Engineered Approach to Life Safety in a Shopping Mall Using Autodesk CFD

DR. M. MUNIRAJULU: I'm glad you could make it to my session in time, and I appreciate-- are you enjoying AU here?

AUDIENCE: [INAUDIBLE]

DR. M. MUNIRAJULU: OK, good. I am a first timer here, both attending and presenting. And after two days of hectic walking around and attending various sessions, I have started to love AU. Yeah. The title for my topic today is "Life Safety using Autodesk CFD." If I'm fast, let me know. I'll try to go slow. I'm going to talk about two or three important outcomes from Autodesk simulation CFD in the design of safe buildings, especially life safety.

OK. Let me introduce myself. My name is Munirajulu. I don't know how many of you will be able to pronounce my name. But in short, we can remember call me as Muni, M-U-N-I. Is that OK? Fine? Yeah. Thank you.

In India, when somebody meets another person for the first time, when they get into conversation, they ask a lot of questions. Where are you from? Where do you work? And then, are you married? And how many children you have. So that is a kind of a culture we have.

So let me introduce myself also in that way. I'm married for the last 20 years, and I have a wonderful family, my wife and two daughters. On the professional front, I am an engineer by qualification, as well as experience. I'm a kind of a CFD technology savvy. For the last 21 years, directly or indirectly, I've been involved with the application of CFD. I have worked in varied industries, products ranging from steam turbines, steam boilers, automotive, [INAUDIBLE].

And I like to keep up with the new technologies, the technology trends. So a few months back, I got myself certified as a IoT professional, Internet of Things. So there is no connection between CFD and IoT right now, but that is my interests. And as I told you, I've been working in the engineering and R&D setup and mainly using CAD, FEA, CFD. OK. Currently I'm chief engineering manager in Larsen & Toubro Limited, India.

OK, real quick information about the organization I work for. Larsen & Toubro, or L&T, is a US dollar $17 billion in revenue. It's a technology, engineering, construction, manufacturing, and financial services conglomerate, with global presence. We are into key sectors-- hydrocarbon, infrastructure, power, process industries, defense. And L&T construction, which is a construction arm of L&T, is India's largest construction organization. And we are ranked among the world's top 30 contractors in that space.

OK. Let me ask you this. How many of you have been to India? Nobody. OK, how many of you are planning to at least visit India, maybe desire? OK, that's good. That's good. Thank you. Thank you.

L&T as constructor designed and built major airports in India. When you visit India, when you land up in Delhi International Airport or the Mumbai International Airport, or Hyderabad Airport, and those airports are designed and built by L&T.

And how many of you are in information technology business? Many of you. Yeah. And you know from India, we have major IT business providers-- TCS, Cognizant, Wipro, and Tech Mahindra, companies like this. And when you visit those companies, to their business facilities, to their offices, you will see landmark structures. So these were designed and built by L&T. OK.

The key word for today's topic is safety. I'm worried about what will ensure safety of people in a commercial building, such as a retail shopping mall, in the event of fire. And that will be based on the application of Autodesk CFD as a design and analysis tool.

I have split my presentation into four parts, four key objectives, learning objectives. And these are based on the CFD project that we have done in the design of a retail shopping mall for India. And one more thing I forgot to tell is that currently we are building Al Rayyan Stadium in Qatar, which would be used for FIFA 2022 World Cup. If you happen to be there, you can remember this. OK.

The four objectives, they cover the fired dynamics modeling. And we will look at key results in terms of life safety assessment, then some conclusions and recommendations that we can draw from. The fourth objective would cover an attempt that we have made to optimize smoke exhaust system design by analyzing different fire scenarios.

Class summary. I'm not going to go into details, but I'll give you an highlight. For smoke inhalation, there are other analysis software available in the market. But here we have chosen to use Autodesk Simulation CFD because our building design data, the 3D models and drawings, we are using Revit for producing 3D models and drawings. So to take advantage of the interface between Revit and CFD, we have chosen Autodesk Simulation CFD.

