130 - Mass timber fire dynamics with Dr. Carmen Górska

Hello everybody, welcome to the Fire Science Show. In this week, fire Science Show is back to the subject of mass timber in fire and it's not just an ordinary interview. In this week I'm interviewing Dr Kamen Guszka from OFR Consultants. Kamen has just recently won the Philip Thomas Award for the best paper of the previous IFSS symposium and, to give the full credit, the paper is also called for by Juan Hidalgo and Jose Tereiro. Anyway, this is a huge award. Actually comes with the physical medal. I was created in ancient Greece in times of Socrates. Amazing, I've seen it with my own eyes. It's really cool. But it's not the medal that we want to celebrate. It's the research and science and knowledge that she has built up and explained in her research paper. In this episode we will talk primarily about fire dynamics in timber compartments, because that's the subject of her PhD, of her work, and by fire dynamics we mean how the physics of fire itself is changed by the presence of exposed timber in a compartment. So many engineers would focus on stuff like charring grates or critical heat flux conditions that help you get self extinction of some sort. Okay, that is important. What was found in this work, and I expect that's one of the reasons why it is our winning work is that she has clearly identified different mechanisms within the compartment fire, and not just identify. She put them against well established opening factor methodology that exists for years and is a cornerstone of many of the fire safety considerations, propose a sound explanation of what she saw and give some really good practical advice to engineers that have to work with timber, mass timber and fire safety. So I hope I don't need to advertise this anymore. This is a very insightful and inspiring talk by a young scientist, by a woman in fire, who has just won one of the biggest awards in the entire fire discipline. So once again, please help me, welcoming Dr Carmen Gurska from all of our consultants and let's go. Welcome to the fire size show. My name is Vojty Wynchinski and I will be your host. This podcast is brought to you in collaboration with offer consultants and multi award winning independent consultancy dedicated to addressing fire safety challenges. Ofr is UK's leading fire risk consultancy. Its globally established team has developed the reputation for preeminent fire engineering expertise, with colleagues working across the world to help protect people, property and environment. Internationally, its work ranges from the Antarctic to the Atacama Desert in Chile to a number of projects across Africa. Ofr is calling all graduates, as it is opening the graduate application scheme for another year, inviting prospective colleagues to join their team from September 2024. By taking this opportunity, you'll be provided with fantastic practical immersion in the fire engineering and unique opportunity to work with the leading technical experts in the field, while learning the skills critical to become a trusted consultant clients. This opportunity is tailored just for you and if you would like to take it, please visit offer consultantscom for further details and instructions on how to apply. Hello everybody, I'm here today with Dr Carmen Gurska. Hello, carmen.

Hello, thank you for inviting me to your podcast.

Very happy to have you here. I've invited you as the new recipient of Philip Thomas Award for the best paper at the previous IFSS symposium. Fire Dynamics is in mass timber compartments, a paper that I know very well and it it made me very happy to see this work being awarded as the best paper of the previous symposium. So today we're going to talk about timber and your experiences in experimental investigation or fire dynamics in timber compartments. Perhaps we can start with why this edge on timber. You know everyone is looking into. You know charring, lamination, stuff like that. You started battling with fire regimes and other very difficult aspects of fire dynamics. What brought you into this?

Well, to be honest, when I just started with the topic, I had no idea where to start just at the beginning of my PhD and basically Jose Dorero said okay, let's just build like a little compartment and let's see what happens, you know, so I literally started by actually manufacturing CLT cross laminated timber myself in a very dodgy way probably, but I remember that very first test already was much bigger than we expected in our lab. So we were all like surprised in that very first step of how big these external flanks, for example, were. And we were aware of many other problems, like for example, that there is this big unknown of connections, how still connections are going to perform with the timber. But with that very first test it was clear to us that we don't know the very, very basics of fire dynamic with the presence of exposed timber inside of a compartment.

Okay, fantastic, that's a great word to start. So in a way, it was exploratory to just figure out what's happening around and how you've chosen your compartment size dimensions. What did drive you into the final experiment design that you've used in your PhD? Was it the iterator process?

