050 - Wind, fire and a surprise

Welcome to episode 50 of fire science show. My name is Guillermo Rein and I am taking over the podcast today. This is an important episode for three reasons. One is because it's number 50 it's a number to celebrate. It's a long series of episodes. It's really good. Second is because Wojciech is not going to interview anyone today. That's unusual in a podcast that he created, that he has been doing every week interviewing people is because be. Who is going to interview someone. And the third surprise is that I am going to interview Wojciech. So he is going to be talking to you, but instead of, extracting information from others and making them talk, it is going to be the other way around. He is going to be talking to us. He's going to be sharing his lessons, his experiences, and his messages. So this is, Uh, an episode that is important in a podcast, a series that is important. Wojciech tells me that he's super happy to know that about 25,000 downloads of the episodes have happened on average, 500 downloads per episode. He's really, really happy about this and keeps growing. And I just wish him the best and that this keeps in this direction and hopefully the interview of the creator um, helps to cement farther the growth of the series. So Wojciech is going to talk us about a specific topic. He is is, as you will listen to in the interview, he's a Renaissance person. he, he does everything. And when he does something, he does it really well. Like I said, um, so all the topics I decided to ask him about one thing that he's very well known because he got awards and really good papers, which is in the modeling computational modeling of fire and wind. The two of them have massive interactions as well. You will tell you. And when they interact, they make the problem of modeling more interesting and more challenging. He is one of the messages he has for the community, especially the younger ones is that people should not be put off by the complexity of combining wind and fire that they should endure. That difficulty is challenging. He's telling us, be aware. He's telling you, he knows that he's telling you, but it's also rewarding and it is important. And as an engineer, he himself can tell us about. Solutions to safety that he has found by not shying in a way from wind, but by embracing when and couple it with, with fire. So there is a lot to learn from Wojciech the scientists and Wojciech the engineer. and just, just assess that. And then I will just, let you listen to what he has to say. So Wojciech I I'm delighted to be here with you. I'm delighted that you are finally being interviewed. You've interviewed so many people, 49 people, I think, or 48. you have an incredible good taste in choosing topics and people that's. I think part of the massive success of the podcast, but the most important thing is your ability to have amazing conversations with multiple different personalities. And that says a lot about you. We adore you. We need you, and we, we want your show to keep going for a long time. Um, so I'm very happy to give a grain of sand and to be the one interviewing you today,

Yeah. Thank you. Thank you so much, man. It's so hard. I mean, this is the same seat, but it feels much different. I'm very happy and thanks for all the kind words. And I'm happy to find finally be able to share some of my own research at the ITB that, that makes me very happy. So what did you pick.

Excellent. So you could be interviewed about many different topics. You are a Renaissance, man. You, you have multiple interests, sewing the podcast, but you have multiple contributions in different topics. I think that definitely is going to be modeling, although you are very good with experiments. And within modeling, obviously fire, because it's a topic off of the podcast, but you, you have these really interesting and early, contribution on fire and wind modeling the two of them. So it's computational fluid mechanics, not just a fire, but of wind, which is an interface. Right. So can you tell us a little bit of how you end up working on this

yeah. Thanks man. Thanks. Thanks for picking this. It's it's very near to my heart. it started somewhere in like 2012, 13. I was not really a scientist back then. I was a consultant and we were doing modeling for four buildings and, we ended up having this, project in the middle of the Warsaw, uh, Hala Koszyki Center we we've celebrated the first, uh Obora in, in that particular building. , And it was an old, historical, uh, shopping mall inside of the town. That was to be refurbished. And because it was a historical mall, to remove the smoke from it, the only choice you had were rows of windows that were underneath the roof, that was the only, the only way you could extract smoke from there.

I remember seeing those.

Yeah. And because, uh, these windows are known for being very inefficient for smoke removal in wind conditions. , my colleagues Grzegorz Krajewski and Grzegorz Sztarbala have started doing modeling for the building to establish. How efficient the ventilation is in it, you know? And, Grzegorz Sztarbala was doing his PhD on, um, the wind influence on tunnels. It was very interesting. So they've applied a lot of knowledge. They got so far, they ended up modeling, , a big part of the city. This is where I came into the project because I helped them build the models and run them. And it was like so interesting, such an interesting case study where you'd actually modeled the wins to truly , answer the question, will it interfere with the smoke control or not? And I think that it was very surprising to me to find, such a rich interaction inside was not a simple, yes, no, it doesn't interfere. It does not. You know, you could see how the flows from different streets would create a completely different, performance of the system. Just fascinating. And we've published many studies about that building, that are available online. I think it was a very nice case study. And, since then I started playing with it. You know, I've built some models. I I've made some buildings, started pushing, wind on them, rotating them, seeing what happens. And the answer was never, never simple. And then fast forward few years, and we go into some conferences, we talk with people about the projects and because we had no idea how to deal with this wind projects, we had to use best practices of wind engineering. And for us, it was kind of obvious you did like. And then you meet fire people and sorry for calling you guys fight people, but then you meet fire people and they drink it completely other way in, in what seems to when engineers, there's a wrong way and you're getting into discussions and you realize that it's not that simple. It's not, such an easy thing to actually perform these studies together in, in a way that is scientifically justified, that you control the uncertainty that you actually model wind and, this is when we've decided that maybe because there as there are no good practice guidelines, I thought it might be a good idea to write ones and yeah, two, three years, but it was worth it it's. So ended up in, in fire technology has ended up being an, a quite well-received paper , and that, that was the entrance to the beautiful world of wind and fire.

thank you for mentioning the paper. I have it in front of me is a two-part paper is, is a absolutely beautiful paper. I just really, I read it again for the interview on, and again, I, I was like, oh wow. It is as good as I remember it even better. It's part two. Right? So it has part one and part two. And actually you got a best paper award.

Yeah. Thanks. Thank you.

do you want to tell us a little bit about that experience?