And as some of you know, CFD is not done for all cases. CFD is done-- analysis is done-- when a building is large, and the geometry is complex. Because there, if the building is large and the geometry is complex, the traditional techniques will not be-- you will not be able to use it for life safety assessment. OK.

OK, before we jump into the details of what we've done, I would like to go through some basics. Because understanding these basics would help us to relate to why we have done what we have done. So these basics are about what is fire-- fire triangle, tetrahedron, and stages of fire. What is fire dynamics? How does fire spread? How do we quantify fire? And then classification of fire.

For some of you, these are very, very basics. But still, I would like to go through quickly. For some of them, I'll give highlights. And some of them, I'll give some more details. OK. Sorry about that.

What is fire? OK, all of you know, if you sit in front of a fire, you feel the heat, and then see the light. But from the engineering perspective, fire is basically a exothermic chemical reaction that produces heat and emits light.

And how do we quantify fire? During fire, heat energy is released. And the rate of release of energy is called HRR. And this is also called fire power. And this quantity is very important because this is an important input into the CFD analysis for the fire source.

Classifying fire. Fire is classified into different types-- A, B, C, D, K. I'll not go into the details. It is there in the handout you can go to later on in the web link I have given, and that will give you more details. This fire classification is based on the type of fuel that burns. And this also helps us basically to decide what kind of a metric to use for preventing or suppressing fire.

Now details about some of them. Fire triangle, when fuel, heat, and oxygen come together, fire starts. So the geometrical representation of that is basically fire triangle. Now, once fire starts, for it to exist and continue to exist, we need a fourth element, which is a chemical chain reaction, uninhibited chemical reaction. So when these four elements are geometrically represented, that is called fire tetrahedron.

Now why I'm showing this is for preventing fire or suppressing fire. If one of them is eliminated or one of them is reduced, fire prevention or suppression happens. So to highlight that point, I have put these pictures here.

Stages of fire. Fire starts at ignition and grows as more and more fuel is burned and reaches a peak value of heat release rate and also temperature when all of the fuel has caught fire. And after the available fuel has caught fire, the intensity reduces, and fire basically decays. This is a traditional fire development curve in terms of temperature and time.

Why this is important, and why I'm showing it here, is because for designing smoke exhaust systems, we design fire, whether it is a steady state that means the heat release rate doesn't vary with time, or unsteady state fire, that is heat release rate varies with time. So one of these two approaches are taken. And in CFD also, we have to use one of these approaches. So it is important to understand what kind of a fire growth we are considering in the analysis.

OK. Once fire starts, how does it spread? So fire spreads by heat transfer. And heat transfer plays a very critical role in ignition, growth, decay, and extension of fire. And as you know, heat transfer takes place by induction, convection, and radiation. Heat from fire, while it is spreading by conduction, basically it spreads to the solids.

For example, in building, building elements like walls, structural elements, so that heat transfer through the solids is called conduction. Heat from fires spreads by convection, that is moment of hot gases. For example, hot smoke close to the fire and to a region away from the fire, or hot air moving from a hot environment to a colder environment. This is called fire spread by convection.

Then finally, fire spreads by radiation also. Heat from the fire spreads by electromagnetic radiation. Some of these already you would know it, but I'm presenting this here so that it would relate to actually what we are going to see with the smoke analysis.

All right. So we have seen what is fire and how fire grows, and now it's time to look at why we are interested in the fire. Fire releases heat energy, and temperature is a measure of that heat energy. And the human body, human skin, responds to temperature exposure in different ways.

I have put up this information here. For example, 44 degrees Celsius, the human skin begins to feel pain. 48 degrees, first degree burn injury. 55, the human skin receives a second degree burn injuriy. 62, that is a kind of a limit where human tissue, the burned human tissue, becomes numb.

So when fire happens, from the point of view of life safety, the temperature exposure to human skin is very important. So that is the reason I'm giving this data here. I've taken this information from NIST. So the weblink has been given there.