Well, you do experiments yourself and I guess that you know that sometimes our choices are not based on physical arguments of physics, sometimes they are more like practical. So there were like few constraints. First was the amount of flow we could take with our hoods, like the fire couldn't be bigger than a certain hitter is right. So that was quite restrictive to do the test in the lab. And another big thing was that ideally I was supposed to be able to do this test myself, because otherwise it would be very hard to do, because we wanted to study like this cessation of putting more and more and more timber and ideally doing several repeats to make a proper study. And ideally I would be. I should have been able to do this myself. So the COD walls I was supposed to be able to handle them myself. They were already quite heavy with these dimensions, but yeah, they were like around 25 or 30 kilos but still possible to handle for one person. But bigger than that I would have needed help. And that probably instead of four years it will have taken me six to this PhD with more than six higher compartments. And then the dimension of the opening they were. We just wanted to make sure that we have post-clash-over conditions, so, yeah, that's how we chose the opening.

I exactly know what you mean about the capacities of the laboratory and the manpower that drives the experimental design. Very similar how in CFD or numerical studies, people usually based on how many processes they have, not the physical constraints of models that they're using. Anyway, let's go into the experiments, because you have done a lot of medium scale fire tests with the CLTs. So let's introduce the audience to the concept of the test. How did the experimental setup look like and why did you call it the medium scale test?

Well, because we were saying that small scale tests of timber are all these tests you know, in cone, calorimeter and FBA and such. Then there was the material scale, which was TGA scale, and it was not a large scale like you know, like where a person can walk in. So eventually, yeah, it ended up being a medium scale, basically, where we can already capture compartment dynamics, but it is obviously not a full scale. So we were aware that maybe there are some phenomena that are restricted by the scale.

But it was not like a fruit number scaling that you do 100kW fire and you say it's 10MW fire. It was simply just a reduced geometrical size and you allowed physics to go on.

Well, we did an attempt of doing a scaling analysis but, yes, as you say, there are so many variables that it is very difficult to keep all the parameters and scale them all down. So we chose a few, which was basically the opening factor, and achieve these temperatures that you would usually achieve. But, yeah, the opening factor was the main thing that you wanted to keep.

Yeah, and when you were talking about this project I remember you were talking about Poland opening factor was like the first thing you introduced during the talk and it ties very nicely to the historical research in fire science. So perhaps for those who are not familiar with the concept of opening factor and how we relate the fire dynamics to it, maybe we can try explaining to the audience the concept of opening.

Yeah, it's opening factor is a pretty simple parameter that basically relates the amount of heat generation to the heat losses from a compartment. So basically, the more heat generation and the less heat losses you have through the boundaries, the higher would be your temperatures. But eventually the fire safety science also discovered that it not only can predict the temperatures but there's also a change of regime for some opening factors. So, as you know, as we go to compartments with huge like, with very big openings, the fire is, in simple words, behaving more like not a fire, not constrained by its boundaries, as in a postplasio fire, and it basically is driven by different mechanisms and that affects the temperature distribution, the flows, etc. And we, like the first check, was like how can we apply this framework for timber compartments and how this opening factor is going to apply to timber compartments?

So we do like we usually would refer to those regimes as fuel control, ventilation control, and there's like a very narrow space between the two of them where the fires have the biggest severity and at the ends of them, at very low opening factors or very large opening factors, the fires tend to be smaller. Let's dig deeper into that because I feel that it's important to the latter discussion. If you have too much fuel, why do temperatures go down in a compartment? Why it's not, as let's say, severe if we measure severity by temperatures?

I mean, if you have fuel controlled and your openings are bigger, then basically you have a lot of heat losses through your openings and it's basically cooling down the compartment right, okay, in average, because these are average temperatures always when we talk about that regime. And as you start to close your openings you transition to this other regime. So basically your temperatures start to build up, build up, build up and you go to this maximum point on the plot and eventually you have this change of regime to ventilation control and this kind of like an optimum geometry for the compartment to have this optimum relation of heat losses to heat generation. So you have, like the good amount of air entering to have a very it's not stoichiometric mixture but the closest as you can get to stoichiometry inside.

Very efficient.

