Juliette Zidek | Portfolio of Selected Work 2016-2020 by Juliette Zidek

CONTACT INFORMATION NAME:

JULIETTE NORA ZIDEK

ADDRESS: 3161 N. HUDSON AVE APARTMENT G2 CHICAGO, IL 60657 EMAIL:

juliettezidek@gmail.com

PHONE: 847.400.6292

PORTFOLIO CONTENTS

PROFESSIONAL WORKS 4 12 20 26 30

THE REED AT SOUTHBANK KENECT PROJECTS ACAW 2020 RIVER BEECH TOWER IDEAS LAB PALACE QUARTER BEIJING

PERSONAL WORKS 34 40 46 50

PLACE OF DIVERSITY [PODS] BIO LUM ACADIA WORKSHOP HEX VAULT

For a complete portfolio of academic design work by Juliette Zidek, please visit https://issuu.com/juliettezidek

THE REED AT SOUTHBANK

2019-2020 / Professional Work Perkins and Will Location: Chicago, IL Client: Lendlease Architectural Design Team: Bryan Schabel, Design Principal Carl D’Silva, Managing Principal Greg Tamborino, Senior PA Justin Wortmann, Project Architect Juliette Zidek, Designer III Samantha Lopez, Designer III Nic Hnastchenko, Designer III Hector Reyes, Designer III Nico Hsu, Designer I Julian Roman, Designer I

The Reed is a 650,000 sf multi-family residential high-rise building slated to be constructed in the South Loop of Chicago. It is the second of four buildings to be designed as a part of the Southbank masterplan, a collection of multi-family projects organized around a central park between Harrison and Polk St. on the east bank of the Chicago River. Lendlease is both the client and general contractor for the project, as well as the owner of the entire Southbank site in the South Loop. The design teams’ vision from the start was to develop an amenity driven residential project on par with the original award-winning building of the Southbank masterplan, the Cooper, which was designed by Perkins and Will and constructed in 2018. Like the Cooper, the Reed includes both rental and condominium units, but differs in its orientation and relationship to the river. The design team worked to develop a unique aesthetic and identity for the Reed that relates to the solidity and industrial nature of the historic masonry building stock in the Printer’s Row neighborhood. As is the nature with developer driven residential projects, the design goals for the project were ambitious but the budget was tight. After the completion of one schematic design phase in late 2019, the client requested significant design changes in order to increase the percentage of sellable area to gross building area - a necessary adjustment to make pursuing the project financially lucrative. Additionally, in early 2020, the client requested a thorough value engineering re-design effort for both the residential tower and podium facades. The following pages describe the computational

design methodology that I employed to manage the tower facade re-design effort. I ended up starting from scratch in a lot of ways, because the original facade design lacked the framework to evolve appropriately with the constantly changing floor plans and varying locations of demising walls and heat pump enclosures at each residential tier. My initial goal for the value engineering exercise was to include necessary performance targets such as window-towall ratio and pre-defined constraints in the design logic from the beginning, so that the facade patterning would generate from these factors instead of having to check them retrospectively. I used Grasshopper to write a script following this design logic, with the goals of minimizing vertical mullion count and the total number of construction modules. Additionally, the script was designed to export specific metrics requested by the client, and to allow manual manipulation with live feedback for a design workflow with the Design Principal of the project, Bryan Schabel. In the end, the redesign effort was successful in substantially lowering the cost of the residential tower facade, with a visual expression that pleased the client and design team. Project Role: In addition to facade design, my role on the project from Schematic Design through Construction Documentation consisted of the following responsibilities: programming and area studies; core and shell design, analysis, visualization and detailing; coordination with structural and mechanical engineers; code analysis and compliance documentation; design documentation; and BIM management.

TOWER FACADE VALUE ENGINEERING PSEUDOCODE 1. Given: Overall Target % Glazing = 60% Overall Target % Solid = 40%

TIER 1

TIER 2

Calculate adjusted Target Glazing % and Target Solid % per floor, taking into account pre-defined glazing at west facade and balconies. Predefined Glazing Area Predefined Solid Area Undefined Area

2. Divide facade length into 2’- 2.5” modules, based on width of structural bays. Calculate Target No. Glazing Modules and Target No. Solid Modules per floor from target percentages. Assign solid modules to columns and heat pump modules at perimeter. Subtract number of pre-assigned solid modules from Target No. Solid Modules to arrive at new target. Assigned Solid Modules 3. Split facade length at room demising partitions to find % facade length of each room. Calculate no. solid modules per room by multiplying % facade length of each room (minus pre-assigned solid modules) by Target No. Solid Modules. 3b. Sum no. solid modules of each room and compare to Target No. Solid Modules. If total surpasses Target, remove the difference of modules from rooms with highest % solid. If total is less than Target, add the difference of modules to rooms with lowest % solid. 4. Randomly generate thousands of options of facade patterns by shuffling the solid and glazing module order within each room facade length. To meet pre-determined design criteria, remove options with: • more than 3 glazing modules in a row (aesthetic preference) • more than 3 solid modules in a row (aesthetic preference) • double glazing split between rooms (reduces number of mullions) 5. Sorting: First sort each remaining option by mullion count per floor. Secondarily sort options with a ‘panel ganging’ caclulation to reduce the number of total construction modules. Adjacent 2’-2.5” modules can be lifted together during construction, reducing construction time. When optimal options are tied for mullion count, module count is used to determine the preferred pattern.