Oh, it I was so surprised. I, I mean, as a young scientist, you know, there are many of these awards where you, battle with other young scientists and it's, it's fun. Uh, it's fun to get these awards, but a Bigglestone award, you battle with, dinosaurs at the best of the best. So I was never thinking of that. This contribution could actually become the best paper of the year. And, it was fantastic. It opened the whole. World of possibilities for us, because we suddenly could say that the work is important to the community that it is, received, by the community. And we had to that it is important, you know? So for a young scientist, thing like that, it's it's life-changing man, I think it's, uh, like you mentioned, the ERC was life-changing for you, maybe not on the same scale, but you know, that, that, that's, that's a similar thing. it's truly changes, not who you are because we're doing the same things, with the same scrutiny, same tools, , but suddenly many, many more doors open than there were open before. So I really appreciate that, that total board of five technology for choosing. These, this paper as the best paper of that year. And truly there was so much work put into that, that paper. And it was a little long ish. Sorry for that.

Well, that's, that's why it's two parts right now, but congratulations, I think this is a success story. you may actually want to write suddenly you could offer a talk, to young academics and young researchers about the success story of a topic that you push forward. A topic that probably at the beginning, people were saying, it's not needed. You don't have to write about this book and then you did it. And then you end up with a beautiful paper and an award and an interview in a famous book test. So part one of this paper is a literature review and I really liked. I literally, you literally scan the horizon. I realized if you go high altitude and then you start going lower altitude into any details, can you tell us a little bit about the structure of why do you choose this? You have six, parts in the literature review. Can you tell

we, we go all the way from, um, modeling, the flows inside the buildings, we're modeling the ventilators and the let's say the interference phase between the building and the exterior we go into tunnels, uh, because that's an interesting twist in the, in the problem. And in the end, we, we move, outside the building into wildfires and, uh, urban interface areas. And then just, dispersion of the smoke in the urban scale, because wind is a phenomenon that, interferes with fire at every scale. And I thought, I always, you know, I come from the ventilator business. So, so my interface is the ventilator. the smoke removal devices in the buildings, I got the dampers and everything you used to remove the smoke. This is my, my core business or use. Now it's wider, but for us, wind is just a resistance. You know, the, the damper, subject to event has worse performance than when it's not subject to wins. And after my studies starting with the Hala Koszyki and then other studies, I immediately realized this is bullshit. This doesn't make sense. Okay. If you just consider a single ventilator, it makes sense the wind is resistance, but when you place that vent on. And the roof has beautiful pressure pattern on it. Some part is negative. Some part is positive. It's so complex. When you put the winds on the building facade where it can push air, or it can block air from coming out of the building, this ventilator will work in a completely different way and it will give completely different barometers of its performance. So narrowing down to the single thing being a vent made, we understood the problem completely wrong. We understood it just for the vent, not for the fire safety and the same goes for a building. If you just model the building in isolation, standing there in an empty space, or just put some inlets around it, say it has this dynamic pressure. This is the wind that acts on it. What you do is you write. Essentially exposing the building to a certain value of pressure, but not wind when there's a complex phenomenon. I always joke that, you know, imagine if we took the complex phenomenon as a fire and just replaced it with a simple time temperature relation, how stupid would that be? And that's what we, as a fire scientist tried to do with, uh, with winds we don't, appreciate the complexity of the phenomenon and with this complexity comes diversity of outcomes of, the wind ethics. So this is why we thought in the review paper, we can not just tell people the story about the buildings. We cannot just tell the story about, um, the ventilators. We need to try and find out what. The fire community at whole has learned about wind. And, and there were like, there were reviews by Pitts from NIST about a war fires, which was amazing. It was fantastic resource on wind and fire interaction. On the city scale, we found brilliant papers about firebrands aerodynamics, they're so rich in knowledge about when we went into fire whirls that's this, this is another fascinating phenomenon. So we thought, okay, if the, if we want to give it the justice, if we want this. The fire community to be exposed to how wind is done and how one should include wind in their analysis. We need to give a broad view on the subject because it is much more rich than most people would think. And I hope that I can do it in like three pages, but no, it was a home, a whole 80,000 character paper about, about this. And we still have not probably captured all, all the knowledge that was there.

Thank you for talking about all the ventilators and the buildings. you touched many beautiful problems. One of them that I invite you to say more about now, if you want is tunnels, you look into when and tunnels and fire.

Oh, yeah, we just recently had, commissioned the tunnel in Warsaw, very big road tunnel. we've spent almost half year that in the tunnel. And one of the main question was how does wind affect, the flows inside the tunnel, you know? And, it sounds like a simple question, but, to answer that you need to answer, what do you mean by winds? You know, that that's a fundamental question. What is wind like? A wind has a velocity. it has gusts because it's not just a constant flow of air. It's, uh, it's huge vortices that move around your, region you sometimes have high velocity, somethings have flow velocities. So it's very complicated, but first you have velocity, so there's a certain probability associated with different velocities. Like, should we design a tunnel for a hurricane? Which don't happen in Poland? No, it doesn't happen. So you need to figure out for what wind velocity would you like to design your tunnel? And the action of each of these velocities will be different, but two to make it even more complicated when has the direction component to it?

when you, one thing, when you, you mentioned wind and tunnels, you really mean the wind outside the tunnel and how it alters the inside. But, but the inside itself is not wind. Right? That's what, that's

Inside you have some flow that is induced by the wind, but the wind action outside, it creates this, some of the, of the pressure, the dynamic pressure of the wind, creates the flow

Yeah, but yeah, because it changes the boundary

exactly it.

And, and this is something is, is fascinating because not many people actually look into this many people look into the modeling of the flows inside the tunnel, but very few people look into the effect of a flow mechanics

and I tell you, this is the issue we've, um, we also felt into, because we've designed the system in 2015, for this tunnel. And we went with the best knowledge of the time, and we didn't really consider that much as a factor or not as such a dynamic factor as, as now, um, in this tunnel we be that transversal ventilation system. We really wanted to have low velocity in the tunnel around 1.5 meters per second. That was our design goal. And the wind could create a flow of four or five meters in the tunnel easily on its own. We had array of jet funds in the tunnel that could be used to stop the winds, but if we start all of them together and there is no wind, then the Jetsons will create this overwhelming force. So we needed to find a balance between the wind and the jet fans in the tunnel. And that led to design of 688 different scenarios of operation of the tunnel for different velocities of the wind outside and different wind direction angles. And only then we could say, okay, no matter what wind is outside the system, if it triggers in the correct scenario, it will stop the winds to do more or less the level we needed. And so the, this. Probably the most difficult wind fire interaction I've done in my career. And this was not modeling. We had the tunnel, we had the measurements, we had the weather station. We could I'll tell you a story. I was, it was like Saturday.

