producing the lamination.

00:39:50.402 --> 00:39:55.157
They must accumulate somewhere and they fall down next to your wall.

00:39:55.257 --> 00:40:13.550
And here you go, you have a perfect conditions to restart the fire or keep the fire going yeah that was very interesting it was, and there's been a huge amount of thought and emphasis placed on avoiding delamination as being a sort of a prerequisite of a good design.

00:40:14.320 --> 00:40:23.260
But I think the point you make there is it indicates it's a lot more complicated than kind of you solve that problem and other things are secondary.

00:40:23.981 --> 00:40:45.851
I think if I was sort of to prioritize the factors that influence whether something is going to stop burning or not, configuration comes before needing to mitigate delamination because, as we say, for a ceiling case, that char is detaching and falling on the floor and it's 2.4, anything that's 3.6 meters away, depending upon floor ceiling height of your building.

00:40:46.632 --> 00:40:51.114
But actually there are situations where you might want to use more heat resistant adhesives.

00:40:51.114 --> 00:40:54.643
They would be the instances of the walls that you reference.

00:40:54.643 --> 00:41:13.291
If you really want to expose a wall, then I would say preventing or being confident you're going to prevent delamination is kind of a prerequisite of a wall self-extinguishing, at least for flaming because of that mechanism that you highlight, because you cannot control how that char falls off or where it's going to land.

00:41:13.291 --> 00:41:22.291
Either that or you end up with some kind of hybrid whereby if you were going to emphasize protection somewhere, maybe at the bottom of the wall is the place to do it as a way of.

00:41:22.291 --> 00:41:33.994
It's architecturally probably not very attractive, but from a purely fire science point of view, a level of protection around the lower portion of your room is probably a good idea to address that.

00:41:34.820 --> 00:41:41.913
That's also that came to my mind like what if we just, you know, isolated this pile of charcoal from the wall?

00:41:41.913 --> 00:41:42.855
Would that be sufficient?

00:41:42.855 --> 00:41:50.353
It's not something we've tested, but I have some CLT in the lab left over and I'll maybe buy some charcoal and we can try it out.

00:41:50.353 --> 00:41:56.108
The critical conditions to prevent that Actually, that's a solid research question.

00:41:56.108 --> 00:42:00.271
Another thing the conditions in which the self-extinction happened.

00:42:00.271 --> 00:42:05.952
How consistent was it with your idea of how it will look like before the experiment?

00:42:05.952 --> 00:42:10.063
Did it surprise you or did it go just as you believed it will?

00:42:10.563 --> 00:42:27.246
We always felt confident with the ceilings, just when you sort of work it through that once in our instance the propane had reached a level where the heat flux to the ceiling was low enough, you were always going to reach a point whereby you just couldn't sustain the pyrolysis at a rate for the flaming to continue.

00:42:27.246 --> 00:42:32.719
Where I got less confident was when we started exposing surfaces very close together.

00:42:32.719 --> 00:42:42.289
So the more surprising one was the back wall and the ceiling which sort of logically you would think in the corner at least they're very close and therefore that would present an obstacle.

00:42:42.289 --> 00:42:44.286
That one did start flaming.

00:42:44.286 --> 00:42:58.650
But then some of the more surprising ones, which was two opposing walls didn't, in one instance partially, because the ceiling ended up collapsing and the dry lining fell off and then we ended up with way too much surface area exposed.

00:42:59.440 --> 00:43:42.523
But one of the tasks and I should remiss of me not to have done this to this point, but credit my co-authors which was Carmen Gorsker and Michael Spearpointacob from from itb and yourself and our funders one of the analyses carmen did was to actually look at what the instant heat flux to the surfaces was in and around the point at which flaming stopped, and so there's a nice graph in the paper that shows, actually, whenever, during the decay phase, the heat flux dropped below something like 43 to 47 kilowatts per square meter, we we had this outcome where the flaming stopped, and that was quite reassuring because it lines up with a few of the sort of smaller scale observations.

00:43:42.583 --> 00:43:46.335
So if you look at rick emberley's work in particular, it's kind of in and around the thresholds he observed.

00:43:46.335 --> 00:43:48.300
It's of in and around the thresholds he observed.