>3 GLAZING MODULES

>3 METAL MODULES

SPLIT DOUBLE MODULE

MINIMIZE SINGLE MODULES

GANG ADJACENT SINGLE MODULES

DISSOLVING OPTION This option was conceived to reduce the amount of metal panel, while also reducing mullion count. Three optimal patterns are selected per tier and alternate through each tier for visual variation. Solid panels are assigned as either metal panel or spandrel glass, with the % of spandrel glass in proportion to the % of metal panel increasing over the height of the building.

METRICS

METAL SHIFTED OPTION This option evolved from the Dissolving scheme and returns the tower facade to its original expression of solidity, with a newly rationalied patterning. It’s expression was also influenced by the work of Anni Albers, a modernist textile artist and printmaker.

This option was conceived at the most rational aesthetic, with an emphasis on the vertical expression of each tier. The most optimal pattern is selected per tier and stacks all the way through.

Three optimal patterns are selected per tier and alternate through each tier for visual variation.

The reduction in vertical mullions and total construction modules led to the selection of this option based on cost savings.

METRICS

NO. VERTICAL MULLIONS: 4,518 REDUCTION FROM ORIGINAL COUNT: 13.7%

NO. VERTICAL MULLIONS: 4,315 REDUCTION FROM ORIGINAL COUNT: 17.6%

NO. SINGLE MODULES VISION GLASS: 696 NO. SINGLE MODULES SPANDREL: 478 NO. SINGLE MODULES METAL PANEL: 1,150 TOTAL NO. SINGLE MODULES: 2,324

NO. SINGLE MODULES VISION GLASS: 696 NO. SINGLE MODULES SPANDREL: 109 NO. SINGLE MODULES METAL PANEL: 1,115 TOTAL NO. SINGLE MODULES: 1,920

%VISION GLASS: 59.7% % SOLID: 40.3%

RENDERINGS BY PERKINS AND WILL

METAL STRIPES OPTION CHOSEN SCHEME

METRICS NO. VERTICAL MULLIONS: 4,281 REDUCTION FROM ORIGINAL COUNT: 18.2% NO. SINGLE MODULES VISION GLASS: 652 NO. SINGLE MODULES SPANDREL: 112 NO. SINGLE MODULES METAL PANEL: 1,086 TOTAL NO. SINGLE MODULES: 1,850 %VISION GLASS: 59.9% % SOLID: 40.1%

KENECT PROJECTS

2017-2019 / Professional Work Perkins and Will Locations: Phoenix, AZ Cleveland, OH Client: Akara Partners Architectural Design Team: Todd Snapp, Design Principal James Giebelhausen, Senior PA Aaron Amosson, Project Architect Lynette Klein, Senior IPD Juliette Zidek, Designer III Rob Deering, Designer III

Akara Partners came to Perkins and Will in 2017 with the goal of developing a new type of rental housing for young, transient professionals. They aimed to create a network of amenity-driven residential towers across the United States, conceived as ‘social clubs with residences above’ that offer young city-dwellers flexibility through app-based security, short-term subletting and leasing, and a national membership platform. Starting with Nashville as a pilot project for the Kenect brand series, Perkins and Will designed numerous buildings to various stages of compeltion in the following cities: Nashville, Phoenix, Denver, Cleveland, and Miami. Perkins and Will also worked with Akara Partners to develop the Kenect Brand Standards for everything from site selection and design; building organization and programming; unit mix, size and layout; structural and MEP standards; and member amenity space design. With the design of amenities, social spaces, and opportunities for connection as driving priorities for the Kenect projects, the building materials and enclosure systems were value engineered from the start to ensure efficient and affordable design and construction. However, every city required different design standards to be met that introduced unique characteristics to every Kenect building in massing, materiality and relationship to site. Parametric tools were utilized in the process of considering enclosure options for each project, with marketability, cost, and building performance as primary aesthetic drivers. Consistent pricing exercises influenced design direction, and it became the design team’s responsibility to think creatively with cost as a constraint from the beginning.

