Algorithmic Mumbai computational optimization of urban and architectural form on the example of the Eastern Waterfront transformation in Mumbai, India master’s thesis by Paweł Unger supervised by PhD Arch. Eng. Jan Cudzik and PhD Arch. Eng. Piotr Czyż reviewed by PhD Arch. Eng. Łukasz Pancewicz
5.11.2019
Mumbai, India
Eastern Waterfront, Mumbai
5km
population Mumbai
London
9 126 366 20 185 064
density
32 303 people/km²
6th the densest city
9 - 12 million live in slums depending on statistic
6% can afford housing
27. Singapore 29. Mumbai 36. Barcelona
29th wealthiest city in the world “Global city GDP rankings 2008-2025�. Pricewaterhouse Coopers. Archived from the original on 4 May 2011. Retrieved 16 December 2009.
$
40% of India’s foreign trade
500 000 imigrants a year
4 months of Monsoon rain
3.5h of commuting to job on average
Methodology XL scale strategic guidelines L scale urban guidelines
research
pedestrain corridors
geo-data
global influences
transportation nodes distribution
society
residential household to public transport distance
transportation nodes distribution
urban angle adjustment urban scale research
filter problems
environment
finding conotations
research and tools bank
XL scale (Estern Waterfron) 750ha
megalocalism
XL scale strategic guidelines
L scale (urban) 100ha
existing buildings use analysis
walkability mobility flooding predictions
algorithmic research
input
selection
analysis / simulation
scale
sun radiation simulation sun propoagation analysis
super services urban blocks widths
urban form
mobility analysis
proximity to the focal nodes
functions distribution in the urban scale
M scale (architecture)
Topology Optimization (TO)
pedestrian network
wind analysis
monsoon response typologies
output
environment
mobility
society
3 months of monsoon rain
long job distances
monsoon flooded public space
humid and hot climate
sinking metro system
0.5 to 4 hours job commuting
extreme air pollution
lack of water transport
6% of citizens can afford housing inefficient electricity
60% living in slums
analysis synthesis
Architectural Form cross ventilation analysis
pedestrian network pedestrian grid simulation
services
visual comfort simulation
quarter or full form builidng solar farms and roof urban farms distribution
offices
indoor farms
final output
XL scale strategic guidelines L scale urban guidelines M scale architecture
main problems
high tides and tsunamies
residential
services
heights of urban blocks
main urban grid solar radiation simulation
architectural scale research
M scale architecture
privacy and access simulation
functions distribution
XL scale: strategic guidelines
750ha
5km
XL
sea level
XL Top cities exposed to coastal flooding in the 2070s
Mumbai sea level prediction
2019
11,41mln (in 2070)
g n i d o o l f r o f d e s o p x e s n e z i t ci
~5m 2,78 mln
0m 0m
2119 ~6.5m e g n a r tide l e v e l sea
~1.5m
Hanson, S., Nicholls, R., Ranger, N., Hallegatte, S., Corfee-Morlot, J., Herweijer, C., & Chateau, J. (2011). A global ranking of port cities with high exposure to climate extremes. Climatic change, 104(1), 89-111. Deloitte City Mobility Index
XL sea range
1km
N
2 C warmer atmosphere in 2050 4 C warmer atmosphere in 2120 o
o
The Gate of India, Mumbai 2050
source: https://www.climatecentral.org
2120
XL
Learning from Chicago, 1850
“Chicago Tribune, February 7, 1866”. Retrieved July 21, 2019
“Chicago Tribune, February 7, 1866”. Retrieved July 21, 2019
XL
ground elevation
underground
on-ground
+3m 0m
initial ground level
infrastructure zone reservoir
XL
transportation and walkability
XL Statistic of mobility type usage in Mumbai:
55.5 % 21.9% 14.4% 3.1% 2.4% 1.6% 0.8% 0.3% walking train bus two rickshaw car bicycle taxi wheel-
XL walkability and train tranport
1km
20 minutes 10 minutes 5 minutes
N
XL Mumbai is a concentric mega-city
XL mega-local cluster parts: -focal point - residential - work - recreationial - services - health care - education
XL