Yeah, it's efficient combustion and minimizing your heat losses through the opening and through the boundaries. But then if you keep closing, keep making smaller the opening, then you are basically the opening starts acting as a valve of oxygen supply and you keep closing, I think closing so there you have less and less oxygen flowing in and therefore the mixture inside is oxygen deprived and your combustion efficiency decreases and your temperatures go down. I can't really have close and fully. You kill the fire. That would be like the extreme.

Yeah, that makes sense. How the ventilation changes, the well, of course it makes sense. That's the fundamental for fire science that we know. I wonder, is there like a science restriction on this relation? Like if you have extremely large open time compartment, it's perhaps going to be different because of the spatial distribution of temperature fields like not uniform anymore.

Yes, how the Thomas and Henselden plot looks like, which has this optimum right.

Yeah.

It is important to always remember that for the regime that is fuel controlled, so the one that you mentioned, with large open floors and big openings, there's a huge scatter of data there and it is true that fires in that regime are not as easy to predict as to say, okay, the temperature is going to be whatever. It seems this correlation, because now we see that the fires in this, in these geometries, they can have like localized effect and be very severe in one side of the compartment or they can be traveling fires, so the fires they get much more complex to solve. And usually that's when we need FDS modeling, because simple assumptions of saying the whole compartment has this temperature and is uniform across the compartment, as we do with the other regime, is not valid here. And the big question with Timber was because for conventional building we more or less know when we are going to transition from one regime to the other, depending on the geometry and opening, right. But now that we have this huge amount of additional fuel because our surfaces are exposed Timber, we didn't know when this transition is going to happen, because maybe the same compartment that is not combustible has a traveling fire, right, but maybe if the ceiling is now exposed Timber and it's just ignites almost instantaneously. Maybe that actually can induce flash over of the compartment. So now basically, we can have the exact same geometry, but just because we have additional exposed surfaces of Timber, the same geometry can maybe afford for a post-flash over fire, and this is not really well defined yet.

I think now we got into your experiments, because this is in a way what you've been researching. You have built this, let's say box compartments, with an opening factor that kind of hits the peak at the Thomas-Hasselden curve, which is the, let's say, optimal combustion rate.

Well, it was not at the peak, it was a bit more to the right to ensure that we are in the ventilation control. Okay, okay so give a bit of margin, because there is always big variability of fire.

But more or less in the top region, not on the ends of the scale. And then you start playing with the CLT. So what were you doing exactly with this research? What were your variations of the experiments? What outcomes did it lead to?

Okay, so in my experiments, yes, I was doing these little boxes, compartment fires, and I was always keeping the geometry the same and the fuel load the same across all the experiments. Well, I did one experimental campaign with woodcrips and one with kerosene, but that's another topic. The geometry was always the same and the only thing that I was changing was the amount of exposed timber surfaces inside the compartment. So I was protecting, in general, the timber with plasterboard and this is how I did, for example, my baseline tests, where tests with all the timber protected. And then I gradually started to expose more and more and more and more timber. And I was having around 100 sensors in these medium scale compartments and I was measuring a lot of different parameters heat fluxes to the surfaces inside the compartment, temperature distributions inside the compartment, temperatures at the opening, workflows at the openings, charring rates. I had a mockup facade above the opening and I was measuring the heat fluxes to the facade. Anyways, what I observed is that by increasing for the same geometry and by increasing the amount of exposed timber, I would first notice that the temperatures inside the compartment also increased because I'm adding fuel right. But then I was increasing even more the amount of exposed surface and then the temperature started to decrease. So these lower temperatures? Now they were because there was so much excess of pyrolysis gases coming out from the timber that the gas mixture inside the compartment was this time it was just too fuel rich and therefore the combustion was not efficient and the temperatures. So in conventional buildings we would usually encounter situations for large open floors with a mixture that is too oxygen-rich and the oxygen would cool down.

And here was like the opposite right so, if I understand correctly, you had a fixed geometry, fixed opening. So the ratio of opening to the surfaces inside, which is the opening factor, was fixed, always the same. Your fuel load, be it the crib or be it the kerosene, it was every experiment, it was the same. So if you looked on this setup from the perspective of the known fire science opening factors, thomas, sassel and work, you should usually get the same outcome of this test, right, because it's dependent on the opening factor, let's say, and you get a scatter with temperatures being higher. So what does this mean for a fire engineer who's designing?