Project Role: I was a design team member for Kenect Phoenix and Kenect Cleveland from Conceptual Design through Design Development, and helped on all of the other projects intermittently. Kenect Phoenix is currently under construction, with an expected completion date and occupancy in the spring of 2021. My roles on Kenect Phoenix and Cleveland included the following responsibilities: initial massing studies; programming and area calculations; core and shell design, analysis, visualization and detailing; coordination with structural and mechanical engineers; code analysis and compliance documentation; design documentation; and BIM management. The next pages depict a computational tool that I developed from the experience of having worked on several Kenect projects, with the intention of streamlining further work with Akara Partners and aiding future multi-family residential work with other developer clients. As Perkins and Will helped Akara Partners to distill their buildings down to particular pro-formas and building metrics, the initial design exercises took on a scientific rigor. With an understanding of the most critical parameters for efficient residential tower programming, I used Grasshopper to develop a tool that would automatically generate residential tower massing and planning for early design phases.

KENECT SCRIPT

26' - 9"

Each Kenect project begins with a feasibility study that helps Akara Partners determine whether or not to pursue the entitlements process for a specific site. Typically, Akara Partners requires confirmation that a target number of residential units, amenity area and parking ratio can be met before investing in further design efforts. These initial planning exercises often lead to the development of multiple residential tower iterations, with metrics that need to be quickly evaluated to provide Akara Partners with an optimal layout.

27' - 0" 26' - 9"

FOLD-DOWN QUEEN BED & TABLE

BATHROOM 8'-2" X 6'-0"

BEDROOM 10'-2" X 10'-1" CLOSET 7'-0"L

LIVING AREA 13'-0" X 15'-7"

12' - 3"

13' - 6"

W/D

MECHANICALLY SLIDING STORAGE UNIT

LIVING AREA 17'-4" X 13'-0"

LIVING AREA 17'-5" X 11'-4"

FOLD-DOWN QUEEN BED

FOLD-DOWN TABLE

W/D

BATHROOM 5'-10" X 8'-8"

STUDIO | 360 SF

Because the Kenect projects all share the same target unit areas, target amenity areas, unit mix ratios, and parking ratios, a common logic can be applied to every project in the initial planning process. I found through outlining the design logic, I could then write a script to automate the planning process. The Grasshopper script becomes useful as the target number of units and physical constraints of each building vary by site and the same design logic must be applied to new conditions.

26' - 9"

MOBILE ISLAND

KITCHEN 13'-0" X 7'-7"

FOLD-DOWN QUEEN BED & TABLE

46' - 3"

BATHROOM 8'-2" X 6'-0"

26' - 9"

BATH 8'-8" X 12' - 3"

W/D

MECHANICALLY SLIDING STORAGE UNIT WALK-IN CLOSET 6'-7" X 5'-10"

LIVING AREA 17'-4" X 13'-0"

W/D

3/16" = 1'-0"

MOBILE ISLAND

LIVING AREA 12'-11" X 10'-0" KITCHEN 13'-2" X 11'-2'

CLOSET 4'-8" L

CLOSET 7'-0" L

27'D MODULE - STUDIO A AND B

UNIT MIX

W/D

CLOSET 5'-5" L

CLOSET 4'-9" L

21' - 0"

BATHROOM FOLD-DOWN 9'-0" X 5'-8" QUEEN BED

FOLD-DOWN BUILDING TABLE LIVING AREA CORE 17'-5" X 11'-4"

BEDROOM 9'-8" X 9'-2"

BATHROOM 8'-8" X 5'-10"

BEDROOM 10'-0" X 10'-0" W/D

The unit layouts depicted on the right and the parameters listed below are common to all Kenect projects:

RENDERED_27' MICRO STUDIO

27'D MODULE - MICROSTUDIO A 3/16" = 1'-0"

BEDROOM 9'-7" X 9'-6"

BUILDING CORE

BATHROOM 8'-6" X 5'-2"

STUDIOS: 60% ONE BEDS: 30%

ONE BED | 560 SF

THREE BED | 1200 SF

THREE BEDS:10% FLOOR PLATE EFFICIENCY: 82-85%

26' - 9"

PARKING RATIO: 75% min. KITCHEN 13'-2" X 11'-2"

MOBILE ISLAND

LIVING AREA 13'-3" X 10'-2"

COWORKING FACILITIES: 20,000 SF + 10 SF/unit

W/D

BATHROOM 8'-8" X 8'-4"

BEDROOM 10'-0" X 10'-2"

PROGRAM INPUTS

22' - 9"

WALK-IN CLOSET 5'-10" X 6'-10" COOKTOP DW

GEOMETRY INPUTS CLIENT NAME The script requires the centerline of the tower and the outline of the podium to run, input as curves. REPORT

W/D

COMPACT REFR. COOKTOP

11' - 0"

STUDIO A - 250SF 3/16" = 1'-0"

STUDIOS, MICROS, CONV. 06/11/18 1 BEDS 06/13/18

The following metrics are reported with each run of the script and are depicted through the script interface: • Typical Unit Mix • Required Building Height • Number of Residential Floors • Number of Units/Floor (for a typical floor) • Number of Units/Floor (for the atypical floor) • Typical Floor Plate Efficiency • Amenity Area RHINO VIEWPORT

COMPACT REFR.