And fire also releases smoke. Smoke contains toxic gases, like carbon monoxide, hydrogen cyanide. And this bar chart shows most deaths during fire happens by smoke inhalation, rather than by burns. Almost 2 to 3 times of deaths during fire happens because of smoke inhalation, compared to smoke burns.

Apart from human loss or personal loss, there is an economic loss. Here, I'll put up an information about the total cost of fire. This is in 2011 data, roughly 2.1% of US GDP goes into managing the total cost of fire. And the stabilizer split up of various components. If you are interested, you can go to this website, [INAUDIBLE], or in the handout, I've given these details.

Fire loss in the US. This pie chart shows during 2016, 74% of the fires are structure fires, meaning the fires that happen in residential buildings and commercial structures. This is 2016 data, again from NFPA.

OK. This infographic I have put here to show that fire and its consequences are a around the clock concern for us. For example, here one structure fire was reported every 66 seconds, based on 2016 data in the US. And then one home fire every 90 seconds. And then one highway vehicle fire every three minutes. So it is a kind of around the clock concern if fire happens.

OK, knowing that, there are consequences to fire. And what do we do? Can we do something to mitigate it when it comes to building? Yes, we can do it. In the design and development stage itself, we can use proper tools. And we can use Autodesk Simulation CFD to simulate the smoke movement and temperature development in the building for life safety assessment. OK.

When we use design analysis tool like Autodesk CFD, a key challenge is, how do we represent a real fire into the software? How do digitally represent that real fire in the software? In Autodesk, the five source is represented as a fire part. And what is fire part? Fire part is nothing but it's a CAD part, a short cylinder enclosed in a steel ring.

And this fire fire part is assigned a resistance material, meaning it has some porosity. And this fire part is kept little above the floor to allow for the air to flow in to simulate the smoke movement. That is, flow enters and leaves through the bottom of the top surface of this fire part. That is why the enclosing steel ring is kept to ensure that.

OK. Now into the details of a real project, for which we have done the CFD. This is where the rubber meets the road. We have done the CFD analysis for a retail shopping mall. This is, again, a transit-oriented development project already completed, designed and developed by L&T in Mumbai in India. So here, this is the ground level entrance to the shopping mall,.

And this is where our replacement air, our fresh air, will enter in case of fire. So that is represented in the CFD model in terms of a CAD part. Of course, we have taken the CAD geometry from the Revit, and then we have done simplification suitable for CFD. And this retail mall building has upper ground and two levels, G plus 1 and G plus 2.

You have a large central atrium. And on either side, you have three small atriums in this retail shopping mall. A little bit more details to understand the geometry of this building. So here I've given the sectional details. Upper ground level, G plus 1 and G plus 2. Here I have indicated location of the fire that is on the ground.

And this hatch portion here, that represent the corridor space where people will be moving during shopping. And they also become exit paths when the fire event. This x- and y-plane I have indicated to show that the CFD output we have taken on these planes, the visualization of the CFD results we have taken on these x- and y-planes passing through this fire.

Now, if you look at the G plus 1 and G plus 2 sections, you'll see that, again, you have hatch portions here, and again hatch portions here. These are all corridors meant for people walking around the shopping mall. Then this [INAUDIBLE]. They are the wide spaces. Basically, they provide interconnecting spaces from the upper ground to the G plus 1 to G plus 2 to the atrium ceiling, the roof.

This is a 3D section again. This gives a better view of the building geometry. You have the fire kept on the ground, and then you have the corridors for people to walk on all the levels-- 2G, G plus 1, and G plus 2. The atriums are covered by skylights.

OK. After looking at the CAD part of it for CFD, now we will get into how we model the fire in Autodesk CFD. Here we have assigned a resistance property to the fire part with a free area ratio of 0.85 Wife and high thermal conductivity, a value of 200 watt per meter per Kelvin.

