Building Simulation in Design Practice

Course Instructors: Roya Rezaee, Tyrone Marshall, Marcelo Bernal (Perkins & Will Research Team)
Team members: Kiran Golla, Prerana Kamat

This course aims at developing a theoretical and practical understanding of collaborative multi-performance building analysis in design practice. Exploring computational techniques to help designers generate a large space of design options, simulate various building performances, evaluate and explore the options, and make informed design decisions in a systematic framework called Design Space Construction (DSC) that was developed at Perkins & Will. The modeling and simulation cover the following domains: Parametric Modeling, Solar, Energy, Air Flow and Ventilation, Daylighting, View, Cost, Statistical Analysis, and Data Visualization.

799 Broadway, New York

We chose the project 799 Broadway New York designed by Perkins & Will for Normandy Real Estate Partners, to apply the methods of modeling and simulation in the domains mentioned above, to learn and explore the best possible outcome in all aspects.

Precedent Analysis

View Analysis: Line of Sight

The existing design is analyzed in terms of the quality of views the spaces in the building received. A landmark monument and a park nearby were considered as targets for the views. It is observed that 82.52% of the facade area i.e., 79812 sq.m. has a visibility range between 0-20% while the rest 17.48% of the facade area has visibility between 20-40%.

Exterior Air Ventilation & Pressure Study

The wind and pressure simulations are carried out using Computational Fluid Dynamics (CFD) methods based on OpenFOAM® technology and related open-source projects.

Two Analysis planes are considered, viz., a horizontal plane at a height of 1.524 meters from the ground level and a vertical plane cutting NS through the building's center. The horizontal plane aims to study the pedestrian comfort at the ground level around the building, while the vertical plane is to examine the comfort levels at the stepped terraces. The simulation is run for both the summer and winter months separately, and the results are compared.

CFD Air Ventilation Simulation results for both the summer and winter months are shown in two different views. The remarks were made with reference to the Extended Beaufort Scale (here).

• Horizontal Plane: It is observed that the wind speeds are higher along the shorted sides in general and are comparatively higher during the winter months. The direction of the wind is seen changing abruptly at the corners of the building. The wind velocities seem to be compromised at a certain point each on both the longitudinal sides of the building.

• Vertical Plane: Wind speeds are higher as we ascend the building and the instances are almost identical during both winters and summers. The terraces are on the leeward side and Wind turbulences at low speeds could be observed at all the terraces.

CFD Pressure Simulation results for both the summer and winter months shown in two different views.

• Horizontal Plane: The pressure difference is prominent across the lateral side of the building during the Summers when compared to winters. Since the wind direction is almost perpendicular to the lateral side of the building during the summers, the transition of the pressure is abrupt and not gradual.

• Vertical Plane: From the pressure map, it is observed that the pressure difference is increasing with the floor level. The stepped terraces side of the building is at a lower pressure when compared to the opposite side. The pressure differences are almost uniform throughout the year (both summer and winter months).

From the above observations, to maintain the gradual nature of the wind flow direction and to mitigate the pressure difference across the building, the ground level could be made porous by introducing semi-opened spaces across the longitudinal side from the form the shorter edge of the building. This could potentially reduce the sharp changes of pressure and velocities.

A detailed analysis report can be found here.

Limitations: In the interest of time and computational resources, the simulation is run only on the desired study building to exclude its context. As the results are entirely dependent on the building in its context, and while this being an initial exercise in understanding the CFD simulation process, the analysis, and the observations are made considering the above-mentioned conditions.

Useful Daylight Illuminance (UDI)

The UDI is calculated with the following assumptions for Window-to-Wall Ratio (WWR) and Glazing Visible Light Transmission (VLT):
WWR_N: 0.95 | WWR_E: 0.95 | WWR_S: 0.1 | WWR_W: 0.1 | Glazing VLT: 0.15

Energy Use Intensity (EUI)

With the below building input assumptions the total Energy including heating and cooling is 287.10031 kWh_m2_yr.

Objectives Derivation

Geometric Variations Explained:

Precedent’s design was interpreted as amassing that is south aligned 3-stack tapper configuration providing two predominant terraces on the Northside of 5th & 9th levels.
With an interest to only alter one of the three stacks, the center one is chosen as it directly affects the main terraces along with the floor plates. These floor plates are then parametrically redesigned to have varying widths while maintaining the floor area constant at each level keeping the overall built-up area unaffected.

Design of experiments (DOE)

Below analyze the complexity of trade-off across different performance objects.

Objective 01: View

Objective 02: UDI

Objective 03: EUI

Energy Use Intensity (EUI) analysis is run for the 16 sets of combinatorial input parameters, i.e., WWR_N, WWR_E, Glazing_VLT, Stacking.
It is observed that the 13th run produces the best results with the least energy requirement.

Conflicting Objectives & Sensitivity Analysis

Value Assessment

Design Recommendation