PROJECT NAME CLIENT NAME

The following inputs are required by the script to generate a prototypical residential tower:

• Target Number of Units • Target Floor Plate Efficiency • Residential Bay Depth • Corridor Width • Floor to Floor Height • Target Amenity Area/Unit PROJECT NAME

11' - 0"

OUTDOOR AMENITY AREA: 10-15 SF/unit

SCRIPT INTERFACE WITH HUMAN UI 21' - 0"

INDOOR AMENITY AREA: 40-50 SF/unit

DESIGN LOGIC

UNITS SHIFTED MANUALLY

DEFAULT UNIT LAYOUT | SYMMETRY OPTIMIZED

33' - 0"

1. SYMMETRY IN PLAN IS OPTIMIZED BY DEFAULT (BUT CAN BE ADJUSTED) CLOSET 13'-10" L

Symmetry in plan increases the regularity of the structural grid which benefits the tower design in several ways. Structural regularity generally reduces project cost and can increase the ease and speed of construction. It also benefits parking design and layout, providing flexibility in drive aisle location and maximizing the number of parking spaces. BATHROOM 6'-8" X 8'-11"

BEDROOM 10'-0" X 10'-7"

BATHROOM 8'-8" X 5'-10"

BATHROOM 8'-1" X 5'-8"

CLOSET 5'-8" L

W/D

BEDROOM 9'-11" X 10'-2"

CLOSET 2'-10" L

However, perfect symmetry is not always possible or preferable in some instances. While the script initially optimizes for plan symmetry, there are built-in controls to shift clusters of one bed units around the plan to satisfy other constraints. CLOSET 5'-9"L

BATHROOM 5'-10" X 8'-9"

47' - 3"

ROOM X 6'-8"

BEDROOM 9'-9" X 9'-10"

2. THREE BEDROOM UNITS ARE ALWAYS PLACED AT TOWER CORNERS

SECOND EXIT STAIR

CLOSET 7'-0" L

DEFAULT CORE PLACEMENT

The typical layout of three bedroom units requires that there is sufficient glazing area for three bedrooms and one living space A ANDonB the perimeter of the floor plate. Placing three bedrooms at the tower corners benefits the floor plan in the following ways:

BUILDING CORE

THREE BED AT CORNER

LIVING AREA 21'-2" X 14'-10"

a. A corner three bed unit is preferable to a linear three bed unit because circulation area is minimized, making the unit more efficient, compact, and communal in nature. b. Higher quality views with wider view angles are distributed equitably to units with more occupants. c. Corner one bed units and studio units often nest in inefficient ways and may require a greater amount of spandrel panels due to MEP equipment. Corner three bed units maximize the percentage of vision glazing at tower corners and therefore provide more value than smaller units in the same place. 3. THE ELEVATOR CORE IS PLACED ON A LONG EXTERIOR WALL, THE SIDE DETERMINED BY SYMMETRY OPTIMIZATION. THE SECOND EXIT STAIR IS PLACED ON THE OPPOSITE SHORT WALL.

A typical elevator core is assumed to contain 3 elevators and 1 exit stair. The size of this core (generally 32’ wide by the depth of the residential units) determines that it must be located on a long exterior wall. While the core is initially placed at a corner, the script is designed to give preference to three bedroom units on the same side. The core can be moved inward from the end of the building with controls built into the script. If the core placement exceeds the maximum common path of travel or dead end distance allowed by the IBC, the geometry turns red, visually indicating noncompliance.

3 BEDS 06/20/18 3/16" = 1'-0"

The second exit stair core is placed on the opposite side of the tower from the elevator core, with it’s short side flush with the 3/16" = 1'-0" short exterior wall, centered on the corridor. This placement maintains plan symmetry while avoiding disruption to the main REALLY CRAZY LONG facades of the building. 4. UNITS ON THE AMENITY FLOORLONG ARE ALWAYS REALLY CRAZY PROJECT CLUSTERED AND OPTIMIZED FOR TARGET UNIT MIX BY DEFAULT (BUT CAN BE ADJUSTED) The amenity floor hosts the remaining number of the units needed to achieve the target unit count. This floor is designed so that the units always form a cluster and there is a distinct residential zone that can be separated from one (or two) amenity zones. The default amenity floor layout is optimized to meet the target unit mix, but built-in controls allow other configurations to be explored. The metric by which the amenity unit mix percentage is evaluated is defined as Δ UNIT MIX, a sum of the deviation of the floor unit mix from the input target percentages. The amenity floor geometry is colored to reflect whether Target Amenity area/unit is met. Red geometry and text in the Report indicates that the target amenity area is not being met.