And in the Resistance Direction menu in the property settings, we have indicated floor direction as global z. This is basically-- software doesn't know whether the smoke will go up or down. So we have to specify the expected direction of the smoke flow in the software.

And the fire source has been shown to grow as a fast I-squared fire growing up to 2.5 megawatts, and then remain steady. And we have used convective portion of that heat release rate. The reason being, why convecive portion alone, that is about 70% of the total fire char. Because the smoke generation and the moment is controlled by the convective portion of the heat release rate. That's why we have taken only 70% in this model.

And since it is a time dependent fire, we have given a time code to indicate that the heat release rate is dependent on time. Mesh details, very important. If some of you don't know what is a mesh, for example, this room, this white space is filled by air. And this space will be divided into small blocks. And in each block, these flow and panel parameters will be calculated by the software. And that is what we call as a mesh.

And we are able to give very fine mesh in the fire part. Here, we have given a uniform mesh of five elements. And also to capture the flow below the fire part, because the fresh air will enter underneath here. So there also we have to give mesh refinement. And since smoke is expected to grow and develop a column of smoke, to accurately predict the smoke development and the spread, about this fire also we have given the regional mesh refinement in Autodesk. The sold ring is suppressed basically to indicate that we are constraining the airflow to take place only from the bottom of the fire part to the top of the fire part.

Material properties. Since fire is involved here, temperatures are high. And buoyancy effects will come into picture. So we have given density variation with temperature. Boundary condition. Now, for the volume occupied by air in a building is the [INAUDIBLE] domain. It will have inlets and outlets. So at inlet, I have specified temperature, pressure, and a scalar boundary condition. These temperature and pressure are basically ambient conditions for [INAUDIBLE] condition in Mumbai.

The scalar value of zero here, this indicates that we are specifying at the inlet, the air is fresh. That means it doesn't have any contaminant or smoke or whatever it is. Now, the outlet boundary conditions. Outlet boundary conditions are given at the smoke exhaust fan locations. And we have considered the fans only in the zone 2 here. I'll come to that later on. And I have indicated here what is the fan capacity for each fan number.

OK. After having created a CAD geometry for CFD and then assigning the fire part properties, now we have to-- based on objective of the CFD analysis, we have to specify the solution parameters. And one of them is whether we are doing a transient CFD analysis or a steady state CFD analysis. So here in the solution mode, we specify as transient. And in result quantities, we have switched on smoke visibility because we want the smoke levels in this building for life assessment.

And when we are switching on the smoke visibility, we have to provide the smoke visibility parameters, given here. It depends on what type of combustion, for a flaming combustion is the value. And this is the signage constant, whether it's an illuminated signage or a reflective signage. And this is a [INAUDIBLE]. [INAUDIBLE], we have taken an average of [INAUDIBLE] polyurethane form for a fuel package of typically a kiosk fire and a sofa fire.

And our objective is to simulate smoke and smoke temperatures. Both fluid flow and heat transfer would be involved. So we have switched on flow and thermal in the physics menu here, [INAUDIBLE] menu. And in Autodesk CFD, smoke concentration is calculated as a scalar quantity. And it doesn't give exact values of the smoke concentration, but it gives you a scalar quantity. So we have to switch on general scalar on and specificy diffusion coefficient.

And this diffusion coefficient value we have taken based on the Autodesk CFD best practice suggestions. And this value is not very critical, but it's good to keep a approximately relevant value here.

OK, initial conditions, since it is a transient simulation, we have specified initial condition of scalar 0, means it is smoke free initially, and then temperature of 32.5 degrees.

OK. So having looked at how we model the CAD geometry and how we specify the solution parameters, the fresh air properties and smoke exhaust fan capacities, the meshing details, it is time to look at key results. For a designer, the most important key results are smoke free space and tenable smoke temperature for people to escape during the event of fire. The smoke free space is defined as a space where the visibility is more than 10 meters, more than or equal to 10 meters.

