255 - Timber load bearing capacity in fire from nano- to megascale with Felix Wiesner

Wojciech Wegrzynski: 0:00

Hello, everybody. Welcome to the Fire Science Show. Uh, the job of podcaster is kind of difficult in interesting topics and, uh, great people. It's a, a lot of work, but every three years this work becomes extremely easy, and that's, uh, when we have an international symposium on fire science by the IAFSS. It's easy because for the symposium we just massive, massive amount of fantastic, uh, fantastic talks invited to give lectures in, and I just browse the conference program like a catalog and, and chase those, uh, people with microphones. Regardless, um, before the conference, we also had, announcements of awards, uh, presented on the, on the conference. Um, we already had Professor Negar Elhami Khorsani f- we-- who, who won the Magnuson Award, and today award presented to, uh, early career researcher, say. Uh, that's Guylène Proulx Award, which this to Dr. Felix Wiesner from the University of British And, as you can imagine, Felix is the guest of this podcast today. I'm super happy to have him. We-- I've been co-chairing the research of the IAFSS with, uh, Felix, had a lot of with him across the last three years, and and I'm very happy that, uh, i-it's my friend who got, who got this, uh, fantastic recognition for his to the fire science so far. And those contributions, they come in the timber, and, uh, in this episode, in this and in the paper that, uh, follows the award you can find online Felix investigates or the state of art knowledge, what we know about in relationship to fire, but timber like material, not, not construction products, not CLT, not, just, you know, wood. What do we know from the large scale starting at resistance and furnaces and, open air, uh, large fire demonstrations through medium scales heating up, crushing the blocks, looking at at, you know, a scale of a plank or, small beam or a cross-section. We discussed the moisture transport, and then we into micro scale. What happens with the cellulose? What happens at those smallest scales and to what actually we know, what we do not know, what we that, uh, we still need to learn. And, uh, I think, uh, this, this journey is a fascinating one. And as I said, we keep it to raw timber, so, uh, limited amount of knowledge on, on adhesives and, and other things. I think it's, it's really great because we also this type of view on the problem. So, um, without further ado, I think you will enjoy this one. The award is being presented tomorrow as the is released on Wednesday, awards on Thursday. So you have a sneak peek of, uh, Felix's I know the award is well-deserved, and this just proves it. So, uh, stay with us, uh, and, uh, enjoy the with Dr. Felix Wiesner. The Fire Science Show podcast is brought to collaboration with OFR Consultants, a independent consultancy dedicated to addressing safety challenges. OFR is the UK's leading fire risk consultancy that this year celebrates its 10th anniversary. As experts in fire engineering, they are fully to delivering preeminent expertise to protect property, and the environment. With over 30 chartered engineers and a team of researchers at their core, they continually explore the challenges that fire creates for their clients and society so that the best research, experience, and diligence can be applied for effective solutions. In 2026, OFR will grow its team once again and is keen to hear from industry professionals who want to collaborate on fire safety features this year. Get in touch at olafarkconsultants.com. Hello everybody. I am joined today by Felix Wiesner from of British Columbia.

Felix Wiesner: 4:54

Hi, Wojciech.

Wojciech Wegrzynski: 4:57

Welcome, uh, back to the show and we need to off with, uh, congratulations on you receiving the Proulx Award from the IFSS, for young, uh, with great achievements. Uh, congratulations on that, man.

Felix Wiesner: 5:13

Yeah. Thank you very much. I, I'm not sure if I would use the word young. It's, uh, just the-- I think the criteria is, uh, five from your PhD at the time of nomination, so it can be any age. not so young anymore. nonetheless.

Wojciech Wegrzynski: 5:27

can. Enjoy while you can. I just dropped off so many young, Uh scholar and I kind of miss them, you know? So Nah, but it's, it's, uh, it's not just, uh, uh, you are just five years of your-- from your It's because the science is excellent, and that's the only thing that goes into the award. And your excellent science is touching also a that is extremely important to modern fire engineering because you are dealing with the and its, uh, structural capacity while in fires. And I think, you know, uh, t- for me, you know, batteries, timber are the kind of main themes of modern fire science or the popular ones right So I'm, I'm, I'm very curious about your research because it, it adds to the overall pile of we know. So, you were giving a, a special lecture on that the IFSS. I will use my time-traveling abilities. It's actually tomorrow So, uh, welcome to lecture, and you can get the sneak peek today. It's, it has a great title. I love the title. "From nan- nano to megastructures: A review of mass timber load-bearing capacity in fire." Uh, me like why are we looking at this from the whole range of scales? Uh, and how, how nano is nanoscale actually?

Felix Wiesner: 6:47

I mean, in this case, down to kind of the, the fibers and actually the cellulose.

Wojciech Wegrzynski: 6:53

Mm-hmm.

Felix Wiesner: 6:53

Um, so not quite atomistic scale, but nanoscale. and I mean, the title's obviously a bit, uh, uh, attracts clicks, attracts the search engines. Now it's, uh, gotta play the game, but also it's-- I, I mean, I generally had an interest to kind of, find out often we, we look at structural capacity very much at a structure scale. Now you at-- if you look at fire resistance, you take beam, you put it in a furnace, it collapses, you measure the time. done. Uh, you record that. If it's above your ninety, hundred and twenty minutes, you're aiming for, you're happy, but you're not finding anything out. Or even if we do strength tests on timber that's we work with the members, and we don't really happening in the material. And now, for three years, I've been at, uh, University British Columbia, and I'm in the faculty of forestry in the sci- in the Department of Wood Science. but I'm not a wood scientist. I don't really know much about trees. so for me, this was an opportunity to look at it more a different lens and see what happens at the smaller that drives that behavior. Um, so basically, the way I have-- trying to address the, the talk and the paper is to kind of start from the and say like, "Okay, this is how we test timber," and work my way down and try to explain what's happening. And yeah, that's how I came to this somewhat title.