No, because there were two assumptions for Thomas Sassel and the main one was that, well, basically, that the boundaries were non-combustible and that the fuel was on the floor, and I am not following these assumptions, so I am introducing, basically, these combustible surfaces, and how this affects the fire is that, first of all, now I have a fuel source which is not only on the floor but also on a wall and or on the ceiling, and on top of that the question was like, because usually we assume that we have heat losses through the boundaries, right. And this is part of our energy balance of the compartment. But in this case, if my boundary a wall is actually providing heat, can I still assume that I have heat losses through this wall? Probably not, so. That's why it basically makes invalid this Thomas Sassel and framework, because it's not really meeting the main assumptions of this framework which we knew, and that's why we were checking how we can maybe change the framework.

Of course Thomas Sassel and we're not working with mass timber buildings, and definitely not with CLT, so we cannot blame them for putting these restrictions. But in the modern world we transition for different reasons sustainability and so on. We try to transition into more use of biomaterials, mass timber included. Yet the entire framework we have for design and testing these compartments is very how to say it historically based on the research starting with Thomas and Sassel and all other research. So it's interesting that today, looking through the same framework on the modern material, we see different outcomes, and outcomes that make sense and we can explain. Another thing that you have reported in your paper were things related to the flows and the discrepancies in temperatures in the compartments. Perhaps we can discuss this more Because, if I read it correctly, you've noticed that the peak temperatures in some of the compartments were not at the ceiling where I would expect them to be, but on a different plane of the model.

Yeah, so this is probably something that is going to be also scale dependent, so maybe this phenomenon is not going to be observed in all the timber compartments that we can encounter, but I'm not only the only person that have observed this either, and this has been observed also already in larger scale compartments. So what's happening is that as we are adding more CLT to the compartment, we are adding more fuel, and therefore what happens is that the oxygen, the air, enters at the bottom of the compartment and by the time it mixes with the pyrolysis gases, there is basically no enough oxygen to react at the ceiling, underneath the ceiling area, and that's why you create these very fuel-rich oxygen deprived underneath the ceiling and you develop these lower temperatures. So basically, you will have the highest temperatures somewhere at the interface of these entering flow and exiting flow, so somewhere around the neutral plane height.

So it's where the combustion is happening, not exactly where the fuel is produced.

Where it's mixing. Basically, it's going to happen where the most efficient mixing happens, with the inflow and the pyrolysis gases. Yes, and this, of course, will depend on your geometry, because if you have, for example, only one wall exposed on the right-hand side from the opening, you will have probably higher temperatures. It's not going to be symmetric.

Tell me more about it. Did you observe something like that?

I observed that, yes, the pyrolysis gases going away from symmetry, if we, for example, only expose one wall on one side, and that is observed in this temperature distribution across the compartments but also on the external flame. So if you have a side wall exposed, of course the flow through the opening is going to be more fuel-rich on the side of the opening where this wall is and therefore most of the flame is going to be tilted or basically not symmetric. So yeah, I've observed that and we've observed this in the small scale, in this medium scale compartments, but also in larger scales. But again, it will also depend on how much of, for example, if you have an exposed wall and a ceiling, in that case you have such a huge excess of pyrolysis gases that maybe you don't notice anymore this lack of symmetry. So it's really geometry dependent.

And were you seeing, because all the surfaces were roughly the same area. Were you observing a significant difference if you were exposing, for example, a wall versus exposing only ceiling or, from your perspective, not that big?

In my experiments I observed that the ceiling was charming less than the other surfaces and that was around 30% less on average. Again, it would depend on if I would compare only ceiling versus only wall or if I would compare tests with ceiling and two more walls. The hypothesis that we developed was that there are two reasons to it. One is lack of oxygen reaching the ceiling. The second reason was that we were creating. There's a smoke layer underneath the ceiling. Basically it acts as a radiation shield. So basically there was less heat relative heat from the fire source reaching the ceiling, and I also cut my CLT samples after the experiments and I could observe thicker char layers at the ceiling, which kind of gives another argument to the hypothesis that probably there was less oxygen in that area. That's why the char could not smolder as efficiently as in a wall.

What would be the consequence of the building scale? Less charring rate means better outcomes, right.

Of course, you lower your cross section and you can have higher utilization factors for your structural members.