And the tenable limit for temperature is taken as 60 degree Celsius based on the BS standards. And Autodesk Simulation, as I told you, doesn't give the [INAUDIBLE] concentration for the assessment of toxicity levels. But the standards mention that toxicity is deemed acceptable if visibility is greater than 10 meters. So we will focus only on the smoke visibility and the smoke temperature as parameters for life assessment in the building.

OK. Key result one, smoke visibility. I have given here smoke visibility plots in the z-plane, which is 1.8 meter above the G plus 2 floor. That is basically the walking height for the people in the G plus 2 floor. And y-plane is basically a vertical plane on which I've given the visibility level values.

So here you see that the visibility values are greater than 10 meters below this 1.8 meter height, meaning this is occupied space for the people moving around in the shopping mall and, the smoke levels are actually above that. So this plot is at about 19 minutes. So people actually would easily escape during that period. And here also you see the walking corridors are free from smoke. These are the walking corridors.

Key result two, smoke temperature. We are looking at a tenable smoke temperature. Here again, the colorful plots-- x-plane, y-plane-- if you look at the values below this red line, these temperatures are below 60 degrees, which is OK.

Key result three, very important. What determines the smoke movement and smoke spread? And related to that, the smoke temperature levels is how this airflow field develops within the building in the event of the smoke fans coming into operation and fresh air being drawn. So here I have shown just near the fire part, the vector plots, and also what happens to the air flow field at the entrance to the shopping mall, which becomes an exit when people are actually coming out.

So here, the standards mention that if the air contracting the smoke plume is more than 1 meter per second, the smoke plume actually will get deflected, and it will disperse into the occupied spaces, which is not good. Here, in the results that we have got is again 19 minutes, at 19 minutes from the start of the fire. So here you see that the airspeed conducting-- or the average velocity conducting the smoke plume here-- is 0.8 meter per second.

Now, when people are exiting through the doors, the fresh air is being drawn in the opposite direction. People are going like this. Fresh air will be coming in. And if this replacement air velocity is more than a certain value, it will hinder the people movement. It will obstruct the people movement. So the BS standard mentions 5 meters per second as the limit for unhindered escape through the exits or the exit doors. And what we've got here is much less than that. So here we have 1.6 meters at the entrance, and down the line we have 1.4 meters. This is acceptable.

OK. So far, the key results I have presented only at a time close to the escape, and there's 19 or 20 minutes I have presented these results. But for a regulator or a designer, the information about these three quantities-- smoke movement, temperature levels, and the air flow field-- these values, these levels, are important throughout the escape time. That is, from the moment of start of fire until the 20 minute escape time.

So I'm going to present to you some details how the smoke develops and how it spreads during that time. OK. These are the smoke visibility plots in the x-plane. I have given the results at various times. 30 seconds, it is just trying to develop. The black area is basically smoke filled area, and the remaining areas are smoke free in this picture. That is, this is the smoke filled area, and the remaining areas do not have any smoke.

And then at about one minute, the smoke rises to the atrium ceiling, and it starts accumulating there as a reservoir. Then from one minute down to 19 minutes, the smoke continues to rise. And it also starts coming down. But what we have seen is, it has not yet started coming down, even at 19 minutes, to the occupied level. So this kind of information, you will get it when you are evaluating the smoke levels from CFD, Autodesk CFD.

This is more clear in the y-plane. So here you will see at about three minutes, the smoke starts getting exhausted here from the exhaust fans. OK. And as time continues, smoke would go out, get exhausted. And some amount of smoke also starts coming down. This is because there is a balance happening from the quantity that is being exhausted by the smoke fans and the momentum-driven flow that happens because of the smoke rising due to high temperature levels.

And again here, the smoke levels are not coming down below the 1.8 meter level. That is where we consider the occupied height in the G plus 2 floor. OK. This gives a nice picture about what happens to the smoke at 1.8 meter level. This level, if you look from the top, this is how the smoke visibility values will look like.