00:43:48.300 --> 00:43:56.550
It's not too far from the thresholds alistair bartlett observed, and these kind of experiments were done at the scale of the cone, not the scale of the compartment.

00:43:56.550 --> 00:44:12.891
So there was whether it was a coincidence or not, that there was at least we are in the right ballpark from those small scale experiments and it was replicated in our in our large scale experiments one thing that gives me confidence in this assessment is an exercise I've done.

00:44:13.311 --> 00:44:18.650
There was this one test in which we were almost on the brink of self-extinction.

00:44:19.190 --> 00:44:23.728
Perhaps you remember it was like going, oh, it's gonna go down, no, it started burning again.

00:44:24.148 --> 00:44:41.132
We've had that wall going up and down, up and down, the temperature is going 20 up, 20 down, and it was just going and going, and going and going and eventually we've decided that now we don't consider this as self-extinguished because it's not going down in the reasonable timeframe.

00:44:41.132 --> 00:44:45.646
And then you've allowed me to spray some water inside the compartment.

00:44:45.646 --> 00:44:48.032
You know, and I remember we sprayed a tiny little amount some water inside the compartment.

00:44:48.032 --> 00:44:58.929
You know, and I remember we sprayed a tiny little amount of water inside, just you know, just to get it this 20 degrees lower, just to get rid of those few kilowatts, and we left it alone and it eventually went down.

00:44:58.929 --> 00:45:07.847
It felt like if we took it outside of this chaotic balance of going up down, we moved the boulder from the hill and it just went down.

00:45:07.847 --> 00:45:21.280
It's not something you can scientifically prove or that we can show measurements, but that was my observation, being there standing with the water nozzle and being able to actually push it towards the self-extinguishing phase.

00:45:21.420 --> 00:45:22.905
Yeah no, it's, it was.

00:45:22.905 --> 00:45:28.306
It was reassuring because as a designer you've got to sort of know your colors to the mast in some respect.

00:45:28.306 --> 00:45:35.851
In some kind of performance criteria, some kind of critical heat flux for autoextinction is things people are adopting.

00:45:35.851 --> 00:45:39.690
Purists will say mass loss rate is what you should use.

00:45:39.690 --> 00:45:48.875
But it was gratifying, reassuring to see it translate across two very different scales and be in and around a similar sort of number.

00:45:48.875 --> 00:45:50.605
It gives us something to work with.

00:45:51.481 --> 00:46:18.072
One more thing that was a part of these experiments was also measurements on the external wall of the compartment, because an obvious assumption is that when you have additional fuel, that is the timber burning inside of the compartment, and we were at a ventilation limit pretty much, so there was not that much more combustion you could do inside, just not sufficient air, and you could obviously see and feel that the fire was larger outside.

00:46:18.072 --> 00:46:22.150
How did that correlate with the number of exposed surfaces?

00:46:22.150 --> 00:46:30.181
And was this again something that went like you've expected it to go, or there were things that surprised you?

00:46:30.605 --> 00:46:32.755
it was one of the more predictable elements actually.

00:46:32.755 --> 00:46:38.539
So when you're heating the compartment, the compartment's getting hotter, the surfaces are receiving more heat flux.

00:46:38.539 --> 00:46:48.155
You uh, you are pyrolyzing more fuel and it will to some extent be proportional to the exposed surface area, because that's that's additional fuel that can pyrolyze.

00:46:48.155 --> 00:47:07.806
And you're right, we very consciously put the ventilation limit right at the limit of the heat release rate of the propane, which is, I think, 1,800 kilowatts, and so that meant the moment the timber structure became involved, that would lead to external flaming, because there is just an excess relative to the opening factor.

00:47:07.806 --> 00:47:19.226
And it was very much the case that the more we exposed, the more external flaming we had, and consequently, generally speaking, the taller those flames became.

00:47:19.226 --> 00:47:41.360
And so when you actually look at that from the perspective of the heat flux, back to the elevation of the building, it increased by more than a factor of two relative when you consider the baseline where the structure is not really becoming involved, through to one of the more extreme cases where we had something like 45 to 65 percent exposed.

00:47:41.500 --> 00:47:51.153
So it's something lots of people said, it's in the original we need to talk about timber paper by Rory Haddon and Angus Law that you will get increased external flaming.