And in regards to layers, peeling off, the elimination of or glue line bond failure, whatever people prefer to call it. Have you observed any effects of that? Was that influencing your fire dynamics?

Okay, so again, because what I find about my PhD like this is a thought that I had quite recently is that when I started my PhD late 2015-2016, with pretty simple experiments, you had like a huge learning curve. And eight or seven years later, we have to do so many more experiments to increase our knowledge in this topic because, of course, we now understand these obvious things that I learned in my PhD and now we want to understand these nuances that you are now asking me. So in my PhD, I tried to exclude the lamination to start with because, we wanted to understand this fire dynamics and we knew that the lamination was going to add another variable and at that stage we said, okay, let's just try to delete this extra variable because it's just going to be too complicated. So, basically, what I did was that the first lamella that was exposed to the fire so the outer layer of my cross-laminated timber was quite thick 45 millimeters, if I remember correctly and I ensured that the fire duration was not going to go beyond these 45 millimeters. However, I was doing usually three or four repeats of the same configuration and I was always leaving one repeat long enough to achieve the lamination and see what happens. Although it was not the main focus of the study, and it is true that in that scale, every time there was the lamination the fire would just go forever, except for the ceiling cases. Okay, if the ceiling was exposed, only the ceiling, and I would have the lamination from the ceiling, the fire would eventually self extinguish. But if I would have a ceiling and a wall, then there was too much fuel and heat feedback radiation and it would not serve extinguish. And if it was only an exposed wall, the phenomenon that we observed in my test and also in other tests that were happening at the same time was that the lamination would do this pile of char underneath the wall. Okay, and you know, these were just perfect conditions to promote continuous burning, because you would have this pile of char at the bottom where the flow of air is entering, going through this pile of char, and then the combustion products would just heat up the wall. But the lamination is a complex phenomenon and also it depends on the type of glue, on the stresses you put on your element, et cetera, et cetera. So well, we are still studying this, as you know.

As I know, of course, and to be honest, when we were doing the large scale experiments one and two years ago together with OFR, when I first saw this char buildup next to the wall, it was so kind of stupid and obvious, you know, because you would expect a complex physical phenomenon to happen to drive the stuff. But it could be something as simple, as you know charcoal falling off the wall and building up a pile next to the wall where it obviously will burn, and that's a perfect place to conserve the fire for a very long time. Science can be sometimes very simple and just observing that is, I think, very beautiful part of science to be able to correlate some effects with such simple mechanisms. You've mentioned self extinction. So, if I can I mean this is a world winning paper, so I have to be very thorough interviewing you about it so you've observed self extinction only in the ceiling settings. Non wall ceiling settings would self extinguish.

No, no, no. I also observed self extinction in compartments where I had a wall and the ceiling exposed, but not if they suffered delamination.

Ah, okay, okay.

So basically to achieve self extinction, it is again. It's very I think it's quite a difficult phenomenon to predict yet because it is governed by the amount of pyrolysis gases generated at the pyrolysis front of the timber. But there are so many variables that are going to control this pyrolysis gas production in the pyrolysis front that it is very hard to predict. But in the scale of a compartment the main variables that are going to govern this are the amount of exposed surfaces, because you have this re-radiated heat feedback between the exposed timber surfaces. It's going to depend on the ventilation because in the decay stage the ventilation is going to indicate the heat losses from the compartments and probably delamination is going to be also one of the main variables because, as we just discussed, if we have this pile of charcoal just lying at the bottom of a wall, probably this is just going to be the perfect condition to just keep burning forever. And in my opinion it also depends a lot like fire time history. So this was actually something interesting because I did this test with wood grips and test with kerosene, and the test with kerosene my fully developed fire was bigger, with higher heat release rates and so on, so more severe compared to the wood grip tests. But once I was shutting down the pool fire there was basically no more fuel coming out of the fuel source and the decay stage was quite steep. And then the heat generated in the compartment was only coming from this exposed CLT surfaces, right no-transcript. Whereas in this other experimental campaign, where again the geometry was the same and everything the same, but with wood crepes, I would not have such a severe fire in the full level of fire, but then the decay stage was just slowly, slowly, slowly decaying for hours sometimes and I had configurations which achieved self-extinction with simple fire, even though the full level of fire was more severe, but did not have self-extinction with wood crepes, because these wood crepes also ended up in charcoal which were delivering enough heat to the compartment to promote continuous burning. But again, this might have been also skill dependent, because this small scale was just also like a perfect context to keep a lot of heat accumulated in a small volume. But an interesting observation, I would say.