And here, you see that throughout the-- from the fire start to the end of the 20 minute period, these corridor spaces are free from smoke. Of course, this is a column of smoke, which is represented as circle when you take the horizontal plane. So this indicates that when people are moving from the start of the fire until 20 minutes, people can easily escape from this building. They can at least reach the exit staircases.

OK. And second most important result is temperature development. Just like we saw smoke development with respect to time, we can also track how the temperature levels exist during the course of this fire. 30 seconds, just smoke getting hot, and then the temperature keeps on increasing. But one thing I would like you to notice is this-- sorry.

This high temperature smoke is confined to only the atrium. It is not getting into the occupied areas. See here, here, these occupied areas, temperature values are not very high. So only in the atrium, the high temperature smoke is rising and reaching the ceiling and then getting exhausted.

Again, on the y-plane, a similar result. But this gives a better visualization with respect to the temperature levels. And you'll see that below 1.8 meter level, temperatures are less than 60 degrees, which is our requirement.

OK. This fresh air flow moment with respect to time again on a plane, which is on a z-plane, which is close to the fire on the ground floor. Actually, this airflow field is very important, as I told you, for the smoke development and movement. At about one minute, the fresh air is getting drawn. And by about three minutes, this airflow field gets stabilized. And from three minutes down to 20 minutes, this airflow field will more or less remain the same.

Why I am showing this is, if this airflow coming in contact with a plume changes, or if it has higher values for some reason, then what it would indicate that at that particular point of time, the smoke has a potential to deflect and get dispersed. So that is the reason I'm showing this airflow field near to the fire location.

This is in more detail. Here you see that it's a vector plot. This is a kind of beautiful results. You will get it from Simulation CFD. This is the smoke plume vectors. And the fresh air coming in contact is shown in the red arrows. And you see that at each level-- UG level, G plus 1, and G plus 2, this fresh air is coming towards the plume, which means it will not get deflected or dispersed.

OK, I just want to summarize. What we see from this analysis is a well-defined smoke plume. And it is rising towards the smoke exhaust fans. And another thing we saw is that make up air doesn't disturb the smoke plume. So no smoke dispersion or visibility problems. And hot smoke temperature is below 60 degrees, which is within the acceptable limits for human safety.

Some conclusions. The smoke exhaust fan capacity and layout is adequate to provide a smoke free space for 20 minutes. And based on the smoke exhaust fan layout and fire zoning logic, the CFD analysis using Autodesk CFD can accurately predict smoke development, smoke movement, visibility levels, temperature distribution. Hence, this approach can be used to validate the ventilation design for anticipated fire scenarios.

And one other thing is, the fire engineers have to actually do some fire zoning first before we take up the CFD analysis. Because if we don't do the fire zoning, and you go by some prescriptive-based design, it may not work. So we have to do some kind of fire zoning first, and then we can use the CFD to optimize this fan layout.

OK. This is the last part of my presentation. We get a lot of insights from CFD analysis-- how the smoke is produced, how it grows, and how it spreads, and what are the temperature levels, and how the air flow field happens, and things like that. But as an engineer, it is important how we use that information. And basically, use that information for optimizing the design.

I'll just show one attempt that we made to optimize the design. Two things we have to remember here. Smoke ventilation is created by the mechanical ventilation fans kept at the higher level. And the fresh air enters at the ground level or the lower level. And then the prior distribution of smoke or air velocity is not known for any given design at the podium level. So therefore, you have to actually consider simulating more than one scenario to complete an optimized design.

Here, design one and design two. Design one is our initial design, and design two is a final design, optimized design. In design one, what we have considered is the entire building on the right side-- that is zone R-- is considered as one zone, and then this one is considered as another zone. And accordingly, smoke fans were designed and kept here.

What happens is, anywhere fire happens in this zone, all these fans will get activated. And in this case, we have divided into a smaller number of smoke zones or fire zones. And if fire happens or detected in any one zone, the fans get activated only in that zone. So I have given the fan capacity in these tables for design one and design two.