Wojciech Wegrzynski: 8:17

Nice. Uh, so the place you're in, uh, that is, is, interested in timber. I w- I wonder, like from non-fire perspective, do we have the, the full comprehension of, of of, of load bearing in timber? Uh, like h- how much do we know before we put and fire into the, into the problem?

Felix Wiesner: 8:39

I mean, we know a lot, but even without fire, this-- very empirical, no?

Wojciech Wegrzynski: 8:44

Hmm.

Felix Wiesner: 8:44

you test it, and then you get your, um, models of rupture, um, which some people also call bending strength, which, which isn't really a strength in its, in its own way. It's, it's a very test-dependent parameter. and you get your elastic models, and then that's and you ideally, you many pieces of timber, and you come up with some relationship between the grading and that way y-you can then say like, "Oh, I have grade

Wojciech Wegrzynski: 9:12

Mm.

Felix Wiesner: 9:12

I'm expecting this range of strength and this range of modulus." still, I mean, obviously, there has been some research, on-- or there has been a lot of research on the microfiber angle will affect these properties, but doesn't really flow into the engineering component. That's just the why. but in engineering, it's very similar to fire that you at the code that tells you these are the limits you to do, these are the calculations that you do to your stress state, and then you compare them. And depending on the safety factor you want, you then "Good," or, "I need to readjust by making the member

Wojciech Wegrzynski: 9:49

Hmm.

Felix Wiesner: 9:50

so, so that's very similar, but there is a lot of that actually looks at, what happens to the strength of the timber based on, um, the orientation of the fibers. I mean, this talk and paper, I actually really simplified it. Um, previously, I worked a lot on looking at adhesive, at cross-laminated timber. Here, I really am only interested in wood, timber, no no connections, no crosswise fibers. So basically take a perfect piece of wood, everything is in parallel, what happens when we heat that up in of its structural capacity?

Wojciech Wegrzynski: 10:27

from my very, very limited knowledge about fire engineering and timber, uh, what, or what I was taught is that it's, uh, anisotropy plays, plays a hu- a huge role, like w- w- what's the orientation of the fibers. Uh, but also there's a lot of elements of because it's a living thing kind of. Like, I, I mean, before you put it in the it, it was quite alive. and then elements such as the moisture, for I was told that they ha- play a great role. to, to get this class of timber, to get these of timber, that modulus, that, that model of do you need to break the piece of timber, or can predict that sufficiently knowing just some properties like the-- It was spruce, it was like amount of moisture, 13%, and density was 570. Can you, based on that only, robustly define the of, of the material in the structure? how much you can tell me about how it's gonna in a structure?

Felix Wiesner: 11:32

Yeah, and I think your question was also do we have to it to find out what it has? Um, I mean, luckily not because we cannot test every of wood, so if we break them all, we can't use them. Um, so usually what, what happens is you do research, you take 100 pieces, you, you measure all sorts of things about them, break them, and then what you do is What they used to do is a lot of visual grading, so who is a wood expert who worked all her life looked at it and see like, "Oh, there's so many knots. This is grade X, and the other one is grade Z." Uh, countries have different-- Sometimes it's numbers, it's letters. and then they would say, "Okay, we've, we've already at so many pieces of wood from this grade, so we know its strength will be in this range and its elastic modulus this." And, today what's more common is machine where basically you have a-- You hit it with a little and you, you measure the vibrations or the sound Different methods are available, and then you have a that then tells you the elastic modulus, which then you the grade you can expect. And then based on that, you can kind of estimate, a of strength that you would expect. And then obviously like, um, we do it in engineering, like to be conservative. You'll take the fifth percentile, or the 10th percentile depending on how confident you are in your wood So generally with engineered wood, you expect a bit of narrower range of capacities. So then you don't have to go on the, on the super low of the expected strength.

Wojciech Wegrzynski: 12:59

let's, let's move, uh, through the scales of, in fire, and let's start where you said you'd to start, which is the, the, the classical way or the full-scale way of us, uh, determining the proxy of, fire behavior of, of, of structure, which resistance. So l-let, let's start there.

Felix Wiesner: 13:19

Yeah. I mean, you, you're already used to word proxies or fire resistance is basically a substitute of, implicit that we're using now. We're saying like, "Oh, something has 120-minute fire then we'll be safe." And that has kind of served us quite well, and it's not very hard to do that from a, at mentally it's very-- the concept is very easy. Now, you, you, you take your timber, you load it, and burn it, and then you, you start a stopwatch, and then the performance criteria are not given anymore, you, stop the stopwatch and then you say, "Okay, now I have 90, 120 minutes fire resistance." And then you put that in a report and now you can use it, and then it's deemed safe for the building that has that fire resistance that you're targeting. So conceptually very simple. The problem is it's very expensive. And now one problem that we're seeing is that as we are building bigger timber buildings, they need bigger Um, so sometimes you can't even find a furnace that can run timber. And then there's also that issue that timber burns other than steel or concrete. So you're adding additional fuel to your furnace, then makes it harder to mix it. So there's, there's some arguments, um, I don't wanna go down that hole now, if that's an accurate but the issue is some furnaces won't even take timber in some countries, um, or large pieces of timber. and then if you, if you think about some of these taller timber building, when you're talking about 18 stories, are now pushing for 20, if all the load-bearing comes um, via timber, you need massive timber columns at the And to test those timber columns, there's not that furnaces that can do it. I mean, you have the, the Mjøstårnet, which for a while was the tallest timber building in, in the world, in, Norway, and they have relatively big columns, and actually tested them on a furnace but unloaded. Uh, so I don't think they ever tested them fully loaded, so we don't really know how they behave under load. Um, another example is there's a company actually from or from Canada, and they have an office in Vancouver. So, so they developed this, concrete timber system, which is, pre-tensioned, and it allows you to do longer but it's a massive piece of timber that already has, into it, so a lot of furnaces won't take it. So they couldn't find a furnace globally that would do fire resistance test for it. if you think about some of the even bigger buildings I mean, Sumitomo Forestry has proposed for their I mean, they still have a few years for one of their I think it's in twenty fifty-four, they want to, to a three hundred meter tall timber tower in Tokyo. So that will need huge elements, and at the moment, can't really test them, at least not with load um, for fire resistance.