CORE SHIFTED MANUALLY

DEFAULT CORE PLACEMENT

PROJECT STATUS

STATUS AMENITY OPT. 1 Δ UNIT MIX = 1.22 AMENITY AREA = 8,902 sf

AMENITY OPT. 2 Δ UNIT MIX = 1.69 AMENITY AREA = 8,637 sf

AMENITY OPT. 3 Δ UNIT MIX = 1.83 AMENITY AREA = 8,497 sf

AMENITY OPT. 4 Δ UNIT MIX = 2.25 AMENITY AREA = 9,510 sf

AMENITY OPT. 5 Δ UNIT MIX = 2.61 AMENITY AREA = 6,523 sf

AMENITY OPT. 6 Δ UNIT MIX = 2.67 AMENITY AREA = 8,092 sf

PODIUM FACADE DESIGN A folded metal panel system was designed to shade the parking garage of the tower podium, while also providing natural ventilation and creating a street facing facade that relates to the geologic nature of the surrounding mountain terrain. The studies below show various scales and configurations of the folded panels, which were studied parametically using Grasshopper and Rhino. The facade was rationalized into a system using 4 typical panels that are folded in two locations along their length. The number of folds and resulting panel depth were studied to reduce the need for a strut at the furthest point of projection, thereby reducing cost. The variation in fold location was found to be acceptable in its impact on fabrication time and cost. The selected facade design is a randomized allocation of the 4 panel types in equal percentages across each building elevation. It is highlighted below and depicted in the building renderings.

021275.000-AP -K enect_P hoenix_detached.rvt

3/30/2018 1:35:52 P M

021275.000-AP -K enect_P hoenix_detached.rvt

3/30/2018 1:35:52 P M

PANEL PATTERNING STUDIES

ACAW 2020

2019-2020 / Professional Work Perkins and Will Architectural Ceramics Assembly Workshop Host: Boston Valley Terra Cotta Architectural Design Team: Mario Romero, Digital Practice Angelica Palecny, Interior Designer III Juliette Zidek, Designer III Ted Hogan, Designer II

Perkins and Will was invited to participate in the Architectural Ceramics Assembly Workshop hosted by Boston Valley Terra Cotta in the Fall of 2019. The workshop’s objective is to have designers consider the properties of terra cotta earlier in the design process and to develop research and design models between manufacturing and architectural industries useful to the efficient production of high-performance facade solutions. Through pre-design and prototype development, teams’ explorations include the use of new digital tools in the production of terra cotta assemblies, the development of unitized façade systems, and how the variable materiality of terra cotta (throughbody color, finish and glaze) can inform and enhance a façade’s performance. Supported by the Digital Practice Group of Perkins and Will, a group of young designers at the Chicago office teamed up to participate in the 2020 workshop. What first inspired our group’s design studies was learning about the history of terra cotta and its earliest use in architectural applications. Through research we found that in addition to being used in sculptures, bricks, tiles, and ornamental objects, terra cotta has been used to transport water for thousands of years. The first terra cotta pipes came into use around 4000 BC, and because of its unique resistance to the corrosiveness of acidic waste water, some of these ancient plumbing systems still exist today. With this historical relationship between water and terra cotta in mind, our team framed our design studies around the question:

‘What are the limits to controlling the flow of water across a vertical façade system?”

Our team studied how a terra cotta façade system could be designed to control, harvest and recycle rain water. Our facade system was designed to act as an inverted system of pipes, exposing the flow of water to view and celebrating its delayed descent to the ground. The 3-pointed facade module that comprised this system was sculpted with ridges to reduce weight in key areas (determined from structurally analyzing the module) and to guide water in, out, and across the surface of the modules. After finalizing the design of the module, our team worked with Boston Valley to prepare it for slip casting and fabrication at the workshop. Typically, the workshop culminates in a weeklong event, with teams constructing full scale prototypes and presenting their work to other participants and peers in the industry. Due to the COVID-19 pandemic, the workshop became a virtual conference with all teams preparing a digital gallery of their work and presenting their design process to all in attendance. Project Role: Our team worked collaboratively through all stages of the project. The initial design direction was decided collectively, with each team member modeling unique iterations of the module. Once a final form was selected, I used SAP2000 to analyze the stress within the module under self-weight and an applied wind load, which informed where material could be removed. Mario modeled the final module massing with Rhino SubD and organized the virtual exhibit, Ted focused on detailing the structural support system, and Angelica led the team in developing a glazing and resist strategy. We all produced drawings for the exhibition.

VIRTUAL EXHIBITION SPACE

MODULE EVOLUTION

DESIGN

WATER FLOW

GLAZE RESIST

MORPHOLOGY

FABRICATION

RIVER BEECH TOWER IDEAS LAB

2019 / Professional Work Perkins and Will Location: Chicago, IL Host: IDEAS LAB Architectural Design Team: Todd Snapp, Design Principal Juliette Zidek, Designer III Gil Song, Designer III Axel Olson, Designer I

Over the past three years, Perkins and Will has led a design research initiative investigating the potential of tall timber construction in collaboration with the University of Cambridge and Thornton Tomasetti. This project resulted in the conceptual design of River Beech Tower, a proposal for the tallest mass timber skyscraper to date, that would radically transform standard methods of sustainable material production, assembly, construction, and urban living. The RBT research initiative began with the question: If we design a high rise building out of timber, what form should the structure take? For the 2019 Ideas Lab, Perkins and Will chose to pose this question to a broader audience and engage the public in the generation of new ideas through education and interaction. Our goal for the lab was to shift the focus of the research project toward the modular application of new forms and typologies, using public input as inspiration for further design development.