At this stage, I would like to mention a very useful tool for optimizing smoke exhaust system design. This is AtriumCalc version 1.1 by John Klote. And based on the design fire size and outdoor ambient temperature, and the height from the base of the fire to the smoke layer interface, with these three inputs, this spreadsheet calculation will give you the volume of smoke to be exhausted based on the fire size. And not only that, the number of fans and the layout can be decided to a wide plug holing. So all these calculations are available in this spreadsheet, and it's very useful.

OK. We'll just compare the results for design one and two. In design one, you see that smoke gets actually deflected towards the right, and then it starts entering the G plus 1 here, G plus 1, here G plus 1 and G plus 2 levels. Whereas in design two, it just rises in the atrium space here. It just rises straight. It doesn't get deflected or dispersed.

This is on the z-plane at 1.8 meter level above the G plus 2 floor. And here you see that by about four minutes, the corridor space is already filled with smoke, meaning the visibility is very less here. Whereas in the design two, the corridor spaces are free from smoke. By about 10 minutes, the corridors are getting filled, and visibility is affected here. But whereas here, in this design two, we don't have any problem. Visibility levels are about 10 meter.

Here we compare the results for temperature, smoke temperature for design one and design two. Again, if you look at design one, because the smoke plume gets deflected, there is a risk that high temperatures can exist in the occupied spaces. Whereas in design two, you don't have that problem. And if you look at the airflow fields, again, this airflow field shows it is getting deflected. And here it is not.

And if you look at the horizontal planes here, you'll see that the air conducting the smoke plume is more than 1.8 meter, whereas here it is 0.8 meter. So that itself actually indicates that in design one, there will be visibility problems. There will be a problem for people to escape because smoke starts getting deflected and dispersed into the occupied space.

This is, again, a comparison of airflow field for design one and design two. Design one, you have the average velocity conducting the plume more than 1 meter. In design two, we have 0.8. And here you see in the design one, the fresh air coming in contact with the plume is very erratic. It is coming towards and also going away from it. Whereas here, the fresh air coming in contact with the plume is always towards it. This ensures that the smoke will rise nicely like a column of smoke to the atrium, whereas here it doesn't. It will get spread. It will get deflected.

OK. Just to conclude the optimization exercise, what we see-- basically, we compare three parameters-- the smoke visibility is one-- and here we see in design one it is not acceptable. In design two, it is acceptable. And second parameter, smoke temperature, in design one it is risky because the smoke temperature can be-- hot smoke can get into the corridor spaces. And in design two, it is below 60 degrees. It is acceptable.

Then airflow field, for design one not acceptable. Design two, it's acceptable. So if we want to conclude about the smoke exhaust system design and layout, for design one it is not suitable. But in design two, it is adequate. Adequate to provide the smoke visibility and tenable temperatures throughout the escape time of 20 minutes.

OK. Now it's time for any Q&A. Any questions? Yes.

AUDIENCE: How big was the model?

DR. M. MUNIRAJULU: How big was the model?

AUDIENCE: [INAUDIBLE] mesh size.

DR. M. MUNIRAJULU: Mesh size, yeah. We tried various mesh sizes. Initially we just did the analysis maybe 6 million. And then to get proper results, we went to 12 million. And then finally, I think we concluded this project using about 24 million. Yeah, it's quite extensive. It's very mesh intensive work. This program would run for two, three days continuously to get results. Yeah.

AUDIENCE: So you mentioned that you [INAUDIBLE] Revit [INAUDIBLE] detailed. You said something about the efforts to get the Revit model in a less detailed [INAUDIBLE]?

DR. M. MUNIRAJULU: Actually, when the Revit model is exported into Autodesk CFD, you would get all the details. But you may not get the CFD volume that you're looking for. OK? Because there would be some gaps in the actual Revit model. And there will be some details that you may not require, which will create meshing problems in the CFD volume. And there will be components which will be intersecting, which will be sitting on edges and things like that.