00 PM. The children are going to sleep and I am dressing up, packing up. My wife is asking me, well, where are you going? And I'm like, I need to go to the tunnel, but why I need to measure the velocity in the tunnel. Cannot someone else do it. No I've called them. No one is there. I need to go there to measure, but, but why do you need, there is very rare winds occurring now. And my wife is like, w w what? what wind is occuring?. I'm sorry. It was like, because we had very, very rare southern wind with a velocity above six beats per second. And I really needed to go to the tunnel to get this data point because I missed it. I, we never had that when we really missed it. And I, I had, I packed my anemometer. I run into the tunnel. I've measured. the thing the guys from security were looking at me like, what's he doing here in the, in nights? But I took the point and it really helped me solve the issue because we've found the missing points that connected, many other together. And we finally had this full, full circle and, and the full understanding comprehension of how winds will affect the tunnel. But yeah, when we imagine a tunnel, that's a pipe, it has to open. And we ended up with almost 700 different scenarios in each of them is completely different. And now imagine the building a skyscraper, which has endless amount of openings, the tunnel is simple. It's in the ground. The skyscraper can be 500 meters up in the troposphere escaping, escaping the lowest, atmospheric boundary layer, how complicated it is then, and in fire, how we deal with problems. We don't understand, we omit them. That that was not a choice we thought is, is viable. we wanted to understand it. We're still, still not there, but, uh, on the route and, and tunnels are a great example , of what you can learn by touching winds, especially if it's a rare wins.

you you're making a very good case for the importance of wind that sometimes it can help in this smoke movement and in the fire movement. And sometimes he can not. you're already alluding to this, but, so what do you propose as, engineer, you're near that obviously you are not proposing that is ignore. As most cases have happened to date. Uh, you, are you proposing to focus on the worst case scenario or proposing a probabilistic approach or, I mean, you you're mentioning multiple scenarios. So tell us more about what you propose of, how an engineer can deal with this scientific, ability to study the interactions of wid smoke and

Yeah, this is a very good question and a very difficult one to answer. obviously as, you mentioned, is, is complicated. There are hundreds of scenarios. I mean, even probabilistic. Too expensive or most cases. So it's very difficult to like capture this phenomenon together completely. we are now running a, a research grant, funded by Polish NCN agency where we're somewhere around the middle of it. And the goal of this grant is to solve that completely for a single building in a single location of the town. We choose an open car park because this is a very interesting case study where the, where the wind can truly have the tremendous impact over the course of the fire. And, it will require from what we. Established at the moment, somewhere between 1000 to maybe 4,000 different scenarios to completely have like a complete view on, on different fires and winds how they interact and what are the outcomes. And this is only once to be done. Like we, we really want to understand the complete image, but, from previous case studies, we had case studies in which we did, uh, let's say two meters per second, five meters per second, 10 meters per second, average wind velocity. And in this case studies, we found the middle one to give the worst results. Like the strongest winds would actually give better results than those lowest one. And that was really. Surprising, you know, because suddenly, okay, so we don't have the worst wind. Like this is uncomfortable for me. I always thought the stronger, the worst, just pick the, the strongest reasonable wind you have. And you're good, but no, this case studies shown it's not the case. It's, it's, it's complicated. so essentially enter a problem that that's potentially not possible to solve, but when the writing this review papers that were mentioned, we also figured out that, okay, let's go to building. when does the wind affect the building a fire when it becomes its boundary condition? Like you mentioned the wind becoming a boundary of the portals. So. Maybe we can have a cheaper way to establish in which wind scenario it is likely to be the worst boundary condition. And then once we determine, this scenarios, you go into full wind and fire simulations, which are expensive because if you simulate just wind, no fire, you can run it as a steady state problem. It is not as computationally expensive. I can do a 20, 30, 50 simulations of just wind at the time. I will do one wind and fire study. So you can actually do hundreds of wind scenarios of your building that you're interested in and determine, in which scenarios you have obtained, the biggest pressure rise on the inlets or outlets when there was the highest under pressure on your events, you know, just investigate the pressure on the building. In which scenarios you had the most extreme values of this, of these pressures. And they usually would correspond to the scenarios in which the wind fire interaction would be the strongest in your fire case. especially when you have, for example, over pressure on your exhaust vents. This is when the wind will block the smoke inside the building and will not allow it to escape. So this is definitely the worst case. And by this, you can narrow down from hundreds of scenarios to be investigated into a number that is manageable by the engineer, and then focus on, on this particular snow. This is how we solve a Hala Koszyki problem. We narrowed down into the ones that are most likely to, be a problematic and then solve for them and see what happens. And I think at this stage, this is probably the best, the best you can do. It's not easy. It is not needed for all projects, for sure. but, when you need an answer, it's the best way you can find one.

Perfect. Thank you Wojciech and you have already alluded to wildfires, but because what if ours is such an, an up-and-coming fire, phenomenon, uh, w w where there's a lot of needs for. And you yourself have interview Michael Gollner and Sarah McAllister and Cathejline Stoof. you in your 2018 paper, this paper that we're referring to, you actually went and review the literature of wildfires. And when modeling, you want to tell us about this, that was one of the parts that I enjoy the

Really? It makes me happy. Uh,

you have amazingness sketch you're really good with the sketches and drawings and you have a particularly good