Well, self-going fires. You know, if you have a beam next to a wall and a small chunk of exposed ceiling between them is very the same scale as your compartment. And from the tests that I know, you know and we kind of talk about in here, we know that the fire can go on in such nice small spaces for a very long time. So even though your scale was, let's say, medium scale, it also represented it for some construction details in parts of the construction where the fire has a lot of surfaces which can interact with each other. I think this finding is actually quite interesting here. The way how self-extinction happens is very fire dynamic dependent. You mentioned the steep burnout curve or very long burnout process. A lot of people would just, you know, want a single number at what number the burnout happens or at what number the self-extinction happens. And we also know that in many design projects the self-extinction would be the number one criteria. Will it self-extinguish or not? That's the question asked, because if it does, then we're great, if not, then we perhaps have an issue. But the matter is of course, much more complicated. What's your take on this? Like having this experience, this knowledge? Self-extinction is also the number one design criterion for you.

I like a lot the approach of Dr Juan Cuevas. He basically gives it like a probabilistic approach. As I just said before, there are so many variables that affect self-extinction. So, for example, if the moisture content is whatever 20%, then your probability goes up to 0.8. And if your heat flux is above 45, then your probability goes higher again. But it's more like a probabilistic thing, I would say, dependent on several variables. And what was happening in my experiments this comparison of the kerosene and the woodchips was that the woodchips had this very slow decay phase and that was inducing, like this, heating condition across the whole CLT slab. So the timber behind the paralysis zone was already hot, so you didn't really have heat losses behind the paralysis zone. And the moment at which I observed self-extinction in this test was not this typical benchmark number that we use as a threshold for self-extinction, which is around 40 or 45. Because this number is, for certain heating conditions, right when we assume that your temperature gradient in the timber basically has a certain shape. But if the whole timber is already preheated and it's not behaving as a semi-infinite solid because the thermal wave has already reached the other side of the wall, basically the temperature gradient is different and you need much less heat coming from the fire to keep your wall burning. You don't have heat losses in your paralysis zone. And now I'm going to get into the details of paralysis zone.

No, no, this is excellent, but this is important for people to realize. The differences are in such tiny details and perhaps just talking about one value of heat flux at which it will self-extinguish, and that's our critical value, whatever. That's perhaps a bit naive at some times. So if the fire was a very long duration and the heat has penetrated deep into your structure, this would be different than if you just ignited and then tried to cut down radiation and extinguish right.

Definitely definitely.

One more thing that I wanted to ask, because you've mentioned that you had a facade outside to measure some stuff happening outside of that compartment. It's not in the paper, but I reckon it's in the PhD, which I'll also link to in the show notes. What were you observing outside? Obviously, if fuel doesn't burn inside, it will burn out. How big the differences were.

So it's just as simple as you named it, and you know this is again one of these very obvious consequences, but it is so obvious that it's not really captured in any standard or codes how to account for that. So that's why I think it was so important to get some data about this back at the time. So, yeah, I would say there was I cannot say proportional, but there was very clear correlation that the more exposed timber I have inside in my compartment, the excessive pyrolysis gases would flow out and they would basically increase external flame but also change its shape. Because for configurations where there was a lot of exposed surface, there was such a. The flows at the opening were so big that it was more like a flame ejection, like a horizontal flame ejection, which later would, by buoyancy, go upwards again. But there was a change in flame shape as well. So, yeah, that was also a very interesting finding, although not every surface contributed the same way because, as I said before, the ceilings would contribute a bit less to this. They were burning more or less 30% less than they were, but again, scale dependency maybe.

Perhaps, and you also observed the flame length outside, like how high did the flame reach?