So it is-- we have tried both. For a simple geometry, we just could get directly into the CFD. But for this geometry, we had trouble. And what we did is we used the existing Revit model to redo wherever we have problems. After creating the CFD volume-- actually, while creating the CFD volume, we come across some problems. At that stage, we would actually modify some walls or the intersection between the floor and the wall and the ceilings and things like that. So we had to do a lot of correction to get this CAD geometry. It's not easy. I'm not telling that it is 100%. You can just get it [INAUDIBLE]. Yeah, it involves effort. Yeah.

AUDIENCE: Did you have to check different fire locations, or [INAUDIBLE]?

DR. M. MUNIRAJULU: Sorry, I didn't get your question.

AUDIENCE: Did you have to check different fire locations where the fire is located?

DR. M. MUNIRAJULU: Yeah. Yeah, we tried I think two or three fire locations. And I'm not a designer. I am basically an analyst. And the designers decided-- they anticipated our probable fire location as something that we have used here. Yeah. So we have analyzed various fire locations. Yeah. Yeah.

AUDIENCE: [INAUDIBLE]

DR. M. MUNIRAJULU: Yeah. Yeah. We have taken the fire to grow as a fast t-squared fire. That fast t-squared, how it grows with time, that decay we have taken it from NFPA 92B. There is a constant, and the convective portion of the heat versus the time, it's basically q is equal to some constant t square. So that constant is there in the NFPA. And for some fires, some real data is also available. So it is up to us to use a kind of realistic data.

So that's what I can say, yeah. There are three types of fire growth, slow, medium, and fast t-squared. So here, we have taken fast t-squared, NFPA 92B. So using that equation, we can actually decide the-- it's not a linear variation. It's actually more like a [INAUDIBLE] goes like this, like a parabola. Exponential, yeah. Right.

AUDIENCE: Question over here. So I'm not used to using [INAUDIBLE], so I'm just curious about the software, and using other [INAUDIBLE]. And I'm curious about the validation [INAUDIBLE] validate it for various simulation [INAUDIBLE].

DR. M. MUNIRAJULU: Yeah, correct. So let me make this clear. We don't use-- we don't validate the software. We use a software to validate the design. That is number one. And number two is, which is relevant with respect to your question-- for example, the smoke development and movement, we have taken one reference for which the data is available. It's basically a standard room with a fire size. And we have used that reference basically experimental data.

And then we did the same analysis using Autodesk Simulation CFD. And we have compared the smoke levels and the values. So with that kind of a exercise, we went ahead using the Autodesk CFD. I can share that result with you if you want later on. Yeah.

AUDIENCE: Sure. [INAUDIBLE].

DR. M. MUNIRAJULU: Sure, yeah. But validation cannot be done for a large complex geometry. It can be done only if you have the data for whatever building it is.

AUDIENCE: [INAUDIBLE]

DR. M. MUNIRAJULU: Yeah, correct, yeah.

AUDIENCE: [INAUDIBLE]

DR. M. MUNIRAJULU: Yeah. Another way to do it is basically you do the same thing in FDS [INAUDIBLE] or some other software and see how you get the results. Yeah. Yes. Somebody there. OK.

AUDIENCE: I just [INAUDIBLE].

DR. M. MUNIRAJULU: All right. Yeah, yeah. OK. Any more questions? Any more questions? OK.

So thank you so much, and I request you to give the feedback to me because if you want me to come next time to AU, definitely I need your feedback. OK. Right. And thank you for listening, and let's keep in touch because we can share a lot of information, and [INAUDIBLE] new tools and new approaches. And we can share information. We can try to solve each other's problems.

So my email address is given here. And if you want to know about my company, what we do and stuff like that, the website is given. We can go through the website. And if you want, my visiting card. I'm here. I can give the visiting card to you. So thank you so much once again, and I hope it was useful to you.

[APPLAUSE]