Yeah. I mean, it was a four for us. It was probably the most difficult one to write you know, because, , we're not a wildfire experts. And, uh, we were worried that we may omit something critical for a wildfire community. And in this particular, chapter w we got help from Mike Gollner and Ali Tohidi who were kind enough to refer us to some, some studies, about when the fire that has been done in that community. And, um, I think it is even if you go back to Rothera mills model, when there's a factory in there,

one of the three most

Yeah, exactly. I, and you don't have to be a, a fire scientist to understand that the wind influences wildfires and how they spread. So definitely, , It was a parallel field to the building fire science that had its own take in fires. however, I think the problem is slightly different than buildings, in, in wildfires. I think, the wind you could, if you could oversimplify the case, like the wildfire case, the wildfire is like a boundary layer problem. You know, that there, the, forest will be essentially a roughness boundary condition on your ground. it will influence how the flow is changing in this boundary layer, but in general, it will not change the structure of the wind so much where in the, in the city, this is something that was the most scary finding of our work in cities. Concept doesn't really work that well, you cannot put an isolated building and just, mimic the city with the roughness criteria. And then just pretend you're doing winter because the array of, of streets in the city will create a completely individual flow path around your building. In that case study that I've mentioned just before, when we found that the middle velocity was the worst one, we also found that, the building was a rectangular building and it had an opening on the, on the shorter wall. So we thought, okay, if we blow directly to the door, it will be definitely the worst case. And it was not the worst case was when the wind was like 45 degrees to the right. And the wind was like hitting a building nearby, running around the building and hitting our door. That was the worst case because of the complexity of the, of the, of the system in there. Here, even though we're, we're talking about wind, this is the same thing. I think the problems we face in the wildfire community faces is completely different in, in terms of, what is needed and in the wildfire community. I think, what's the most beautiful thing is they truly do forecasting. You know, they, they work, to forecast how fires will grow. Uh, where will they go? Where will the smoke go? how will it spread? And. They do something we never do in buildings. They measure the wind and try to incorporate that measurement in their forecasts, in the reduced order models. And for me, this was absolutely mind blowing how beautiful these solutions can be, in our case, you have to make up the winds. You know, you have to have this logarithmic wind profile. You have to make up completely based on science obviously, but you make it up and here they deal with something they truly measure. And it's something that you can like the closer you are to the forecast, the more, you know, the more accurate it should be.

You mean that because wildfire has happened on very large timescales? No, they will be spreading for days. You are saying that they can measure the wind profile in real time and then do a

They can, they can incorporate the momentary metrological data in their analysis to refine their, forecasts. And for buildings, you don't have a building model readily available when the fire is ongoing not to mention metrological data. So I think only for the biggest, most difficult fires, you will go into any, any sort of modeling of a fire during a fire. Uh, while in wildfires, I think this. A simpler model because it's not only CFD can use Eulerian and models. You can use, uh, Lagrangian models. There is, um, I mean, even simple Goshen dispersion or puff models are, are being used, to forecast. Uh, I actually was very happy to write a chapter four on that for Brian Meacham's handboook and Margaret McNamee's is a handbook that's coming up soon. Ish. I hope, uh, where, where we go in depth in, how we can model this, um, wind and smoke interaction. And, I think in, in welfare committee, the wind is maybe even more essential than it is for us. How are I think the solution that may be even maybe a little easier than, than for us, because for us this, Specific flow paths of, city wind is, is something that can drive your fire behavior. And that that's unfortunate because it makes the problem so much more complex than, than just, uh, you know, a simple boundary condition

Cool. Think of Wojciech you are. You're such a Renaissance man. Confirm, um, also in the interview,

in the render sense you would die for forecasting weather. So that's not great. That's not great, you know?

following the structure of your paper in part. you go be joined the literature review and you're going to something which is very nice and very rarely done in science. You actually provide our best practice guidance. this is amazing and I encourage all engineers in the field to take a look at no a leader in the topic is actually saying what he has learned and he sharing, the, lessons and the hints. So in that, in part two, there are you, you literally, we could be talking about this for days, but let, let's try to make it just the podcast episode. So for example, you talk about a topic that is really important in modeling and CFD is there this discretisation and do you want to tell us about the importance of discretisation in the specifics of when you are doing fires smoke?

Yeah, this was a huge challenge for us and, uh, my co author Tomek Lipecki, who, who, uh, uh, actually, uh, should have mentioned him in the start, because this is, this was a two man or at the, at some three men work with, with . But Tomek I guess, has done immense work on, on, on this review along me. And it was a huge pleasure to work with him on this. And we're still working together on wind and fire. So, so high five to, to Tomek um, Thomas was a researcher who did not do any fire is a wind engineer. but he did a lot of work on scaffoldings, around building. And funnily in the issues in discretisation of space, when you model scaffoldings and model fire are seemingly the same because you have this tiny thing that is inside your building or around your building for him, it's a scaffolding for me is the scale at which the fire phenomenon happened. So it's, it's in the centimeters, let's say, and you have wind and you can really do a centimeter mesh. When you were modeling two by two kilometer, large domain in wind engineering, you would go to 15, 25 meters cells. That's not very odd that you go into such a cell dimensions. Whereas in fire you would like to use relatively small meshes. We tend to use 10, 5, 20 centimeters, based on the star, which is not the measure of mesh quality, by the way. That's I always feel horrible when I see people justify their choice by that. But yeah. Regardless, we use meshes incentive meters, whereas in fires you would use meters

you mean the same to me, the same fire where we meet.