The RBT Ideas Lab was structured in several parts to engage the public in various ways. First, a lecture was given to introduce the RBT research and the case for building tall in wood. Then participants were introduced to several unit types through virtual reality tours and asked to complete a survey based on their preferences for living in the tower. The survey cards asked participants about their preferred amenities, location in the tower, and to list any preferred neighbors (if they had come to the lab with a friend or colleague.) The surveys acted as a tool to elicit unique responses about new typologies, while also collecting information for a computational design exercise.

Once their surveys were collected, the results were entered into a live excel document, which was then read from Grasshopper to run a sorting algorithm using Galapagos as an evolutionary solver. Participants were able to watch the solver work for the duration of the lab exercises, running in the background as they learned and explored. Just before the conclusion of the lab, the solver was terminated and the results of the algorithm were presented. Individual participants were able to see where they were placed in the tower, as their assigned units were tagged with their name. The units were also shaded according to fitness, with lighter units in closer proximity to their stated preferences than darker ones. Project Role: I helped organize the RBT Ideas Lab program, while also directing and executing the computational design exercise. I wrote the Grasshopper script used for the lab, entered all data during the event, ran the solver and presented results at the end. Gil and Axel were responsible for the RBT unit development and virtual reality exhibition of the tower and units.

PALACE QUARTER BEIJING

2016 / Professional Work Thornton Tomasetti Client: Will Bruder Architects Analysis Team: Edward Peck, VP Nicole Peterson, Project Consultant Juliette Zidek, Design Intern

This work reflects my collaboration with Nicole Peterson, the sustainability consultant for the Facade eEgineering group at Thornton Tomasetti. As consultants for Will Bruder Architects, the Facade Engineering team developed a collection of sustainable strategies for the design of the Beijing Yihe International Business Park (called Palace Quarter Beijing,) a prospective commercial campus to be located in the Haidian district. These strategies ranged from site design guidelines to building scale studies; specifically providing passive strategies for heating, cooling, natural ventilation and enhancing pedestrian comfort. Beijing, located at a latitude of 39°N, has a monsoon-influenced humid continental climate with hot summers and dry, windy, cold winters. The climate is somewhat extreme, and is both heating and cooling-load driven. The site strategies diagrammed to the right were included in the final sustainability narrative submitted to Will Bruder architects and were developed with the goal of achieving LEED Platinum accreditation. Our team used the Ladybug and Honeybee suite of GH plugins to generate solar altitude angles, wind roses and psychrometric information throughout the year that informed the development of horizontal shading options, building orientation studies, and wind access strategies.

A method for shading the east and west faces of the non-shaded buildings with vertical fins was proposed, but the architect wanted to understand what variety of design opportunities might be available with this strategy. With advisement from the Facade Engineering team, I scripted, modeled

and ran a series of analyses to generate a variety of vertical fin configurations. These studies examine the effect of vertical fin arrangement, sizing, and angle on the levels of annual incident solar radiation, daylight autonomy, and heating and cooling loads in an office environment of selected buildings from the master plan. A 10m x 19m, one story module was used for the west and east facing vertical fin shading analyses. The analysis models extend halfway through the building to analyze the full depth of daylight penetration. The diagrams on the following page show the analysis module in context. The radiation, daylight, and energy analyses were conducted to examine 240 combinations of spacing, angle, width, and depth parameters. The geometric parameters used in this study are listed and diagrammed at the bottom right, while the metrics, targets, and assumptions used in the analysis are detailed below. Average annual solar radiation, daylight autonomy, peak heating, and peak cooling loads of each of the 240 vertical fin iterations were recorded and then analyzed with Thornton Tomasetti’s Design Explorer. This web-based tool allows the user to set boundaries for analysis results, thereby facilitating the optimization of several criteria at once. The graphs to on the following pages show the process through which vertical fin iterations were selected based on their performance. The north façade radiation values for the Beijing climate analysis were used as a target for the vertical shading systems with an allowable range of 300-550 kWh/m2.

