Welcome back to Sports & Building Aerodynamics in the week on building aerodynamics. In this module, we're going to focus on natural ventilation, and we start again with the module question. Let's focus on natural ventilation of a large building situated in an urban area with small ventilation openings. And this ventilation should be assessed and improved. Which of these methods do you think is most suitable for this purpose? Is that on-site measurements? Reduced-scale measurements in an atmospheric boundary layer wind tunnel, or CFD simulations? Please hang on to your answer, and we'll come back to this question later on. At the end of this module, you will understand the advantages and disadvantages of several methods to assess the natural ventilation of a building in an urban area. You'll also understand the important effect of urban surroundings and wind direction on the natural ventilation of a building. The case study that we'll focus on here is that of the Amsterdam Arena stadium in the Netherlands. A study that we performed some time ago. More information on this study can be found in these articles, and in the next slides I will provide a short summary. So, what's the problem statement of the stadium? Well, it's a stadium of course where football games are being played, but it's also a stadium where concerts and festivities are being held, and you see a concert in preparation on this figure here. So the stadium actually has a semi-transparent retractable roof construction. It does not have HVAC systems, so no heating, ventilation and air conditioning systems included. And indoor climate can become a problem. Certainly during concerts, when the roof is closed, and the light and sound equipment is mounted on the roof, and 55,000 people are inside that emit a lot of heat, a lot of vapor, a lot of CO2. Because there are no HVAC systems, you need to count on natural ventilation to ensure a good indoor environment. So the research aim here was to first analyze the current situation, and because ventilation there was actually insufficient, to analyze alternative ventilation configurations and their effectiveness. Just a brief description of the stadium to start with. Well, it was completed in 1995. It has a capacity of a bit more than 51,000 seated spectators. As you can see here, there is a roof with two panels that can be opened and closed. As mentioned before, no HVAC systems, and natural ventilation can occur through a few openings in the stadium facade and the stadium roof. Some main dimensions of the stadium; it has a length of 226 meters, then there is a width of about 190 meters, a height of 72 meters. At the corners of the stadium there are four large openings called the stadium gates that connect the inside of the stadium with the outside. And one of these gates you can see here on the right side. So it's large enough for a truck to drive through, which is indeed needed when you need to move sound and light equipment inside, for example for concerts. Then there are a few ventilation openings. Of course the roof, when it is open, is the largest one. The next largest opening are the four gates actually. Summed they give an opening of 166 square meters. Then there is a very narrow opening between two steel plates near the roof gutter, and although this opening is actually quite narrow, it is present over the entire length of the stadium. So overall, it's also quite a large opening. And then finally, there are some openings between the fixed and the movable part of the roof. So these are the potential ventilation openings. In this study, however, we consider the roof to be closed, and we therefore have to achieve natural ventilation through the other openings. Then there is the urban surroundings, as you can see here, it's not a very densely built area but, the surrounding buildings are high-rise buildings, so they will have an important effect on the wind flow around the stadium. And then in the wider surroundings, we can assess the aerodynamic roughness length by using the Davenport-Wieringa roughness classification for a distance of ten kilometers upstream of the location of interest. So, then the methodology was that we started by assessing natural ventilation both driven by wind, but also by buoyancy, the thermal effects, which during the concerts are very important. We included in this study a large range of length scales from the smallest scales of centimeters, referring to the size of the ventilation openings, to the kilometer scale of the urban area. And then there are three methods that we could use: on-site measurements, reduced-scale wind-tunnel measurements, or CFD simulations. And they each have their advantages and disadvantages. On-site measurements have the advantage that we can measure real conditions and we don't have scaling issues. But we cannot test, of course, future configurations of the stadium. And we only measure in a few positions. Reduced-scale measurements in a wind tunnel offer high controllability of the boundary conditions, but there are serious problems here with scaling. Because if you want to put and the stadium and the urban surroundings in the wind tunnel, we have to scale it down to such a large degree that the small ventilation openings would become so small that they are maybe impossible to manufacture. But also that the flow through these openings might switch from turbulent to transitional or to laminar. Which means we would severely be violating similarity requirements, which is clearly unwanted. In this way, we cannot get accurate results. In addition, simulating thermal effects in a wind tunnel is very difficult and as shown before, also in week two, it's not possible to satisfy in many cases, the Grashof number similarity. Then there are CFD simulations. We can perform them at full scale, so we don't have scaling problems here. We can also assess future ventilation performance. We get information on the whole-flow field, but we need to perform solution verification and validation. So what was the approach followed here, well for the current situation we combined full-scale measurements and CFD simulations. There also the CFD simulations were validated, and then we applied CFD to predict future situations. So starting with the measurements. Measurements were made inside the stadium, but also outside the stadium of air temperature, relative humidity air