Wojciech Wegrzynski: 16:08

yeah, i-if you think about it think about trees, sequoias. You're in the part of the world where you perhaps have, like, a short 11-hour drive to them, but, you know, at, at the bottom they're pretty damn And, uh, and it's just a, you know, a, a live of timber transferring, uh, its own weight perhaps wind, which is not, uh, you know, not as at this scale, but, but that's it. It's not handling, uh, humans, floors, everything else, and it's, it's like, uh, meters in diameter at the bottom. So that gives you an idea of how big the elements have to be at the bottom, uh, when it vertically up. In terms of a furnace, I'm-- we're pretty crazy we have huge furnaces. Send them to us. I could do a two, two-by-two meter column and it to whatever you like. So yeah, a, a short, uh, input perhaps, uh, like get a bonus. I don't, I don't think so. but, uh, anyway, I see a lot of issues in, in timber in, in, in furnaces and there has-- had been entire podcast episodes on me and my guests the fire resistance as a proxy and why I it's the best proxy we have. For me, the biggest issue perhaps not structural engineering, but it's more like I- can I say that? Uh, like people are-- just would assume equivalency, you know, 60 minutes of concrete beam, 60-minute beam, that's the same thing, and, uh, for me it's very not the same thing. Uh, but that's, that's just another thing. How, how does, uh, in practice this fire resistance of timber elements translate into structural capacity of a building and structural, you know, response of the building to fire? I mean, yeah, we-- it's a proxy. We assume it's gonna be good, but in reality, if wants to go further, how much they've learned their structure knowing only the fire resistance of the timber?

Felix Wiesner: 18:09

Well, I I don't think it features into the structural that much because basically the structural is done by the structural engineer, and they will memos appropriately, and then they will check is the fire resistance good enough. And if that's a yes, then tick, um, job done. But there's no direct connection between what happens to the structure in a fire and fire resistance. It's just, okay, we got the sixty minutes, that's fine. Um, like I said, I don't really-- you already said this has been discussed a lot, so I don't really wanna spend too much time on the, the whole merits of fire resistance as a thing. But if we-- even if we accept that, okay, fire works, it's what we use as a safety, we're now at a that, okay, we have a lot of timber, either we, we are to test it. Um, I mean, we can maybe send it all to you, Wojciech.

Wojciech Wegrzynski: 18:57

Very happy.

Felix Wiesner: 18:59

o-okay, um, or we can test it, or it's also very So what happens actually in practice is, um, we use to, to calculate the fire resistance, and also now we proxy. and, and that's kind of where the review starts. It looks at, like, methods how to calculate fire um, that has been done for a long time. Some of them are just empirical. I mean, you can go back to the '50s where, um, in the Maholtra and Rogowski tested, like, various columns, and they came up with a fairly simple system that was just species, size, adhesive, and some sort of empirical that then gives you a fire resistance of those. uh, obviously then through time, people have developed sophisticated methods that actually look, okay, if we load timber that much, we have this stress, and we need to that much timber, to then again turn that into an correlation to get fire resistance. But we always gotta remember that only gives you fire So it doesn't predict anything that will happen in a real fire. It gives you the capacity in the furnace. Now it kind of, saves you the, having to do a furnace And of, of course, initially you still have to test your calch-calculations work. And now that we're seeing a lot of innovations in the space, new sizes, cross-laminated timber, LVL, new of connections or load-bearing capacity, that also a lot of questions of like, okay, how well do these calculations work for new timber products? New... Okay, I said I didn't wanna talk about adhesives. I have new adhesives and everything. Yeah. So, so that's another thing where we're kind of running into questions. And, so one of the most popular ways to calculate fire is through the reduced cross-section method, which we assume we know the charring rate, Uh, so around 0.65 or 0.7 millimeter per minute is, what happens in the and that's generally fairly accurate. that doesn't mean that will happen in a real fire, but we're in the furnace. We just look at what's happening in the furnace. And then we also need to account for heated timber below the charring, and then the reduced cross-section usually just adds a fixed of zero strength to it, the, famous or infamous zero strength layer. And so you say like, okay, my charring rate times my fire resistance that I need plus whatever that zero layer is, gives me a residual cross-section, and then I calculate if that can still carry the load. If yes, then that's fine. So now here I actually have a, a direct link capacity and fire, but again, it's only valid for the And we're now also saying in the new Eurocode, they're changing those values. Now for a long time, um, this was based on Schafer's in the US, which-- where it was like 0.3 inches, um, then translates to roughly seven millimeters. So you say, 0.65 or 0.7 depending on the element you times the time you have of the fire you have or the resistance, um, plus seven millimeters, and then you just take that off and treat the rest of the timber as So then that's your, your substitute for running a fire resistance test. But now the new Eurocode actually changes it, and so for example, for columns, you should use fourteen if you look to Canada or Australia, for example, they have the seven millimeters, and I don't think it's change that much. And you gotta ask yourself, okay, but, like, do they something we don't? Like, why do we... Like, why does one jurisdiction decide much more than the other?

Wojciech Wegrzynski: 22:31

is this number related to the type of the timber, the conditions in which it is, the, I don't know, the location of the timber element, or it's just, you know, a, a nice number to work with because always the same and then you're, you're done? And, and what's the safety margin on it?