yeah. Centimeters in fire. And whereas it would be meters in wind and you know, this tiny detail, I mean, what is the consequence of that? Well, first of all, in the wind, you would run. steady state simulations, like I mentioned before, or some sort of periodic simulation. So you would measure like 20 seconds of the flow and you assume that it periodically repeats. So this is how you would model wind in fire. We, in, in practical engineering, we model ASET/RSET. We need to know the fire development. You know, you need 20 minutes of the fire, what will happen, how it will change over time. So suddenly you're talking about a completely different simulation. That's why I mentioned I can do 50 wind simulations when I can do one wind fire simulation, because for the reason I need to model my fire, on a transient scale, uh, and now the size of your mesh, if you use FDS, that does not only dictate your, spatial resolution. It dictates your time step because of Courant-Friedrichs-Levy criterion FDS is used in MDs. so changing the mesh, you change the timestamp of a simulation. You're adding iterations, do simulation by having, um, a finer mesh. So. quickly becomes a very complicated matter that dictates if you're able to solve it or not. You know, uh, because often when people will receive a CFD that will last a week or longer, they just give up, they rarely pursue such simulations. And I hate it when they see a paper, we had to use a really horrible mesh because my laptop is not strong enough. This is not the, this is not how science is done. Like, you know, imagine certain set. Uh, yeah, we think this particle exists, but we cannot confirm because the accelerator is, is just, uh, bad and we cannot afford the better one. It's not that that that's a restraint. You have to work around, but not, something you can justify a choice with. Uh, so, so the choice of fish is really horrible. And I think in this. I'm highly inspired by Bert Blocken he's a leading scientist in the field, the field of wind engineering and field of, sports aerodynamics. And Bert um, he, he was a pioneer who were doing this best practice guidelines for a wind. And this is what inspired us. And he showed very practical solutions to problems. You know, they've run case studies. They said, okay, this mesh is sufficient. This is not, this is how you should build it. This is the most efficient way you can do it. And we wanted to transfer it into, into fire. The thing that I think is. Quite, surprising for fire people for us is that it's not sufficient to just make the mesh. You have to measure the quality, and this is something I'm missing in many modeling studies. Like people just choose and assume it's going to work. But in wind you have to measure you. You have to provide some metrics that this mesh is good. Not on like the best is if you go convergence study and see, um, how a change in the mesh affects your outcomes. But that in the fire is sometimes very difficult, to obtain because uh, many solvers do not really convert with, but like better mesh does not necessarily mean a better solution in fire. Yeah. And, and they have these metrics for like two blends, Y plus, uh, the scales of analysis. I found it really interesting that, such metrics can be used and I wanted to expose the fire community. White guy seeds, like the mesh meshes, like, uh, it's an important thing. let's mesh it up. Let's check it out before we go into modeling and there they're spending more of these best practice things in, in the paper, for sure.

yeah, that's key. You, you are that you're describing the problem. Rightly so as a multi-scale problem. You have two phenomenon of importance, the fire, the wind, and each of them have different temporal and a specialist scales. And as a modeler, you need to do both. You cannot choose one or the other, you have to do both. You're absolutely right. I really like it. How you emphasize how many people drop the wind when they try to do this and then not just going fire, maybe not for the best reasons, just because of the difficulty of the task, which means that the people who do it, like you must be celebrated and highlighted because this is the way

And my, you know, when we went into this OPOs grant, when we submitted that we want to study wind and fire interaction, we didn't just go that we want to see if a carpark will burn worse than, uh, when there's wind or when there's no wind. We wanted to understand the consequences of low to meet, velocity wins on, on the fire, on smoke, to be able to, you know, give this better informed engineering judgment. When we go into something like field experiments, you know, Often see, in papers from Windex, from fire experiments that a wind was moderate up to like five meters per second. So we assumed it has no effect on, on fire worse. We didn't know. We have no idea if it had there not no one has ever done it in a parametric study. So we have no idea. What is the consequence of low velocity winds and from the initial work we've done, we've, we've seen these consequences. Like we've seen that, one or two meters per second wind can already change things a lot. And that's disturbing because that means everyone can, like, it's very rare that you would have wind conditions that would not affect you at all. Now, when you do fire experiments in a laboratory, if you take my laboratory and you do the experiment With, a wooden building that I'm currently burning, like the ninth time. If you show me the picture of that, of that fire plume, I can tell you which doors were open, you know, based on the shape of the plume, this is how often fragile fire is to external conditions. And, when this overwhelmingly there, it's always there. It is not the condition that appears and disappears. It's just there in the background. And if we want to understand, what are the consequences of it being there? And I don't mean that it invalidates the data or something. I just mean you need to know the uncertainty. I had Francesco here last week and, uh, you seem to terrorize him about uncertainties and he's really good at them now, but this is important, you know? No, no, you're your set and this and that, that gives you a better way to, to understand what you are.

Now thank you. You, you are picturing a truly fascinating problem. Multi-scale complex problem where the worst case scenario is far from evidence, which I think is very interesting because obviously in engineering, focusing on the worst case scenario is a tradition of itself, not the only one, but when the cannot be found, that is a hard, a good justification to invest into when modeling them. Excellent. So from the following on the multi-scale and the complexities of combining fire and wind, and one thing that wind and fire having common one, then definitely they do is that they'll never laminar. The two of them are always turbulent. So good. So you are definitely going to be including turbulence in your model, but which model of Trulance and which model of combustion

oh, huh. That's a tough one. That's really a tough one. I like this philosophy that, we need to understand physics to remove it from our, from our modeling, you know, I think Jose said something like that in Princeton lectures. And, for combustion, we forget, for example, what is the added value of modeling combustion? When I model a fire in my building, where I predefine the heat of combustion of my sample, I predefined the total heat release rate and I predefine . No. So I predefined what amount of heat I generate I predefine what amount of smoke I generate. And by my model, I give it geometrical constraints where it will burn.

It will just tell you where the flame is.