PLACE OF DIVERSITY [PODS]

2020 / Personal Design Work Phil Freelon Design Competition Shortlisted Entry The Design Leadership Council of Perkins and Will hosts an annual design competition for all staff (except those currently holding design leadership positions) to enter. The 2020 design competition was named in honor of the late Phil Freelon, architect and owner of the Freelon Group prior to merging with Perkins and Will. Phil is remembered as a lifelong champion of beautiful, democratic design and the competition was designed with these ideologies in mind. Team: Katie Coffey Amanda Ko Juliette Zidek

The brief for the 2020 competition asked participants to design a housing solution in a co-living model that would incorporate concepts of shared economy, social networking, and collaboration, while also increasing density above current zoning requirements in one of three sites. Our team chose to locate our project on the Denver site because of its close proximity to parks and community green space; cultural and commercial amenities in the downtown area; and its scenic views to nature; all of which we considered important to the development of a healthy, dense, diverse co-living project. Fundamental to the programmatic organization of our co-living project is the idea that group living, currently referred to as co-living, is actually an ancient model of habitation. In fact, anthropologists suggest that humans have lived in communal bands ranging in size from several dozen people to several hundred, for at least tens of thousands of years. Even looking to more recent human history, precedents of group living are found in many forms, from convents and dormitories to communes and senior living communities. Group living then, can also be understood as a natural model of habitation that offers the kind of social connection essential to human health and survival, indiscriminate of age, wealth, gender or race. However, the current co-housing trend in urban areas is largely targeted toward one particular demographic—young, single millennials and gen-Z’s. This may make financial sense to developers who can sell the social benefits and affordability of coliving to debt-stricken, transient (and often lonely) young professionals, but this narrow

focus on one demographic not only excludes other equally isolated, cost-burdened groups from the financial and social benefits of co-living, but also misses an opportunity to create more inclusive, diverse, trusting and healthy communities in urban areas. With this in mind, our team’s driving thesis for our proposal became, How can the current model of co-living evolve to offer a greater diversity of inhabitants the same benefits of communal housing? Our answer to this question was to design a new inclusive model of co-living that is both multi-generational and supportive in nature so that people of all ages and living circumstances are able to live together in a mutually beneficial, symbiotic way. The design of PODS began at the scale of the living units, both programmatically and formally, as each unit was designed to accommodate diverse demographics while also providing improved views, sunlight, and social connections for all. The living Pods are both adaptable and efficient, prioritizing quality of life and affordability. Project Role: The competition effort was structured in an egalitarian manner wtih each team member taking on roles that were essential to the development of the design. I designed and modeled the residential tower and living pods with Grasshopper and Rhino, and rendered exterior perspectives with Lumion. Katie took on unit planning and photoshopping of the perspectives, while Amanda modeled the podium of the tower and managed the competition board layout. Each team member contributed to the design of the whole, with each piece of the project evolving in synchronicity.

BIO LUM

2020 / Personal Design Work Urban Confluence Competition Launched in 2019, this international open ideas competition invited people from around the world to imagine a new iconic landmark for the city of San Jose, CA. Team: Wyatt Beard Katie Coffey Juliette Zidek

Bioluminescence has an almost inexplicable way of capturing human attention and imagination. Countless childhood memories recall the magical dancing of fireflies during precious summer evenings. Tourists trek to the remotest areas of the earth to witness the spectacular illuminance of glow worms in the underground caves of New Zealand and the seasonal algal blooms on shorelines from China to California. Foxfire, a bioluminescent species of fungi, has fascinated forest-goers for nearly a millennium, often used in place of candles as a navigating light source in the night.

bioluminescent organisms. For research groups and academics, BIO LUM will act as a physical hub for observation, collaboration, information exchange, and experimentation. For the residents of San Jose, BIO LUM will serve as an educational, civic, and cultural center where visitors can immerse themselves in the enchantment of the natural world.

Beyond the compelling beauty of bioluminescent organisms, however, is their value as a latent resource for understanding and developing sustainable models of illuminance. There is a growing body of research dedicated to the engineering of bioluminescent flora through various means, from embedding nanoparticles in plant leaves to inserting specific fungal genes into plant DNA. This research suggests society is at the cusp of a lighting revolution - soon plants will illuminate our homes, businesses, and streets, a shift that has the potential to greatly reduce the energy demands of urban infrastructure around the world.

The BIO LUM submission proposes a continuous, vertical ramping promenade supported by a timber exoskeleton and bioluminescent scaffold. In a methodology similar to the fire rating of mass timber with sacrificial layers, the BIO LUM structure is proposed to be pressure treated in a nonuniform manner to allow for the controlled decay of the outer layers. In a process illustrated in the following pages, the outer layers of the timber structure are designed to retain moisture at a higher rate than the inner layers. Introducing a specific species of Fungi to the moist wood is proposed for the outer layers to decompose and produce ‘foxfire’, a bioluminescent glow resulting from the oxidation process. New and advanced techniques in the digital machining of wood are proposed to achieve this end, in the creation of an iconic tower that truly glows.

With these biotechnological advancements in mind, BIO LUM proposes to build upon San Jose’s identity as a hub of innovation and ingenuity, while reconnecting to the city’s agricultural roots and relationship to the surrounding natural ecosystems of Silicon Valley and the Bay Area. BIO LUM, the world’s first bioluminescent conservatory, will serve as a center for biotechnology research, development and exhibition pertaining to the engineering of

Project Role: I planned and organized the competition submission, with notable contributions from Katie Coffey and Wyatt Beard in photoshopping the renderings for the boards. I scripted the entire design of the timber structure and ramps in Grasshopper, allowing large scale changes to matriculate to detail level rapidly. I was responsible for the modeling, rendering, drawings, diagrams and writing for the final submission.