Felix Wiesner: 22:46

couple of questions here, so I'll try to order them in Um, type of timber is not usually considered. So type of timber, some jurisdiction give you a charring rate for type of timber, but the layer is usually not affected by it. the value comes from basically looking at the, the, that drops pretty steeply, in timber and kind of the assumed strength loss that happens in that region one section. So rather than having to deal with a gradient, you just say, "Okay, I take a certain amount of that, make it all zero, and that should amount to the same loss than if I take the gradient." Yeah, and it comes from Shaffer did, simulations on glulam beams and, and that was it. And, then we have ran with the seven-millimeter since. and you asked, like, what's the safety margin And, you Fahmi et al. have done some tests where they actually really tried to look at what are the actual properties of the timber. So they had big timber beams, and then they took out of them. So they had a really good control over the ambient and the elastic models, and then tried to the zero-strength layer. Um, or they did back-calculate the zero-strength from fire resistance tests. And then they, they s- they actually said like, "Okay, the average zero-strength layer that we get was nine," so pretty bang on to the seven millimeter. The problem with that, though, the coefficient of was like something like sixty percent. I, I don't have the number in my head, but it? was pretty large, 'cause they sometimes got a layer of, of twelve, sometimes three millimeter. So the spread was huge. So, you could say, like, pretty good accuracy, pretty precision. but when we talk about factor of safety, precision no? We, um, we do want to, to work with that. So, so that's an issue and, and the new Eurocode kind works around that by having loads of different layers for different situations, for cross-laminated for columns, bending. Um, so it's, it's all different, which kind of... I mean, it's a-- The original concept is very nice it's very simple, no? Everyone can kind of do it. Every can-- Everyone can multiply a charring rate by time and then add a constant value. Um, the Eurocode kind of goes a new way where you now have to look up loads of things and loads of situations, so lose a bit of that simplicity. Um, I haven't actually seen the data what it's so I can't say how merited it is. But for example, I, would say that the increase to mil-millimeters or doubling the zero-strength layer for columns, is definitely justified based on seen also in my own research that compression is just affected by heat. So it doesn't make sense to have the same zero-strength

Wojciech Wegrzynski: 25:26

you, you just touched the, the, question I wanted to ask. So i-is there a big difference in bending and for, for that purpose? Can, can you go a little bit further why is, Um, more affected than men- bending?

Felix Wiesner: 25:37

yeah, I mean, just the way that the fibers work and and the, the individual components. yeah, we can go into that when, when we move down the a bit. Yeah. Yeah, to, to, to find that out exactly. But yeah, in general, I mean, another complication when you work with timber, because if you work with steel, your tension and your compression, at least the reduction in strength and stiffness works the same, so you to, to do it. And, and actually with timber, if you have a timber in bending, uh, it can be quite tricky because, if trying to fi- if you're trying to actually do the calculation, you need to know where your neutral axis is. So the neutral axis separates, where you have, tension compression. Yeah. And for that you need to know the elastic model or through the section, but the elastic models itself is dependent on, on, on compression or tension. So first you have to guess where your neutral axis is, then you have to iterate because you don't know yet, how much reduction there is throughout the beam. So it, I mean, in the end, people just assume the neutral axis is somewhere and it kind of works because the is not that great initially. But, uh, in more complex, uh, elements like CLT, stress distribution is kind of all wonky, it does become relevant if you want to know, because especially if your neutral axis is in a crosswise layer or not, um, again, I'm, I'm talking crosswise even though I promised I but, uh, um, here we are. Yeah.

Wojciech Wegrzynski: 27:01

No, but I mean also the thermal field, if you if you're working with something like CLT and face an issue of delamination or heat-induced you also will have those, steep gradients of eventually appearing, uh, so that, uh, I guess uh, charring front and in consequence the zero layer is gonna move quite, rapidly. It's not gonna be this nice single line from start of the fire till the end of the fire, connect dots, that's my, that's my line. It's gonna be, it's gonna be way more complicated i-i-in time. And we're talking about, that to the best of are not that easy to predict, like when exactly gonna happen.

Felix Wiesner: 27:41

Exactly. Yeah.

Wojciech Wegrzynski: 27:44

Le- let's move, let, let's move to the another of the large scale, uh, uh, fire behavior that, i- is considered, and that's the decay it's also like, uh, we also had a podcast episode in here with, uh, Thomas Gernet, with Johann and, uh, they shown-- uh, th- there was a video of column crushing in a furnace after the fire test, and that was a massive, uh, LinkedIn phenomena of that video. I, I was mesmerized by it, uh, as well. te- tell me h- how this is important and, and, uh, why you need to go to nanoscale to understand it if you, if you need.

Felix Wiesner: 28:21

Yeah. I mean, so far we've talked about, different calculation methods and, we kind of landed on the reduced method and how it differs a bit, and I'm not the fan of that method anyway. Um, but yeah, in the end, like we said, they're all for calculated fire resistance in a furnace for the fire, yeah? So you can't do performance-based design with that your performance is what's my fire resistance, which, isn't very interesting. Um, but if you want to consider a different fire and fires have a decay phase, then you're now looking at, that work. Um, yeah, so, Thomas Journet actually, took that video I will use it in my presentation because it's excellent to show that. it's actually something I also saw in my own PhD I also had, uh, CLT kind of wall strips, that first I to failure and then I did some tests where I heated them a bit and then turned off the heat and then some of them failed, long after they're, they're, they were heated also we, we did some tests previously at CERIIG where the same thing happened. I mean, in that case it was smoldering. Um, so that's kind of two things here. Obviously smoldering, if you can't stop that, then you, will lose your element eventually if it's because then the heat keeps generating. Um, but more important is in a decay phase, if manage self-extinction and you can stop all your timber might still collapse as that work has And I think, um, through a series of papers, um, Journet's work, I mean, there were lots of other involved, but he's kind of the, I think the combining Um, they tested 12 different columns, um, that all were designed with the newer, more onerous Eurocode to have fire resistance, and most of them were just heated for 15 minutes of fire resistance and- So basically run the for fifteen minutes and then switch it off, and they all failed. None of them survived. And now we need. to ask ourself, okay, how good of a proxy is our, uh, cross-section method, um, for a proxy of fire resistance when we see all these failures? and obviously people can say like, "Oh yeah, but fire is not a real fire. It just, uh, tells you like a ranking. And if we say this ranking stay, it's good enough, and service sprinklers all of that." But we need to have conversation that well, clearly, even with a low fire you will see collapse if you can't get water in there enough. and if you don't do that, you're putting a lot of work a lot of expectations on your sprinkler system and your fire service. and generally that's not how we should be doing fire engineering. If we want to do that, that's a different discussion, I don't think the fire service would be very happy that. Um, yeah, so, so that work has been super important other works. Um, I know also, uh, Andrea Lucchini has done some work. he hasn't really done any actual tests, but looking at, okay, what should be the zero strength there in a compartment fire based on calculations? Yeah. So very often if we can't do tests, what we do is, and done it a lot, is you take a compartment fire, you have the measurements or the models from the temperature, and then you use the known reduction curves of, okay, models goes down this much, strength goes down this much. When do I have failure? And then you back calculate that and either you say a problem here or you come up with a method that I modify my reduced cross-section method, um, to a