yeah, I think combustion to the. It's like the same thing, but with additional steps in the meantime, which you have paid for. So, if I do predefined everything there, I don't need combustion to do that. I would need combustion if I wanted to know the oxygen concentration around, or I would like to know where the flame is, or I would leave the, the soot generation to be dictated by the physics model. Or I would like to know rate of reactions, or I would have underbite, ventilated fire. And I would like to understand if when pushing on the, small, uh, holes in the room can give sufficient oxygen or not, then sure. I would model combustion. But when I'm doing a simulation like that, I cannot model like I can model combustion, but it's just a fancy thing that gives me not that much and costs me a lot. And with turbulence, it's the same. You would have, two typical ways to model turbulence your flows. When the air flows is giant vortices of air that are constantly spinning, that are moving around, dissipating into tiner and tiner ones until like, uh, there is at Kolmogrov scale, there is none left. So, so that's, that's turbulence and someone defined it like turbulence is, is like pornography, you know, when you see it and he know what it is. So it's very difficult to define that, but you have to model that, that, that the way, how, how air moves, and there are two ways you can do it. You either average it over time, which we do with Reynolds Averaged Navier Stokes equations models, or you average in, in space. And you solve for the movement in time, uh, which is used in Large Eddy Simulation. That is the backbone of FDS, for example, The averaging in time approach is very efficient because you can go for a very long time steps, like one 10th of a second, maybe even more where you just assume that, the flow happens in the space, uh, you know, the average velocity, you know, The turbulence intensity. So how turbulent the flow was, and you know, that this patient rates, so how this larger vortices is, has shed smaller ones and how this energy is spread around. So this gives the solver a way to, to solve the momentum equation and give you a final flow field, which is averaged in time. So when you take a picture of a fire, in this model, average in time, you don't see a beautiful fire. You see like a cone or something with the different values, with different gradient of velocities or temperatures inside. So that's one way, and this is very time efficient. The only thing it cannot give you is the momentary value. And they've already mentioned before that wind is, uh, gusts, which can be much higher than the average velocity. So there's absolutely no way you can capture this with, with RANS modeling. You can to some sort the biggest vertices, because you see them forming with this solution. Especially if you go into very complex models, like Reynolds stress model, , then you have many more equations for every direction, how these dissipation happens. You can capture, but not perfectly in LES modeling, you do the opposite. You understand that there are multiple sizes of these vortices in the world And you understand that the big ones are the important ones because they move the air around. And the tiny ones are not that much important. They just like swirl until they dissipate and you just cut them. Like I solved the big ones. I forget about the small ones or I solved the small ones with. a sub scale model, uh, and that's how FDS works. It solves the big ones. It doesn't, it approximates the tiny ones and the interface between them is related to the size of your cell. So once again, we go back to the importance of the choice of your grid. now this solution is expensive because the timestamps have to be very short in order to capture the movement of this beautiful, um, vortices. What you gain for that is that you gain a beautiful image of, you know, this flames, puffing, smoke, moving in vortices. it's, it's a much more beautiful, uh, when you look at it and it's much more realistic and you can capture the gusts in that, but the price you pay is. Times 10 times hundreds of, RANS simulation. And the funny, if you, and we'd done that if you take a LES simulation and you average the results, you get the image like in RANS you know, because you suddenly time average everything. So, the choice of the turbulence is a difficult one in fire. People often make this choice by choosing software. If you choose FDS you use Les modeling, no other real choice. If you go into like ANSYS, that's a fun thing because you have to choose, you have to tell there is no default model. Default is, is laminar. So you, you need to choose your model and to choose, you need to understand, for us. we have. Usually using either, K Epsilon RNG, that's a Renormalization Group, not random number generator. How many people think it doesn't give random way as is the name of a group or the other one is k-omega SST and also works quite well for wind problems. So, so these are the two that would be models of my choice. And one day I would love to use Les for everything, but cannot afford it now, unfortunately.

And what about DNS? That reg numerical

Yeah. I mean, , that's

No, for wind,

um, this has been done till like Reynolds number of few thousands and, many people would use it direct number. The simulation means you just solved the Navier Stokes equations, for every scale that this in your analysis, up to the tiniest vortices The question is, do you need that, the, the best practices of, of modeling, flows with larger, the simulation there's this beautiful paper, thing, 10 questions about Les, uh, by, by Pope, I would highly recommend to anyone who's dealing with, with flow problems in fire and, somewhere in there, there's a thing that if you solve 80% of the energy within the vortices, with Les, with the direct model and 20% you approximate with subrogate model, you're pretty much good. So essentially four pieces you're solving directly. One piece is approximated. So your error is not that huge. And that justifies why you don't need, DNS and DNS. I mean, you can do DNS simulations for. Problems, but it's prohibitively expensive. It's just not with today's computers. Not, not reachable and maybe not necessary. I wonder, I don't think it is a, it's a nice validation technique,

this is a joy to interview you, but you, because I'm learning. Um, and also I'm in best agreement with the things that you're

That was the point of the podcast you learned so much, man.

you are indeed a man of a Renaissance. in modern times, the, there is a lot of things that you do that I value tremendously. Um, one of them is that you are a scientist and an engineer, something very important for me also that you do experiments and you do modeling. I know you, you, you, have more of a modeling background, but you also are involved in experiments. , so in the work that we have discussing today of, of wind and fire modeling, what is the role that you see to experiments on the topic of validation?

that's, that's a good one. That's a good one. You actually once said about us that, we are a center for experimental fire science, and I saw that tweet and I was like, oh my God. Yeah, that's actually what they want to become a center for experimental fire science. That sounds so nice. Let's build that up. And, and we're trying our best to be, to be such a hub for wind. Wind engineering is essentially an experimental field. , almost everything we know about it comes from experiments in wind tunnels. So even the best practices for CFD in many cases, refer to the best practices of wind tunnel engineering, and experiments in wind tunnels like the blockage radio you would like. how much of the space in your domain, you can block with buildings to make sure that, you don't affect the flow inside. So that comes from the, from the tunnel experiments. So that's beautiful. Now when with the wind and fire, experimenting is very difficult for two reasons, you either can achieve, the scale. So you can go into full scale and you can achieve full scale. And include the wind effort in that fire. If you burn a building outdoors there's winds, you can have wind FX in that experiment, but you do not get to choose the wins. You know, you, you work with what the nature has G gave in at that particular moment. Or you can go into wind tunnel and you can choose the wind. You have the perfect choice of what wind boundary condition you have. But you have to go into scale fire, which is not a real fire, which is difficult in interpretation. So I think the only people who can do it both are IBHS in us, they, they break the rules because they have a giant wind tunnel where you can put burning buildings inside. So, so th they, they broke the law there. It should be prohibited. It's, it's unfair advantage over the rest of us, but, but these guys can do both everyone else on the planet has to either, run the fire experiments in whatever, when do you find, or, or scaled down. And th this is very difficult. And in our project, you know, we choose the route of, numerical modeling as the fundamental tool of research, supported by wind tunnel modeling for, In terms of pressures on the buildings. So we know that we are unable to model wind and fire together, but we want to build a mock-up of the city that we are modeling in a computer in the wind tunnel, run, wind experiment in it, and figure out if the pressures on facades are accurate. If the flows are the same in the CFD and in the wind tunnel. So it would give us the capability to say, okay, to some salt, we have validated the wind model. That wind solution is correct in here. , and again, for me as someone who loves to work with, with full-scale experiments, this is also, as I mentioned, important thing to understand, to what extent when you meet on the field impacts your far expand meant, and to be able to quantify that in, in a way or another, that's fun. But yeah, there's definitely in the future in experimenting with wind, it's just hard to put fire in there unless you are IBHS

Awesome. Thank you Wojciech I do have two more questions, for the interview today. I wish we could have a follow-up. Uh, maybe when, when we have the episode a hundred, you invite me

yeah. Uh, to, to clear all the misconceptions that were said in episode 50 and why it's completely different.

so the first is, do you have a message for young engineers and young scientists that are entering into a topic of Fire and wind. When do you, is there anything that you want to tell them specifically? Not to the old dinosaurs, like, like us step to the two younger ones to the much younger ones.