Yeah, yeah. So I mean it's difficult to say that. To quantify this update, we could say that for the same geometry the baseline tests compared to tests where 70% of the internal surfaces were exposed, the flame was five times bigger. But you know this term bigger it can be visible flame, heat fluxes, what exactly was bigger? But yeah, let's say that the visible flame, where your flame is basically continuous and not intermittent, this length was maybe five times bigger and the heat fluxes maybe were like three to four times higher in some sections of the facade, the Ryan heat flux. Yeah, so basically a much more severe exposure of the facade.

And now I'm wondering you take non-combustible structures, let's say different structures, but they have more or less same opening factor. They have more or less same fuel load let's say their offices. You would expect a very similar behavior of the external heat flux to the building coming from the flames vending outside, because the size of the fire perhaps would be similar. If you get more or less same temperatures inside the compartment, you would perhaps have more or less the same thing outside. Here the change seems very dramatic, right?

Yes, it is true that so far we had over the years we had tools and models to more or less predict external flaming from compartments. You know, margaret Law, cecilia Abecasis and there is a we have a history of studying this and these tools and models. They usually assumed a fuel load that was typical for typical residential, commercial, whatever compartments and then the model was doing correlations with the compartment geometry only. But again, here I actually tried to apply these models to my compartments and of course they were underestimating a lot the external flame because again, the assumption of the model was that basically your structure is not adding more this extra fuel. So it's definitely like a huge parameter that has to be added to these models that we've been using so far.

Yeah, I'm pushing you because you know I'm very how to say it? Sensitive about the current paradigm that emerges from Noncombustible structures. All the experiments done 60, 70s, 80s, 90s, you know that gave us all the knowledge we have. We base some rules that go into the law, to the codes. You know the separation distances, downstones, things like that which worked in the previous paradigm. And now comes the timber and, and seemingly it's the same thing. It is completely different when you start looking at the, at the fire dynamics, and suddenly, perhaps in some cases our assumptions perhaps still hold, but other are Could be so far away from, you know, working that it's actually quite disturbing. And then this is stressing me, stressing me a lot, and it's very difficult to get that message through. It's very difficult to tell people it's not just the fire resistance, it's not just the firing rate. Look, there are so many different components. Look at that. And again, it's also not that we cannot handle it, we don't know anything, we know a lot. It's just you have to meticulously consider those aspects and if you do, there's a good chance you're gonna have a safe structure.

Yeah, I fully agree with you. It's not to blame anybody, but I find fires a physical phenomenon that are, you know, we all understand many things because it's based it's not too complex physics, the the difficulty of fires is that there are so many variables that this is what makes them complex. Right, and of course, to be able to somehow manage this whole fire problem was to narrow it down to certain scenarios and to encapsulate Different fires into different regimes so we can apply models to more or less predict. But now we have introduced a huge, huge, new, big variable which was not present before, and that's why I call myself like a person that has contributed to define a new Fire dynamic model for for timber compartments, because everybody that studies fire safety we understand what fire dynamics in compartment is. But this fire, this typical framework that was used until now has changed a lot for for timber compartments. So the fire dynamics has changed and therefore we have to also update all our models for Temperatures inside, for external flames, for self-extinguished whatever that are going to represent this new fire dynamic, because this new variable that we introduced, this new parameter, exposed timber. It's basically influencing a lot and it's not negligible anymore.

Yeah, so now for a grand finale of the episode. Let's talk about the final finding, because you have a plot on which, based on the Thomas Hesselden opening factor, your points are scattered in the vertical line, making no sense, let's say, and then, one page later, you have a plot which Shows they suddenly fit this growing and decreasing curve of, let's say, thomas Hesselden framework. So so tell me about the modification of the opening factor that you've proposed and how it Can be used by engineers.

Okay, first of all, it was the grand idea was of Jose Torreiro to do this adjustment. Okay, I was not thinking alone, alone, in front of my desk, the one having this Brilliant genius ideas. But it's a very simple proposition. Basically, in this Energy balance that is represented as an opening factor formulation, the term of heat losses is represented as the area of your boundaries, your walls and ceiling. And we said, okay, it doesn't make sense anymore to assume that an exposed seal, the surface, is gonna be a Heat losses surface, so we are going to subtract from this term the area of exposed ceiling. So basically, we modified the term that represents the heat losses from the compartments and we no longer assume that exposed timber acts as a Heat sink so.