Wojciech Wegrzynski: 31:53

I, I have a experience related to that. Um, not, not in measuring that the residual but the importance of that capacity. We were doing a large CLT compartment experiment we had a large CLT slab, like, I don't know, 60, 70 square meters, quite large. Uh, it was-- It had this huge, uh, beams, glulam It was very nice structure. did the experiment, everything. We, we, we were happy with it, and then, uh, like we're finishing up, and I know that there's like... I see there's smoke coming from behind the beam. I only had one opening to that room, and I see smoke coming from behind the second beam. So I know there is some combustion processes there, and you know, based on my experience from fire laboratory, if I have my CLT smoking it's gone by the morning, you know? And it's-- Or, and we're talking about structure loaded with like 20 tons on it, or even more. Uh, and, and I'm like, "You know what? Let's, let's try and, and take it down, uh, and, let's play firefighters." we've tried you know, how to-- how can we bounce the water the wall so it reaches up that little point where I knew there's combi-- I, assumption is I cannot go inside. It's a loaded post-fire structure. There's no way, like standing three meters away from the structure was already stressful, you know? And the way how we, uh, did it is w- it was, uh, um, a structure built with, with ma- masonry. So we just, you know, the hole at the side of it and put the nozzle and just sprayed the water directly through that nozzle. But we had the access to external wall through we could h- hit a hole in. now my mental, you know, exercise is that, uh W- what if we are, uh, in a timber building, and you're not really able to cut a hole on the side of the to, to, to introduce the water exactly to that All I needed, it was a cup of water, Felix. That's all I needed, was a cup of water on that spot.

Felix Wiesner: 33:57

Well, you need it in the right place. Well, because we, we did some tests in Australia, they unloaded, so we could actually go in and we had some somewhere. Um, we knew roughly where it was, and we put so much on that, and it took us still one and a half hours to get that smoldering spot out. Uh, we, we took off the plaster board from the inside, water on. In the end, probably used as much water as if we would filled a big tank and just lowered a compartment into the water, splashed it in. So it is an issue, but again, not really. Though smoldering is a separate issue, you need to it, because like I said, I mean, that's always the step when we talk about timber. And one of the things when... So when I got the award, I was very happy, and then asked me, like, uh, "Write a review paper on a topic of your expertise." And I was like, "Well, my topic of is, uh, mass timber and fire." But therefore, there have been so many review papers and so much work on that. Um, but all of these reviews have really focused on the fire dynamics, um, be it smoldering or delamination or the secondary flashover, whatever you want to call it. And that makes perfect sense because if we can't that, then you can have the strongest timber in the the strongest structure, the, the biggest fire If your fire is infinite, it doesn't matter, And we that actually loads of-- When we look at examples where timber failed structurally in a fire, it's usually it's still on fire. Uh, I mean, obviously the decay phase, we know from tests is, is an issue. There have been some examples also of, uh, structural after a fire. But really when you have a catastrophic fire loss of a building, um, it's usually because we can't deal with fire. Yeah. But I really wanted to focus here on, okay, if we can that, lots of people have already worked on that. We have, like, a better idea now what we need to do, the structure still has to stand up during the fire and during decay. even without smoldering, you need to make sure because even if you can get self-extinction, you can self-ex-- so you can get self-extinction both of flames and smoldering, your structure still has to stand up, um, and it has to survive that thermal wave that we're seeing in, in compartment fires. yeah. And obviously then, so earlier we talked about, um, kind of empirical methods and reduced cross-section methods. But now when we actually look at, okay, what's the in the timber? We need to explicitly link the strength and elastic of kind of each little sliver of timber to the that exists there. And then we can come up with what's my distribution of modulus for the section? What's my distribution of remaining strength? And then I can do structure engineering, compare, uh, and strain, um, calculate the bending moment, the bending moment, how much capacity do I still have? Yeah. So those are the kind of calculations that we set. We, we now use our, our theory of, uh, mechanical and, uh, structural mechanic, uh, structural mechanics, um, to kind of make that all work. Yeah. But to do that, we need to have the link between strength and elastic modulus. And that's where we, where we go down to, I guess we're still on an empirical scale and the most common one there is actually the one that was in the old Euro code, um, is based on work from, uh, uh, Koenig Valle, Valleg, Uh, I may get the name wrong here, I'm sorry. Um, but, uh, yeah, so which, which is also one of the more onerous ones, which basically says you lose most of your strength and elastic modulus between the range of 20 and 100 degrees. So at 100 degrees, you've already lost 75% of your and 65% of your stiffness or the other way around. And stiffness, I'm saying stiffness, which, um, is, not the same as elastic modulus, but I'm just using it, um, here synonymously for elastic modulus. Okay. So, so then I wanted to know where do these come from and what are the alternates? And I looked a bit into that and, if you look in the there have been loads of work on that. And then they also look at different, uh, timber that different moisture contents, but there's a lot of outcomes. And actually, again, from Thomas Janné's group, they something where they looked at the strength from a probabilistic angle. So they looked at loads of different works. Um, I think, uh, Garcia Castro was the lead author. I hope I got that right. and so they, they, they collated all the data, which great, and then came up with a probabilistic model, which actually turns out less onerous than the current one was in the Euro code, um, that kind of drops off very Um, but then the problem is when you look at where data come from, I think they had 30 data sets, and of 30 data sets, only two or three actually tested So do you know when I talk testing transiently versus ta- steady state, do you roughly know what I'm referring to?