speed, globe temperatures, CO2 concentration. We also measured the irradiance of the sky in the roof gutter actually. And then let's look at some results. What you can see here are temperatures that were recorded on three consecutive days. And the lowest curve, the dotted curve is the outside air temperature. The four other curves, the higher curves, are the inside temperatures. So what you see here actually during the day, is that the roof is closed, shortwave radiation comes in through the roof, through the transparent, polycarbonate sheets, and is then absorbed by the inside stadium surfaces, emitted as longwave radiation, which cannot pass through the polycarbonate sheets. So you get, a mini green-house effect. So it get's very warm inside the stadium. And you clearly see, that during the day, the stadium air heats up, more rapidly than the outside air. Then, at some point, the solar radiation decreases, so the temperature goes down. Until the moment that the 55,000 spectators come in that emit a lot of heat, so, then suddenly you see the air temperature going up again. And at midnight, when the concert ends, and the people leave the stadium, through natural ventilation you get, actually, the ventilative cooling of the stadium by the outside air. But you also see that the inside air temperatures do not reach the low value of the outside air, also indicating that natural ventilation in the present situation is quite limited. You can also see the effect of the concerts and of the 55,000 people coming in and being present very clearly in this figure that shows humidity ratios, so this is the water vapor concentration actually indicated here. Then we started the CFD study. This is the computational geometry. High detail in the stadium vicinity, less detail for the surrounding buildings. Then the high-resolution high-quality grid was made with only prismatic and hexahedral cells, so not a single pyramidal or tetrahedral cell. A grid-sensitivity analysis was performed. The discretization error was estimated. Here you see a detail of the grid near the roof gutter, which is important, because there we have one of the important ventilation openings. Some other views. This is outside the stadium, so the deck surrounding the stadium. You see the grid and the corresponding photograph. And also the surroundings, the high-rise buildings were taken into account. The computational domain was determined based on the best practice guidelines. And then a set of boundary conditions; logarithmic wind speed profile, appropriate profiles for turbulent kinetic energy, turbulence dissipation rate, and then the correct relationship between aerodynamic roughness length and sand-grain roughness height and also fixed surface temperatures were imposed. And then some additional computational settings, well the 3D steady RANS equations were used, with the realizable k-epsilon model, standard wall functions and the required second-order discretization schemes. Then vaildation was performed and I will not go into detail here, I would directly like to go towards the results. And the results actually will be presented for different ventilation configurations. So, we have the current situation. For example, with this ventilation opening near the roof gutter that is very narrow. And then we look at configuration two and three, where either this top seal plate is reduced by 50%. So, it's slightly reduced, or it's removed completely. And this way, we can rather easily in this stadium enlarge this ventilation opening. So let's see what kind of effect this has on the ventilation, which is quantified in terms of air change rate per hour, which is the number of times that the entire indoor stadium volume is replaced by outside air. These are the values that you can see here for different wind directions, and what you can see on average is that by removing the steel plate, the top steel plate near the gutter, you can increase the air change rate per hour by 43%, which is quite substantial. You also see from this graph that the differences in air change rate are quite large, depending on the wind direction. This indicates the importance indeed of wind direction. As you can see here, these are wind speed ratios made dimensionless, in horizontal planes. For this particular wind direction, you see indeed that the stadium is situated in the wake of the high-rise buildings that are situated upstream. So, they will substantially affect, negatively affect the natural ventilation. Then to analyze the effect of wind direction more in detail and also the effect of the urban surroundings, we performed simulations for these eight wind directions. And the results in terms of air change rate per hour divided by the reference wind speed at 10 m height is what you see here. And for example, you see here that for the 16 degrees wind direction, and for 198 degrees, the difference between those two is almost a factor two. So the conclusion here is that it's very important to take urban surroundings into account. And this should be stressed, because often in wind tunnel measurements, natural ventilation of buildings is assessed by only including the building itself, which clearly can give rise to erroneous results. So that's an important conclusion here: always take into account urban surroundings, and assess ventilation for different wind directions. So let's go back to the module question now. So which is the best method to assess and improve natural ventilation? Well for the current situation, full-scale measurements are recommended. But also Computational Fluid Dynamics simulations can be used. For the future situations, we have to use CFD. Certainly in this case, because wind tunnel measurements are not possible because we would violate similarity requirements too much by scaling down the small ventilation openings. In this module we've learned about the advantages and disadvantages of several methods to assess natural ventilation of a building in an urban area. We've also looked at the important effect of urban surroundings and wind direction on the natural ventilation. In the next module we'll focus on some important characteristics of rain and wind-driven rain, on the importance of wind-driven rain. The parameters determining wind-driven rain, and the complexity of this process. Thank you very much for watching and we hope to see you again in the next module.