00:47:51.153 --> 00:48:08.894
That will present challenges for, in particular, external fire spread across the boundary, because that's additional energy that your neighbors receive, but also in in terms of vertical fire spread and what that means for the evacuation of your building, your structure potentially being heated over multiple stories simultaneously.

00:48:08.894 --> 00:48:20.971
And so, again, reassuring, because sort of our ofr position to this point is being we do need to do something specific with the facade for mass timber buildings.

00:48:20.971 --> 00:48:37.360
The good old spandrel that was once a thing should make a comeback, purely in terms of managing that heat flux increase that you get from the external flaming, managing that heat flux increase that you get from the external flaming.

00:48:37.380 --> 00:48:45.884
So, yes, one of the more predictable elements and, and again, reassuring from a design perspective when you say it was like 50 percent or some or whatever um higher, it doesn't give you the finger.

00:48:46.385 --> 00:49:06.351
For me, the the important was that when you had the exposed surfaces in there, the the larger parts, like we had two walls or a wall on the ceiling exposed you would find that at 1.6 meter mark, higher heat fluxes than that you would have at like half a meter when there was no timber burning in it.

00:49:06.431 --> 00:49:23.922
Right, yeah, so the separation distance at which you're safe again vertically is no longer like a meter, meter and a half, it would be perhaps two meters, three meters, I I don't even know how high would you go which essentially is the height where you have your next compartment.

00:49:23.922 --> 00:49:29.119
So this external fire spread between the levels is so much more likely.

00:49:29.119 --> 00:49:44.081
And assuming you had exposed surfaces of your timber in the, the floor that you you underwent fire in kind of kind of, you know, alludes that you're going to have more exposed surfaces on the upper floor, so perhaps some more targets for radiation.

00:49:44.081 --> 00:49:55.184
So I I would say that this, to me at least, a subject for consideration for practical design and perhaps a real challenge with those timber buildings yeah, it's, it's a good point.

00:49:55.244 --> 00:50:22.641
I think what, architecturally, many people have been trying to do with timber buildings is to make timber buildings look like non-combustible buildings from the outside, and so they expect the elevations to be comparable, to have the same amount of glazing, to have the same sort of facade systems, have the same types of solutions and and I think that's a bit flawed and we should be treating mass timber buildings as as unique things in all facets of the fire strategy.

00:50:22.681 --> 00:50:27.057
And that means, potentially, that the external wall needs to look different.

00:50:27.057 --> 00:50:36.485
Perhaps it's the case that having consecutive openings on stories is something we want to avoid for the purposes of managing that increase in heat flux of two, three meters.

00:50:36.485 --> 00:50:47.483
That's something that we can implement if we think about it early enough, and it's just a case of articulating the facade in a different way to how we would do if it was a steel or concrete building.

00:50:47.483 --> 00:50:53.427
So I think if you look at what the hazards are and you look at the data, there are solutions to the problems.

00:50:53.427 --> 00:51:04.603
It just requires a certain amount of compromise from all the other members of the design team to sort of recalibrate their expectations for what their building is going to look like, and that's the missing bit for me.

00:51:04.603 --> 00:51:08.237
At the moment, we're still kind of just trying to make them look like every other building.

00:51:08.719 --> 00:51:20.469
So, after all this we went through, did it change the way of our design timber, or does it just give you more confidence in the strategies you are employing?

00:51:21.114 --> 00:51:22.661
In many respects it was validating.

00:51:22.661 --> 00:51:27.748
We've been incredibly cautious when it comes to mass timber buildings.

00:51:27.748 --> 00:51:33.235
The majority of the demand has been in commercial buildings.

00:51:33.235 --> 00:51:34.697
Majority of the demand has been in commercial buildings.

00:51:34.697 --> 00:51:52.028
So actually a lot of the work that we've done around the offices high opening factors, large compartments, generally exposed CLT slabs the workers informed our positions on that and actually it's been able to it's allowed us to make more informed choices around when we choose different adhesive types.

00:51:52.028 --> 00:51:56.637
So we know we don't need to be so concerned with it.

00:51:56.637 --> 00:52:07.148
If it's an exposed ceiling, if we have an interaction between a ceiling and a wall, then that's definitely something that we do want to to an uncertainty we want to remove.