Wojciech Wegrzynski: 38:57

uh, for me, the, the mental model is like I can like a, a furnace which heats up and cools like a electric furnace that I just set temperature

Felix Wiesner: 39:08

okay, so that's steady state heating where the change. But what I mean is you... So transient and steady state in in structural fire so you can either, you can heat something up and then can test it. Like let's say we have a compression member, we put it an oven, we, we set a target temperature, we measure the internal temperatures, we wait. Okay, now it's at 200 degrees, now we crush it. Or you can test transiently, uh, where you load it you, then you increase the temperature. Yeah. And only-- And the majority of, of studies that are out there do steady state testing. Why? Because it's much easier.

Wojciech Wegrzynski: 39:43

you know, my transient model is related to the that I have in use, which is the furnace.

Felix Wiesner: 39:50

yeah. I mean, I mean, obviously the term transient is not to structural fire testing. It's just, yeah, we, we... I mean, if we're really accurate, it should be, um,

Wojciech Wegrzynski: 40:02

wh- when we're working with, uh, with the for who do structural, I am, I'm mostly working concrete people on that, but they, they do that. They either test the concrete blocks, like they it into an oven to 400 degrees, keep it on, in, in that oven for two hours, and then take it out and crush it. Or the ones with an expensive fancy machine can crushing as it's heating. your model.

Felix Wiesner: 40:28

Yes, exactly. and you just say it. like the, the machine is expensive and fancy because doing the testing something with a steady load and temperature is more difficult because you kind of have heat up your, sample while you keep everything else cool, and you need-- you, you basically need a way to do load testing while you heat it up, which is what a furnace Most furnaces that can do loads, now you put your in, you load it, and then you heat it up. So o- those outcomes are transient. Um, but then in a lot of the testing of the actual of what the majority of studies has been steady state, there's also when we talk about elastic modulus, um, some of them have been, um, just on measuring the dynamic modulus. So not even doing load on, But, just, correlating the elastic modulus or back calculating it from

Wojciech Wegrzynski: 41:23

But, but in timber it's such a good insulator. It's, uh, it must have a s- very steep gradient insi- inside. I mean, it-- I know it has. So, you know, the duration of heating and the of the heating curve will also be a massive factor

Felix Wiesner: 41:40

Yeah, I mean, that's also what makes it difficult because to actually-- if I want to test something at 200 if I have a block like in mass timber, generally the is mass- massive and it has that steep gradient, so I never be able to take the block and say, "This is at 200 degrees when I heat it while loaded," because it will two hours to get there. So unless I have a very low load, which is close to um, and then I won't see any failure, which doesn't tell me anything, I can never get a uniform temperature and say, "This is my failure temperature." So I kind of have to, I have to kind of look at where does the

Wojciech Wegrzynski: 42:20

how how upscalable are those results? Like does the size of the element you test matter? Is it like a small block, a big block, uh, just a fiber of a timber?

Felix Wiesner: 42:31

Yeah, I mean, the size matters from a structural point view, because generally the bigger the timber more likely you Is you have knots in there. Yeah? So if you get a, a deeper timber board, um, you can get deeper timber boards, with less knots, but it will cost you more. So the... generally the bigger the timber member, the more likely you have some defects. But if you just take a perfect sample of, um... I mean, if we just want to know the pure material of, of timber, we do something like that for our you can easily get between 70, 100 megapascal of, of strength of a clear specimen. But in reality, you will never get there because, uh,

Wojciech Wegrzynski: 43:14

Okay, but when you're heating it up from the interview we did, I know that there is this transport phenomena. Like there, there's a lot of happening. So if, if technically you have a little sample your furnace, uh, where, where you're pulling or the sample, that moisture has also little space go, where in-- if it was a part of like a one-meter wide column, perhaps this internal interplay of temperature, everything would be different. So, so I wonder like, is it like if you get this of megapascals in this small scale, can you easily translate it into load-bearing properties of the whole structural element?

Felix Wiesner: 43:52

Uh, again, that's the tricky part. Now, now you're, you're touching on, on the issue at which is the moisture. So one of the issues between, like we talked a lot about transient and steady state, but I have actually said why they're different. The difference is that if I heat something to 200 and then I crush it, I drive out all the moisture. And what one thing that actually really weakens timber getting it moist, not getting it wet, um, in like we call the fiber saturation point. Up to that point, strength of timber and stiffness goes down with moisture. So now what happens if I load and I first and then I I drive that moisture deeper into the timber and I the moisture locally, potentially even while also heating it. Um, even worse than moisture seems to be steam. so then I'm, I'm creating, weird strength gradients and I might not have my uniform compression or tension I want anymore. So it gets really hard to do those tests in the first It's hard because like we said, you need a big, uh, a machine, uh, expensive fancy machine, but also of those results get difficult, so. And the kind of used model by, by Koenig Walzhe that's in the Eurocode, that's also comes from back calculation. So they kind of had loaded timber studs in a furnace. They measured the temperature, and then they saw when failure happens and based on the temperature and the they back calculated, okay, I need to apply this to this member to get this, and that's how they kind of

Wojciech Wegrzynski: 45:25

The practical consequence of that is understanding how we better can translate those items of into load-bearing considerations of Like, or, or we should just accept that moisture into timber, let's just, uh, have a better and, zero strength layer and we're done. Like, Uh, where, where is it going, Felix?

