Study on Cross-Ventilation Performance of Residences in the Passive Town Kurobe Model Based on Measurements and CFD

The housing complex in the Passive Town Kurobe model was designed for residents to enjoy an environmentally-friendly lifestyle, and cross-ventilation was used to improve indoor thermal environment in hot seasons. To understand and maximize environmental improvements through cross-ventilation, computational fluid dynamics (CFD) simulations and measurements were carried out. It was confirmed that CFD simulation using the modified k-å model reproduced the observations with sufficient accuracy. The turbulence model used in the CFD simulations in this study was the standard k-å model, which incorporates the Durbin limiter. This model had the most accurate correspondence with the measured values. The wind catcher (WC) used to increase the cross-ventilation flow rate was found to be ineffective in the dominant wind direction. Further, it was found that if the room targeted for environmental improvement was on the upstream/downstream side, the flow rate was maximized if all windows in the room were opened.


Introduction
There is increasing interest in passive houses that efficiently use natural energy and are superior in comfort, health, and energy conservation. The 1st and 2nd blocks in the Passive Town Kurobe model were designed to distribute ventilation throughout the entire house using window arrangements and a wind catcher (WC; Fig. 1). The seasonal wind in this area has been considered previously, but the effect of opening specific windows was not clear.
Therefore, based on a comparison with actual measurements, it was necessary to understand the predictive accuracy of ventilation performance via computational fluid dynamics (CFD) analysis and to determine whether opening windows (and which ones to open) would affect ventilation.
①External wind collides with wind catcher ②Increase of wind pressure near the opening ③Increase of inflow air volume ①External wind collides with wind catcher ②Peeling off airflow ③Reduction of wind pressure near the opening ④Increase in outflow volume  Fig. 3 shows a plan view of the target dwelling unit on the 4th floor of the 2nd city block, and Fig. 5 shows the window type used in the room in Fig. 3. For the second block area, the LDK on the south side of the dwelling unit was considered. In the second block, the inflow and outflow openings were selected from window 8-I at the end of the corridor, window 5-F of room 5, window Ⅱ-H of the LDK, window Ⅱ-G and window Ⅱ-H. The difference in opening way was considered. Measurements were made using a wind direction anemometer installed on the roof of the residential building in the 1st block and on the roof of the neighboring company house in the 2nd block. For the indoor wind speed measurement, an omni-directional anemometer was used, and in the 1st block, wind speeds were measured in room 1 and the LDK-Ⅰ; in the 2nd block area, measurements were taken in the LDK-Ⅱ. In addition, the ventilation rate was measured using an ultrasonic anemometer, and the wall pressure difference between the outdoor wall surface near the opening and the indoor wall surface was measured with a differential pressure transducer.

CFD analysis
After analysis (model L), which reproduced 1,000 m 2 of the urban area surrounding the study target building, we analyzed the boundary condition of the reduced calculation area (model S) around the target building via data mapping. The conditions analyzed are given in Table 1 for the 1st block and in Table 2 for the 2nd block. The target wind direction was set to 0° (north), 22.5°, 45°, and 270° based on the summer wind direction of the breeze map for Kurobe (Fig. 6). For the inflow boundary condition of model L, the approach flow measured by the wind tunnel experiment was used.

Table1. First block district analysis target conditions
Table2. Second block district analysis target conditions

Results of Measurements and CFD analysis
We investigated the appropriate turbulence model using the 2nd block. The turbulence models compared included a standard k-ε model (hereinafter referred to as "improved k-ε model"), realizable k-ε model, SST k-ω model incorporating the Durbin limiter and SST k-ω model. The turbulent energy k and turbulent dissipation factor ε obtained in model L were converted to the specific dissipation factor ω and mapped. Fig. 7 shows a comparison of the results of analysis of each turbulence model as compared to actual measurements.

.1 Study of Effectiveness of the WC using CFD
The wind pressure difference acting on the opening is the driving force. Because the WC is a device used to stop the air flowing along the outer wall and to convert the dynamic pressure into wind pressure, the direction of installation with respect to flow is important. Figs. 8 through 9 show the opening wind pressure coefficients according to the presence or absence of the WC in the wind direction called "Ai no kaze" in the Kurobe area, i.e., at 22.5° and 135°.
We first consider the 1st block. When the wind direction is 22.5°, the wind flows from west to east due to the wind reflected from east to west on the windward side and the second downtown area on the leeward side. As is clear, the WC increases the positive pressure on window 1-C up, but it decreases the positive pressure at windows 1-C down and 4-C (Fig. 8). When installing the WC in the opposite direction, it is difficult to improve the ventilation performance because the WC exerts the opposite effect when wind flows in the same side. For a wind direction of 135°, the wind flows from east to west, so the positive pressure increase determined by the WC on window I-C up is remarkable.

Ventilation performance evaluation method
With ventilation, the goal is to ensure that a large amount of natural ventilation occurs by opening windows, such as in the middle of a summer night, thereby allowing residents to directly force air currents and improve the environment by cooling the indoor air. With regard to air currents, the indoor airflow velocity is a standard, but in order to cool the indoor air, it is necessary to consider that the ventilation effect decreases as hot air on the upstream side of the ventilating room flows to the downstream side. Therefore, when assuming a constant contamination occurrence per unit volume in the entire ventilated housing space, the amount of effective ventilation obtained by dividing the contaminant generation amount of the target area by the average concentration of the area is converted into an effective ventilation frequency and evaluated (Table 3).

Ventilation performance evaluation method
We consider the second district. Table 6 shows the results of ventilation performance for a wind direction of 22.5°. Regarding the LDK, in case 1 where only the LDK south window was opened, the amount of ventilation is poor. However, opening the northern corridor window at the same time as in case 2 was effective, and in case 3, the effective ventilation frequency of the LDK reached a maximum. This is considered to be due to the influence of pollutants generated from the room when opening the window of the other room as the upstream side in the following cases. In case 6, in which all windows are opened, the total ventilation frequency was a maximum, but the average wind speed of the LDK and effective ventilation frequency were not optimal. Rather, the average wind speed increased in cases 7 and 8 where window 7-J, which is on the leeward side, was closed. In case 9, the average wind speed in the south became larger in case 10. This is thought to be due to the distribution of ventilation airflow to the south side of the LDK, which decreases because the ventilation that has flowed through the corridor changes greatly toward the eastern side when window II-H is opened. For room 5, the total ventilation frequency reached a maximum in case 11, but the average wind speed and effective ventilation frequency did not reach a maximum. The average wind speed and effective ventilation frequency were larger in cases 12 and 13 where the north windward opening was reduced. In this case, back flow occurred in all LDK south windows except II-GI and 6-I, and the average wind speed and effective ventilation frequency reached a maximum in case 16 in which this window was closed. With regard to room 6, in cases 18 and 19 in which the LDK leeward window was opened, and in case 17 as well, the average wind speed and effective ventilation frequency both decreased. In case 22, the average wind speed and effective ventilation frequency reached a maximum. This is considered to be due to the addition of the inflow opening as compared with case 17. Regarding room 7, the average wind speed and effective ventilation frequency were at a maximum in case 29. This is because we closed window 6-I other than window II-GI of the LDK South window where a reverse flow had occurred in case 28.

Conclusion
The following results were obtained from this research.
(1) The turbulence model used in the CFD in this study was the standard k-ε model incorporating the Durbin limiter, which had the best correspondence with the measured values.
(2) In Passive Town Kurobe, the WCs are not necessarily installed effectively in the first and second blocks.