So if you, let's say, had a hundred square meters of walls and 50 square meters of them would be CLT, then you would just calculate the walls as 50 square meters, because that's the area where you will have Heat losses.

Yes. Okay and it's also important to mention. So I think that for the test, the baseline test and the test with I don't remember exactly 15, 20, 25 percent of exposed timber, these temperatures they match Thomas and Heseldon because they are still within this ventilation control regime and Beyond that point the temperatures start to decrease. In my case they start to decrease because the excess of pyrolysis gases makes this inefficient combustion inside the compartment. But the coincidence of these points with Thomas and Heseldon plot is just pure luck. Okay because the Mechanisms Lowering down these temperatures are completely different. In my case, the temperatures go down because there is a these excess of pyrolysis gases, the the combustion is like super fuel rich and it's just too much for. And in Thomas and Heseldon what was happening was that he was having huge heat losses through the openings and also a lot of air coming into the compartments. So instead of being fuel rich, it was oxygen rich. So it is kind of the the opposite mechanisms and it's more like a coincidence that these Two loads are kind of matching each other. It's a coincidence, but yeah, it's a nice one this time.

Well, it's a nice one, and it's actually great that you have a sound explanation of what's happening. So, in this case, evaluating the severity of the fire using this approach, the, the CLT, shall be taken account into account, not a surface, that that takes away the heat from the fire. Plus, on top of that, you should consider all the other aspects that CLT brings into the table that we have discussed so far. Is that something I can recommend to my designers?

I think that we could very simplistically look, look at what are the principles of any fire Strategies for buildings means of escape, compartmentation, no vertical flame spread, like these points. And I think that Probably for most of these principles, the presence of CLT is going to have an impact and it's not only about suffix extinguishment, which very often that's what architects thinks, for example, it's the CLT is introducing new challenges in all the stages of a fire safety study. You know different, different flame spread over surfaces. Vertical fire spread, once the fires fully developed, over the facade. Nos of cross section. If your building is a high-rise building and requires to withstand Burnout, you have to ensure suffix extinguishment. How do we do that? It's a jet. So it's, it's really at all levels of fire safety strategy fantastic.

Once again, thank you for joining me and Congratulations for your award once again. I believe after this episode is quite obvious why it is a well-earned award, because this is not every day. Someone that defines new, new, new view on Fire dynamics. So much appreciated and thanks for being here and I'm super happy to work with you more on that. In Short future, perhaps once the stuff becomes public, we will meet again and discuss the newer findings on on this field.

Thank you, thank you very much. Thank you again for inviting me. I also look forward to work in the future, in future projects, and Again, thank you to everybody that helped me to do this Fantastic research and that's it.

Thank you for listening. I guess after listening to the interview, it becomes apparent why this particular paper was chosen as the best from the previous symposium. It's not your everyday thing to find a new fight dynamics or find a new correlation to Thomas and Heslidan data, which is a cornerstone for many fires of the concentrations, as I have said at the beginning of the episode. I hope you found this interesting. I hope it helped you understand why the mass timber in fire is something really new. I mean, it's not just a new construction material. It changes the fire itself and how it changes the fire itself. We need to understand that. We need to take this into account when we are designing mass timber buildings, your next future mass Timber skyscraper. You need to be aware of what is happening to fire regimes, to the fire itself, by introducing timber and Carmen, as she said, when she started this job, this work, there was so many unknowns. She was doing that when we literally have had very little idea on what will be happening and it's really funny that today some of the things that we have discussed with Carmen they they're kind of obvious. You know, we know that, but we know that because of her research and other research that has happened in the past that moved us further, and today we can dig much deeper into technical details and you know some specific, specific issues around mass timber. But it's it's those big studies at the front, exploratory ones, that put us on this pathway, and I'm very, very pleased that I had the privilege to interview Carmen about one of such stories. Thank you very much for listening. That would be it. I am on my way to Slovenia for SFP conference on sustainability and fire and Looks it's gonna be an amazing event. I'm looking through the list of guests and I'm just astonished about the speakers list. So if you are in Slovenia, please say hi to me and I would love to meet you in person, and if you're not, well, you will have chance to meet me once again next Wednesday in the next fire science show episode. Thank you for being here with me. Cheers, bye you.