Felix Wiesner: 45:47

Uh, I mean, we can-- if we want to make our life easy, we can say, "Okay, fire resistance, that's the gold has been there for a hundred years. We're kind of happy with it Very few buildings collapse, very few buildings collapse in a time where we have very few tall timber buildings." and now we just, we just built these massive timber buildings, and we use the of fire resistance even though we know, like we've for the last thirty minutes or more, that there may be issues with that, no? Like, the translation of fire resistance to real fire uh, and now also that, okay, does that even work in a phase? uh but now we can say, "Okay, we-we are engineers, we to do it better," or, "We have a concern, we want to do performance-based design." All of those are reasons us to ask, "Okay, how do we do it better?" Then we need that link between, strength and stiffness and And we have some links, but now we're asking like, too onerous? Do they still work in the decay phase? Because they have only been derived, either from, small-scale tests, often from steady state, um, which I think are not onerous enough to really, like, show how the material behaves in a real fire. Or we have them from back calculations, but then only standard fire. So we don't really know what happens when it cools. Like how much strength is coming back, how much of the stiffness. Yeah. So now we need to kind of see, okay, what happens if heat the timber and we cool it? What happens then? And we don't really know. No one has really done much. I mean, there's a lot of work in the past that kind of it. Uh, and that's why I also wanted to look at like, okay, how does timber actually fail at the, at the, micro and nano scale? and then again, the paper was limited to sixteen and seventeen pages, so I couldn't really look at all of the nano scales. So I really focused on what's happening to the cellulose because that's the main co-component of kind of the which make up the cell wall. and in there you have cellulose linked with hemicellulose and lignin that kind of keeps it all together.

Wojciech Wegrzynski: 47:49

are we even capable to study that in elevated at this scale? Because it sounds like you would need some very uh, scanning electron microscope or something to, to really study, uh, perhaps even more if

Felix Wiesner: 48:05

Yeah. I mean, obviously, I mean, cellulose is in itself a super interesting and complex material. I mean, through like looking into this and talking to my colleagues, I have some colleagues on my floor, they, work a lot with cellulose and they turn it into, it into adhesives, they turn it into medicine or yeah, they do all sorts of things of, with like they manipulate it to, to do new materials. so there's loads going on. There's, there's multiple states that cellulose can, be in. In wood luckily we only have one state, but it's still quite clear what that looks like. So there's theories, 'cause one- once we go to the um, we can't really look at it in too much detail, no? If we're at the microscale, we c- we can kind of see cell walls, um, super nice pictures of wood that's sliced open. Um, but if we go deeper it gets harder and yeah, and obviously doing heat testing, we always have to imply no? We can heat test it, but we can't take like a strand of cellulose and, and, and test it. Um, although cellulose in itself is super strong. Yeah. But it's not about getting a definite answer here, it's about finding out what actually happens. And one thing that through looking at it is kind of, I, I went down and I found this paper that says like, does wood fail on stress?" It's like, "Oh, That's That's just what I'm looking for." Uh, and one exciting opportunity about, uh, this talk and this paper for me that I could actually do research. Now you, you know that you obviously, as you, you go a bit in the academic world, you, you, you turn more into a research manager and you, you, you work with your and they do, do all the grunt work, and then you help the results and you don't have much time to, to, go into too much detail. And so I could actually go down to the library and at those papers, um, because you can't find them online. Um, luckily we had some copies at UBC in one of the and, and it's super interesting because they kind of timber as this perfect composite, where all these layers and aspects work together and they act as a per- perfect composite when they all break together, maximizing the strength of each component. But then they said like, hey, for example, when it's um, the membrane between the cell walls gets a bit so then you get individual fibers being pulled out others break.

Wojciech Wegrzynski: 50:23

So, uh, the future, what's the next biggest question that you are looking in? W-w-what's the next step? Going smaller? Going larger?

Felix Wiesner: 50:35

I mean, I wouldn't go go smaller. I mean, I, we have a, we have a, pulp and paper institute. They also work a lot with that. They also need to know how strong, uh, wood chips stay uh, like they take out some of the lignin, and then need the cellulose to be strong. It's kind of how paper's made. And they have a cool machine where they can kind of the compression of the microfibrils, and they actually like, "Hey, we have this temperature module for it that we have never used." And I'm like, "Oh, wow." I mean, uh, let me have a go at that. So, so that might be interesting. Um, in terms of like low-hanging fruit, something we should look at is what happens to the recovery of and elastic modulus after heating while it's still Yeah Loads of people have looked at residual strength, usually the way that looks like is they burn the timber, then they take It out, get off the char, and then it. Okay. That's obviously it has some interest, but it's not relevant to what happens in a real fire because in a fire, that timber is loaded while heated. So burning it while it's unloaded and then testing the afterwards doesn't really tell you how it affects it. Because if you look at it, one of the things, again, down to moisture, and this is where we get to the nano is that if you look at how the cellulose molecules, uh, or not the mole- the cellulose chains are connected both to the hemicellulose and internally, um, if you add it breaks some of the hydrogen bonds. if, if these are like my top hand... Well, obviously this is, this is a podcast so, but you have, you, you have two cellulose chains, um, on top of each other, and it's unloaded, and I remove the hydrogen bonds between them, and then it dries and hydrogen bonds reform. Nothing happens. If that's loaded in somehow and I remove those hydrogen bonds, they will move relative to each other so that will deform and so will its, its capacity to, resist Yeah? So you're changing the structure, the nanostructure And then when I dry it while the load is still on, that kind of deformed state remains in there. So whether you have load or not is super critical to the outcome. So if you do unloaded tests, it gets very difficult to say like, "Okay, this is what will happen in a fire." And what I've seen now often is that people do compartment tests unloaded, and afterwards they measure the charring depth and they say like, "Oh yeah, this charring depth actually less than what we would get in 120-minute fire in the furnace, so therefore my fire resistance must be as good or better as 120 minutes. Job done." But obviously that doesn't tell you happened if I had loaded a compartment, and also it what happens beyond the charr depth With heated timber, moisture migration. Or also, like you said, a thick piece of timber, the moves, uh, that Well, It will move on the Like, let's say you have a column that's heated from all sides, the some of the moisture will obviously evaporate, some will go inside, but now the timber, let's say the fire dies so now a previously heated section cools down, so now moisture can move back. So now you have timber that's being heated, dried, timber's being wetted, wetted timber dries, later tim- uh, dry timber is wetted again, all of it whilst load. Yeah? Plenty of opportunity for those cellulose chains to move, which will affect your outcome. So I'm not saying as engineers we go down and, and model our, our timber beam to at the nano scale. Um, I mean, if you have a huge supercomputer, be my but yeah, I know there's research out there that looked at kind of modeling timber at the, at least at micro scale, like very fine detail, but obviously to find out what's happening, and that's super but we're not gonna do that in engineering. But we need to know what's happening at the nano and scale to inform what's relevant for our engineering