00:52:07.148 --> 00:52:13.045
Or perhaps it's it's a ceiling with an exposed glue and structural frame, for instance.

00:52:13.045 --> 00:52:14.128
So we do.

00:52:14.128 --> 00:52:21.605
It's definitely informed our decision making around things like adhesive choice and actually what conservatisms we need to adopt where and when.

00:52:22.025 --> 00:52:28.085
From a residential point of view, it's not really changed anything, and that's not because we've we felt like we knew all the answers.

00:52:28.146 --> 00:52:42.590
It's more the case that the confidence to build and again just justifiably so in my view medium to high-rise mass timber residential buildings is still not there and therefore it's not really happening and so there isn't at this point a need to apply it.

00:52:47.054 --> 00:52:56.500
Design guides have come out to try and enable more sort of low to mid-rise timber buildings, as the new model building which was sort of done by Warthisselton and UCL and Bureau Haphold, which is a fully encapsulated structure.

00:52:56.500 --> 00:53:15.615
I still think, even if we accepted the commercial downsides and the sort of carbon and sustainability consequences of full encapsulation, there are still some annoying little details that remain unresolved for me, in particular how we deal with penetrations.

00:53:15.615 --> 00:53:23.981
So if you think about how we test a pipe or a duct passing through a wall, the test focuses on what's happening on the unexposed face.

00:53:23.981 --> 00:53:43.762
It doesn't take a look at what's happening at the exposed face, whereas we know from an encapsulation point of view, the critical surface is the interface between the wood and the protection on the exposed side, and so it almost needs to rethink in terms of how you test or detail penetrations if you're going to fully encapsulate something.

00:53:43.762 --> 00:53:49.563
So there's some annoying realities of building buildings that haven't been addressed to this point.

00:53:50.356 --> 00:54:19.670
I would even re-emphasize what you've just said, because you have an impact on the design phase and execution phase, but you don't have the impact over 30 years or 50 years of a person living in the home and they may want to put another electrical socket in the compartment, they may want to hang a painting with a nail, you know, which inevitably leads to some sort of damage to the encapsulation, and I'm not that sure if a person would be aware of the consequences.

00:54:19.670 --> 00:54:27.557
Not even to mention a case where someone eventually discovers there is a beautiful timber underneath their plasterboard and decides to expose that.

00:54:27.557 --> 00:54:29.626
But that that's another story so there's.

00:54:29.646 --> 00:54:43.925
There's definitely an onus on designers passing on that information and being able to communicate the importance of that lining, but there there is an inevitability that you highlight, that people want to use their homes as a functional place.

00:54:44.005 --> 00:54:59.900
That means they do want to attach their tv to the wall, they do want to hang things up, they do want to put shelves up and things like this, and all these have the potential to undermine the performance of your lining, and so a good designer, I suppose, is going to have to foresee that and counteract it.

00:54:59.900 --> 00:55:03.791
And and the way in my head that you counteract that is you almost.

00:55:03.791 --> 00:55:44.784
You have your life safety critical lining system and then you have a somewhat sacrificial aesthetic lining that sits somewhat separated from that and all that's doing is introducing yet more plasterboard or geofiberboard into the discussion, because we've now got this aesthetically sacrificial element to it as well definitely an interesting world and a no one silver bullet, but a lot, lot of engineering to be done and I'm really really happy that, thanks to this project, it's science-based, evidence-based engineering and not some of our predictions.

00:55:45.454 --> 00:56:04.021
Let me thanks once more to the sponsors, structural Timber Association, sta this time I got it right and the store, enzo KLH and Bindeholz, with whom we were working for the CLT for the project and who've been working with us on this, and huge thanks to your team at Ofer.

00:56:04.021 --> 00:56:06.710
It was a huge pleasure to work on this with you.

00:56:07.492 --> 00:56:09.197
Yeah, I've got a feeling we're not done either.

00:56:09.197 --> 00:56:11.802
I think we've created more questions than we've answered.

00:56:11.842 --> 00:56:19.338
that's for sure Not only this always happens when we do research, like every time we answer a question there's like three follow-ups.