Um, it's complicated, but it shouldn't hold you back from doing it the best you can, the interaction is so rich. There is probably not a single solution for that. You know, it will always be a problem of your. Architecture of your wind boundary conditions of your fire. It will always be something very unique that you find, and it is, it is hard, but it is worth it. I think, I think it gets you on a completely different level of understanding the phenomenon you're trying, especially the wind can be the driving force for many phenomena. I guess. That's, that's the first question you should ask yourself. Do I, by the literature study by the overview of the field, they think a wind can be the driving force in here, because if it's not, then maybe it's not worth it. But if there's any chance that wind can be a driving force in the thing you're trying. If you are omit it you just ommited the driving force. So, that's not, a great solution and it should not admit it to save time or to save computational resources. You never should do that. You should, try and use it. you should also try to see if there are simpler models than let's say CFD that we're using that can do both. Like if we go back to the tunnel example that we've discussed, if you wanted to solve that problem with, CFD, it will be very expensive, difficult, and long solution solution to do. It would be ridiculous to do. We had the luck, the tunnel was built and we could take measurements and approximate the coefficients for portals with, with measurements, but it would seem that it would be ridiculous. However, if you want it to solve the same thing with the simple momentum equation, You know, and, and put a wind as one of the boundaries in that reduced order model, you can do it in one, two days and have really decent solution. And that's what we did essentially in the end of the, of the test. So, it's complicated, but they may be reduced or their models that can allow you to solve your issue and understand this holistic problem of wind and fire interaction in your case. And then when you go into final modeling, maybe , you will know when is the driving force and what is not. Then if the driving force is improbable, because that's one in a thousand a year hurricane, you can cut it out and you're safe, but you have a proof why, if it's a driving for steel, then maybe you need to do some experiments, do some modeling, go, go deeper in it and see to what extent it is influencing your solution. Or if you don't feel like that because it's overwhelmingly, because I know you can jump into the loophole, which will be overwhelming. I mean, that's where we jumped and I know what I have went through to achieve my results. So, it can be an overwhelming, local, but you know, there are people there who do that and just open your mind and collaborate. And then like I found my Tomek who's the brilliant wind engineer. And he helped me tremendously. You need to find one for yourself. And, uh, he doesn't have the, it doesn't have to be an expert in fire, but they may be an expert in solving your particular problems. So, yeah. And you don't have to take everything on your own. Just collaborate.

that's an excellent call for the younger generations to embrace multidisciplinarity. Um, and not just do it by themselves, but during a team that sounds like a very modern approach to science. Also, I like that you are inviting them to go into a topic that is you are warning them. That is challenging, but you are promising them. That is reward. And that's fantastic. And then the last question then your audience and I, all of us would like to know what's next for you on not, I mean, you have a whole lifetime of fire science.. We know that what we mean is in wind and fire modeling. What's what's next for you?

Yeah. W within the phone, a couple of modeling, part three, hopefully in fire technology.. so we're in the middle of this, and granddad was mentioned where we want to see complete image of wind and fire interaction on over a single building. The what extent the wind would be a driving force in the building. Can we define in which scenarios the wind was the driving force in which it was not, then that the mind, the probabilities of this scenarios to go into quite the complete risk. Based analysis on, was it a major factor in overall safety of the building or not? And if it was, to what extent that that will answer a lot of questions on how wind influences fires in general, because this will be a nice case study to refer to. and I, I hope like we're completing the preparation stage. So there is just this few thousand CFDs to be done and we're, we're good. So, so that's, that's the nearest future for that. And I have brilliant postdoc Paulina Jaminska-Gadomska , who's working on this with me. She's a, wind engineered, trained by Thomas, my, my friend from Lublin. So she's very good at it. And then she was a little overwhelmed with fire aspect, but, it's under control. She's very happy doing that and she's discovering cool fire. Like we are, we're discovering wind and that's an exciting thing for her, I think as well. So that's the project. Uh, we also build a wind tunnel in, in ITB for modeling, natural ventilators. it's more like a business thing, you know, to, to investigate the performance of natural ventilators, according to standards, but we are doing funny things with it. Um, we're doing, Photovoltaic panel natural ventilator interaction. we want to, do scaled facades, especially scale facade tests, you know, and, and, and quantify to what extent an outdoor test is, uh, is changing the outcomes.

I don't have that question with you too.

that's an interesting question. This isn't it. So this such a topic to be answered. And, I really liked the interaction of things on the building roofs because I honestly hate the current way, how natural ventilators are put on the market based on their individual CV value. When, we know, Different the performance of the ventilators is on the roof. Like every ventilator has done performance. You cannot just standardize that. It's very difficult. So one day I would like to abolish that probably as evolution there, the, the standard to test a curve. It's not possible it's overwhelmingly on the market, but yeah, these are the next, the next nearest steps in, in wind and fire. And we're taking it one step at a time, going deeper and deeper, understanding it better and better and hope to provide a broad answers to this narrow question. This wind influence

Are you, are you going to have open tours to your tunnel facility, to your tunnel? wind tunnel?

If you're ever in Poland, the wind tunnel is in Pionki and,

I'm definitely, I'm already in the in the list. I wonder about the audience.

Everyone's in invited. If you ended up in Poland, just give me a call. We can arrange it. It's a good photo opportunity. The table rotates now. So you can take a like musical music video type of, shoots in insight. It's a great thing. We can measure your aerodynamics if you want. My mine is horrible.