ACADIA WORKSHOP

2019 / Workshop Participation Architectural Geometry and Machine Learning Mashup Instructors: Alicia Nahmad (R-Ex) Vishu Bhooshan (ZHCODE) Federico Borello (ZHCODE) Shajay Bhooshan (ZHCODE)

While attending the 2019 ACADIA conference, I participated in a workshop taught by computational designers at Zaha Hadid Architects and Architecture Extrapolated (R-Ex.) The goal of the workshop was to explore synergies between machine learning and the generation of architectural geometry. Due to the increasing digitalization of the AEC industry, the field of Architectural Geometry now commonly integrates structural and fabrication constraints within the shape design process. Complex material and machine constraints are abstracted as intuitive geometric parameters, making geometry based reasoning insightful, didactic and designer friendly. To play with data-driven design workflows, participants in the workshop used the Force Density Method to generate efficient forms of compression only structures. We then explored how Pix2Pix, an Image-to-Image Translation Neural Network, would perform in predicting the form-finding results after training it with data sets populated with the results of our explorations. The first part of the workshop was structured around developing topologies for the Pix2Pix form finding training. Each particpant used Maya to craft various topologies through a series of modeling operations. Once the topologies were selected, participants used an MEL script to paint a gradient on the final meshes that would assign re-mapped force density values according to the ranges specified. Boundary constraints were assigned as well.

Once the topolgies went through a formfinding exercise, another MEL script was employed to export 100 various

configurations of the original mesh and resultant mesh. Data from the mesh generations were stored as a pixel matrix, with coordinates stored as RGB values and FDM values stored as grayscale values. 80% of these images were saved as a training data set and 20% were used as a test data set throughout the training process. Each participant trained their Pix2Pix network overnight, for a total of 150 epochs at 200 seconds per epoch. Once the training had been completed, our workshop regrouped and used the Pix2Pix network to predict the form of the resultant mesh from an original topology mesh. While the results did not appear to be especially accurate, they did exhibit some formal fidelity. The workshop instructors believed that with significantly more training images, Pix2Pix could produce more accurate results.

HEX VAULT

2016 / Personal Design Work AA Aarhus Visiting School 2016 Rethinking Patterns: Moving Towards Heterogenic Structures Published on Parametric Architecture Instagram, 2017 Instructors: Ryan Hughes David Reeves Asbjorn Sondergaard Florin Stan Team: Guy Gardner Daniel Sneddon Satyam Satyam Juliette Zidek

As a part of the Edward L. Ryerson Traveling Fellowship agenda, I attended the AA Aarhus Visiting School to engage in robotic fabrication firsthand. The Visiting School’s program in Aarhus focused on the integration of robotic tools and explored how these manufacturing techniques could influence the future of design and fabrication in architecture. Particpants were trained to use large and small-scale Kuka robotic arms to hot-wire cut foam for the construction of architectural installations that were to be displayed at DOKK1, a public library and cultural center in Aarhus, for a period of two weeks. This international program formed small teams of participants from around the world, and challenged the groups to design an installation that would explore the process of robotic hotwire cutting. Our group, comprised of Guy Gardner, Daniel Sneddon, Satyam Satyam and myself, decided to design and fabricate an inhabitable vault structure made of faceted hexagonal modules. HEX VAULT seeks to explore the advantage of the mass customization that robotic fabrication allows. Each of the 308 hexagonal moduls that compose the vault is entirely unique but designed and fabricated with the same script. The individuality of each module allows a maximum gradient of aperture size and surface texture without detracting from the efficiency of the fabrication process. HEX VAULT was conceived to frame views to surrounding points of interest from DOKK1: Universitetsparken, Isbjerget, and Lystbadehavnen to the north; Jelshoj to the south. It purposefully opens up toward the east to act as a threshold towards the Aarhus Bugt.

Since each piece of the pavilion was cut individually from larger foam blocks, a mechanical fastener was designed to hold the pieces intact. These fasteners were laser cut from bass wood and were easily wedged into each piece for quick assembly that relied on the precision of the robotic fabrication process to ensure alignment. The pieces were ordered rows by height relative to the base of the vault so that large sections could be preassembled before transporation to DOKK1. Project Role: I participated in all phases of the project’s design, from conceptualization through scripting of the vault form, panelization, and simulation of the hot wire cutting process. Our entire team worked (tirelessly!) to cut each peice of the vault, laser cut the mechanical fasteners, and assemble the structure at DOKK1. I was responsible for producing the drawings and diagrams presented in the following pages that were exhibited on a board with the HEX VAULT assembly.