Wojciech Wegrzynski: 54:36

Well, you know, as I, as I said at the beginning of the, of the interview, my limited knowledge on, structural fire engineering of timber, what, what I learned is that it's, it's quite variable. There's a lot of randomness and anisotropy into you know? And you can just live with the fact that there's lot of randomness, or you can go into, you chains of, uh, of cellulose and try to figure out, like, what are the conditions at which those effects can happen, which perhaps like if you a ultra-fine supercomputer and you were-- you be able to model a beam loaded up to a level of single cellulose string in the beam, uh, would get it, like why or why there is a

Felix Wiesner: 55:22

Yeah, but I think there are two different things here. Randomness, we will always get that. Now, even if you have the perfect model, your actual beam may have slightly different orientation. We can deal with that, safety factors, distributions, probabilistic approaches. But if we have a fundamental understanding wrong happens and that informs our model, that's a bigger So that's why I think we should try to consider the scales.

Wojciech Wegrzynski: 55:50

Perfect.

Felix Wiesner: 55:51

And sorry, just one more thing I wanted to say. I mean, you say, like, you don't have a good idea, still understand the concept of structural fire safety There's loads of people out there that have way less an idea of you, and they actually design those buildings based on those empirical and more simple models talked about earlier. So, and, and that's kind of the issue that, um, we're groundbreaking buildings and we're kind of saying "Okay, we use this proxy of a proxy, but we're not sure how it, uh, translates to a real fire." And kind of where the work is, and there's loads of fruits there, I think, for further research, which I as researchers we're always interested in, aren't we? Always like, "Okay, what's the thing we want to, to find

Wojciech Wegrzynski: 56:35

I, I, I'm working way too much with structural I know they would kill me if I claimed any, uh, in this space. So I'm just, you know, uh, a tourist entering world and enjoying the views. But, uh, I, I do enjoy them and I find them very, very interesting. I find them, uh, rich in, in the amount of that's in there and, and especially the way that you carry your research looking across the scales finding answers at different levels. And, uh, you know, uh, I, I always say, uh, i- if you consider anything in fire and you ask why at four to five times, you're gonna get into interesting place. Uh, and I, I think, uh, this is, uh, where you with the, with the timber and I hope you will-- you still have a lot of whys to ask to, to get into

Felix Wiesner: 57:28

Yeah, I, wouldn't even say I found any answers. It was just for me, I wanted to go down and, and ask those whys. But, um, as usual, with any good research project, you

Wojciech Wegrzynski: 57:43

there are just questions. Like we should teach that to students first year,

Felix Wiesner: 57:49

Yeah. No, no, no. With students, you don't want them to give too many

Wojciech Wegrzynski: 57:57

Exactly. Perhaps, yes. Yeah. Okay. Uh, Felix, looking forward to yours-- and congratulations on the Prue Award I'm looking forward to have a beer with you in

Felix Wiesner: 58:10

All right. Yeah. Um, looking forward to see you there. Thank you very much, Wojciech. day.

Wojciech Wegrzynski: 58:15

And that's it. Thank you for listening. I am actually in La Rochelle, and I've already Felix and he looks great. And I think he's getting ready for the uh, big day when he will receive the award, uh, his, uh, acceptance, uh, lecture, and, share with the participants of the IFSS conference You got it first at the Fire Science Show in I would say, an expanded version, which makes me happy. Um, So this talk summarizes years of work and by Felix. Um, he works with a lot of people out there in space, and now we're being in a-- in University of British Columbia, where there seems to be a interest in, in wood in particular. I think he's in a great position to actually expand the fire view into, you know, more mainstream science and, you know, interact with those more to, to, to get more. I'm excited to hear that he has access to new and stuff that will allow him to look even deeper into the, the s- the scales of the problem. That's, that's very, very exciting. I think in here we've covered a lot of scales, or Felix covered a lot of scales from fire resistance to cellulose strings. That's a hell of a range. Um, there's a paper that follows this, or this follows a paper which is already available at the Fire Safety Journal. The link is in the show notes. I think if you are interested more after the you should read the paper because that's gonna everything that Felix said by a factor of two, because everything is there even in more Is-- There's even more content out there, and I highly, highly recommend that. And, uh, for me, I'm going back to the enjoying, uh, La Rochelle IFSS and, uh, capturing more guests in here, recording interviews, enjoying fire science, learning new stuff, and, uh, talking to everyone around. If you see me there, well, let's, uh, let's chat uh, and, uh, have a good time. Thanks for being here with me at the Fire Show this week and for the next week. I have a really, really good one coming. So all, all are good, but this one is particularly good. You should be looking forward for that. See you there next Wednesday. Bye.