00:56:19.338 --> 00:56:29.858
I guess our jobs feel pretty secure and thanks for coming to the Far Science Show, mate, and I hope to see you soon, sooner than in 150 episodes from now.

00:56:30.240 --> 00:56:30.682
That's the plan.

00:56:31.054 --> 00:56:31.637
That's the plan.

00:56:31.637 --> 00:56:34.036
Thanks for having me, and that's it.

00:56:34.036 --> 00:56:51.954
Just as I promised, the episode was full of insights into mass timber, with observations on the experiments and good advice on how to design them, perhaps not everything lining with what we know of or what we do, especially the partial protection strategy.

00:56:51.954 --> 00:57:06.150
I mean it felt really odd when we seen those bolts falling off and the fires changing its course, and I kind of resonate when danny says that he would prefer exposed than partially protected ceiling, for example.

00:57:06.150 --> 00:57:12.123
I mean seeing all those fires firsthand and I literally seen them from like five meters.

00:57:12.123 --> 00:57:18.565
I've stuck my iPhone into the door opening, felt the heat on my back Seeing all of that.

00:57:18.565 --> 00:57:30.175
I kind of agree I would prefer exposed surface that I'm pretty sure that's going to self-extinguish over an uncertain protection that could make stuff much worse.

00:57:30.175 --> 00:57:32.860
Of course depends on what structure I would be designing.

00:57:32.860 --> 00:57:43.648
If it's like two-story building or three-story residential building with good access from all sides, ability to fight fire from the exterior, perhaps not such a big deal.

00:57:43.648 --> 00:57:52.356
But if that was like 20-story tall building, boy, that's a completely different problem and that needs a lot of thinking, a lot of thinking.

00:57:52.356 --> 00:57:58.028
I also like that this is positioned for residential buildings, for perhaps smaller buildings.

00:57:58.028 --> 00:58:07.684
I think these are the goals to decarbonize the construction economy, and I think this is what we should focus on as researchers.

00:58:07.684 --> 00:58:30.458
Google and others are rich enough to fund their own research programs if they want to build their new magnificent headquarters, but I think, as scientists, we should provide answers to the society that are important and valuable to the society, and those answers you'll find in the run of the mill buildings.

00:58:30.458 --> 00:58:35.070
So I'm very happy that we've contributed to this part in run-of-the-mill buildings.

00:58:35.070 --> 00:58:36.492
So I'm very happy that we've contributed to this part.

00:58:36.492 --> 00:58:37.094
Anyway, that's it.

00:58:37.655 --> 00:58:39.231
I won't summarize everything that has been said because that's just too much.

00:58:39.231 --> 00:58:44.184
There are papers that you can read and I will link the papers in the show notes.

00:58:44.184 --> 00:58:47.155
There is Work Package 5, paper and Fire Technology.

00:58:47.155 --> 00:58:50.440
That covers everything that has been discussed in this podcast episode.

00:58:50.440 --> 00:58:54.106
There is Work Package 6 in Journal MDPI fire.

00:58:54.106 --> 00:59:00.065
That covers the previous experiment where we weighted the slab when we were burning it fun stuff.

00:59:00.065 --> 00:59:05.402
And there's a paper from some global timber conference on the impact of the beam hat.

00:59:05.402 --> 00:59:11.981
Danny mentioned that beam changed a lot when we introduced a beam into the flat ceiling that we've previously tested.

00:59:11.981 --> 00:59:19.746
And, yeah, there's a paper about that that you can look and see for yourself like how much beam changes in the outcomes of the fire.

00:59:20.175 --> 00:59:22.543
So that is it for today's episode.

00:59:22.543 --> 00:59:27.708
I'm just heading to Barcelona for the European Fire Safety Symposium.

00:59:27.708 --> 00:59:40.362
If you are there, let's meet and talk Really happy to meet all of you, and if you're not, I hope you're having a great time and, of course, next week, more fire science coming your way in the fire science show.

00:59:40.362 --> 00:59:41.085
Thank you, bye.

00:59:41.085 --> 00:59:54.882
This was the fire science show.

00:59:54.882 --> 00:59:57.264
Thank you for listening and see you soon.

172 - Lessons from mass timber experiments with Danny Hopkin

Listen On

Recent Episodes