That's fun. You should organize a conference, and you can invite, some of your most successful podcast episodes to, to, to be, part of a conference.

that may be a possibility we'll see in the future. I think it will be it's generally, it would be great to finally meet again. People like real people, not on the screen. And I cannot wait for, for the, for this chance to finally reemerge the damage already. So I'm happy.

What'd you, what was, what is the name of your postdoc? You

Uh, Paulina Jaminska-Gadomska

Perfect. So I think I'm going to propose that there is, if there is a quote for this episode, I think the quote should be, you might want to actually repeat it if you agree, and then we make it, your quote is, is the court is find your Tomek and find your Paulina

yeah.

And then if people want to know what that means, they actually have to go through the.

Yeah. That's the that's brilliant. And, , many people are meeting a wall trying to solve issues where others have solutions for and. You don't need to know everything. Sometimes it's, it's way more than sufficient to know who knows it and then just work together. And I believe in this collaborative minds of scientists, and it's also a philosophy that you did, you implement over your students, this openness, collaborative environment in which, the best things can emerge from. I think it's very rarely that a single person creates a groundbreaking

not not in the 21st century. Yeah. Maybe Newton, maybe Newton could do that. Now.

that was very hard. Yes.

And the last question is there is another type of fire smoke. And when, which is related to the Russian invasion, Ukraine, Putins war, that is actually as we speak ongoing, not far from you, but, I mean, this is probably not something that you want to model, but, is there other ways that you are thinking of how you can help these horrific events that, that are happening very close to.

In terms of, fire research, you know, and especially fire wind research. There was awfully lot done in terms of understanding how efficiently we can burn cities of our enemies probably more research on how can we efficiently destroy them? How can we efficiently protect? Unfortunately, the war time research, the post-war time research, one of the major threats related to nuclear weapon use is the radiation costs and the ignition cost by this, uh, this weapon and how fires will spread. There was immense amount of work done on how fire spreads in cities, how fire, these choice cities, but also how to, in a way, make sure that your city is hit less than the others city. fire engineers have surprisingly long, history of researching that field. I just think it's not my generation and not yours generation. It was the generation of the dinosaurs before us who, who were much into that. And there, there are NATO guidebooks about that. There's the Pitts review that I mentioned that there's lots of literature that can be used. And, yeah, fire is, is a weapon. Of course it is a weapon. the second issue that I think that is related to war, but not directly to military actions is the safety of the people who escaped the world. And there's 13 million Ukrainians relocated by the war. Imagine that's the country that has 40 few million people. And 13 million of them are not in their home because of the war. One third of the country. And. To the best of my knowledge, they don't have huge camps yet, but probably it's Inevitable. if this continues and then we go into the struggle of providing fire safety to them and they have, they have bigger problems than fire on their head, you know? they don't really have a chance to care about it that much. And I know your director at Kindling . And, I admire that company a lot because you guys are doing what these people cannot or it's not on their mind at the moment. And I think providing fire safety to those who escape or is tremendous value, and it is very important to understand the challenges they go through. And if there are solutions that can be given advice to them that will solve some of the problems that they may not even consider a problem, but they are problems. So, I think it's a difficult subject, but I'm very, very happy that there are people like, like people on Kindling who, who actively work on that and, already are looking into solutions. So,

yeah. I know Daniella is who spoke in your, in your podcast. It's just always talking to you in how they can help. It's just that as, as engineers know, we, we try to fight fire as an accident. it is very, very sad to have to work sometimes into fire protection or protection of people when it's not an accident when it's the military, abuse of someone else that that is very hard. That's very unfair. And there's one, they call is higher, right. To help those that are being, hurt, uh, left and right. Um, by,

yeah,

for very valid reasons.

But thankfully, you know, the work that has been done, the whole fire protection engineering industry, all the work on the fire resistance, all the work on the reaction to fire of materials, our cities are built different than a hundred years ago. They are more fire resilient. They are more fire safe. So thankfully in a passive way, by stopping this accidental fires, we're also making the job of the incinerators more difficult than that. That's probably one, one good thing again, from our fields too, that the people in Ukraine at the moment, there's always a war in the world. So many others are affected by wars everywhere. And the fire is a part of them always.

Okay. Thank you, Wojciech was an absolute joy to interview you. Um, there was my first interview and I interviewed someone and I, I did adjust. And following your book, I listened to you weekly, so I learned how you do it. I thought I'll try something similar to that.

thank you, Guillermo you did a great job on that. It was a huge pleasure to be interviewed. I have a fresh page, no notes on my eyes. So refreshing, you know, time I do it, I it's. It's a new today. Yes. Thank you. Well, thank you for, for doing that and, uh, yeah. Happy to lend this, tube to spread important messages in the future as well. Cheers, man.

thank you. Thank you so much.

Yeah, that's it. It's me again in the driver's seat. Thank you so much Guillermo for taking it over for one episode and thank you. So much actually for proposing thing this and organizing it, then. I was a nice surprise and. It was way more fun than I thought it will be in. Actually it's surprising how much the perspective changes from host to guest. It was a, I guess, a much needed experience for myself. So I can torture my guests better actually. And anyway, I hope all of you enjoyed the wind and fire things. I had to say. It's a fantastic topic of science. It's a subject that's never ending because as many buildings as many fire and wind the interactions there will be. And I think if we want to build a fire safe society, we need. To understand. This phenomenon better. My team and I. We are working very hard to understand it and build guidance for the rest of the community. And we are very. Very happy to see other groups interested in the subject. Uh, other researchers pursuing. The research in the field of wind and fire. It makes me very happy to see how this field. It is growing and so much. Great stuff will come out of that. And yeah, I will not torture you anymore. With wind and fire science. I think if you sustain so long, uh, You're good. And, thank you for doing that for thank you for enduring. That thank you for being here and yeah. 50 episodes of the podcast. It makes me very happy and I hope. You enjoy this show as much as I do. And if you do well next Wednesday, next time I see you episode 51 and it's going to be. Awesome. As usual. See you there. Thank you. Cheers.