multi-functional mega-local districs
XL
mega-localism High potential for the center of each Megalocal district:
Sassoon Dock Local Fish Market
The Gate of India
Horniman Circle Garden
Indira Dock
L scale Clock Tower
Silos park
Coal Burden
Hay Burden
Industrial Park
Vidyalankar Polytechnic Park
1km
N
XL
mega-localism High potential for the center of each Megalocal district:
Sassoon Dock Local Fish Market
The Gate of India
Horniman Circle Garden
Indira Dock
L scale Clock Tower
Silos park
Coal Burden
Hay Burden
Industrial Park
Vidyalankar Polytechnic Park
1km
N
L
L scale: urban guidelines
100ha
5km
L
L scale: urban guidelines 2D
L
grid setup
100x 50m urban block 15m offset for streets
100m N
L
W
main axis, the green corridors and tranportation
average wind direction
W S N
S E
train stations train connections with monsoon-proof areas water tram connections pedestrian green corridors with enhanced ventilation
100m N
L
inland green areas
100m N
vs
shortest path on Hippodamus grid
494.96m max distance 270.83m average distance
shortest path on high density grid
100m
100m
N
N
382.69m max distance 211.75m average distnace
27.9% shorter
L
L
N key: designed buildings existing buildings Eastern Waterfront parts outside of the study area Mangrove parks
L
L scale: urban guidelines 3D
L
density
Street widths
the furthest from train station
the closest to train station
150m height
buildings distnace
30m
Building’s height related to distance from public transport
15m the furthest from train station
the closest to train station
16m
urban ecology
generative roofs
urban farm
60m² of urban farm is enough to feed one person
solar farm
solar energy
L
urban ecology
L
solar radiation simulation
annual average of sun radiation
11 10 8.9 7.75 6.6 4.75 4.41 3.33 2.17 1.08 0
[hours]
L
urban ecology
key: solar farms urban farms
290 000 people powered by the roofs 25 000 people fed from roofs
mangroves
https://cms.qz.com/wp-content/uploads/2018/09/mangrove-Apple.jpg?quality=75&strip=all&w=1600&h=900
generates oxygen 1ha = 1760kg of CO2 less yearly 2m strip = 90% lower waves
L
mangroves
https://cms.qz.com/wp-content/uploads/2018/09/mangrove-Apple.jpg?quality=75&strip=all&w=1600&h=900
1ha = 1760kg of CO2 less yearly 2m strip = 90% lower waves
tetra-pods
https://alicegordenker.files.wordpress.com/2014/04/main-photo.jpg
1tone = 900kg of CO2 emmited 2m strip = 100% lower waves
L
urban ecology
mangrove parks
20 480kg of CO2 less a year
functions and focal points
buildings use proportion summary
residential 65%
services 25% super services 10%
range of groundfloors services 50m range from squares
high potential commercial zone 100m range from train station
L
low typologies elevated train platform
volume slider
checkerboard
motion
100% 100%
5% 33%
36% 37.5%
55% 54%
cross ventilation dry public space [% of a footprint]
high-rise typologies
volume slider cross ventilation dry public space [% of a footprint]
23% 92%
checkerboard
motion
random
80% 37.5%
59% 203%
48% 149%
L
M
M scale: architecture
M
?? ?? ?? ??
M Train station 70m
N key: designed buildings existing buildings Eastern Waterfront parts outside of the study area urban block chosen for the architectural development (M scale) Mangrove parks
Mangrove park 100m
Thane Creek 20m
M L scale to M scale: architecture based on the urban strategy
2D passage
from the train station to the bay
M
3D maximum volume
40x70x110m
generative roof type
functions
solar roof
1/3 residential 1/3 offices 1/3 services
M step 1 : form finding Sun radiation analysis of the maximum building’s volume to reduce sun exposition of the building
step 2 : structural optimization Finding optimal structural for with Topology Optimization for the given geometry from the setp 1
step 3 : programme optimization A set of simulations to discover the optimal function for each part of the building.
step 1 : sun forms geometry
sun radiation simulation process:
M
M
step 1 : sun forms geometry
analysis results
annual sunlight average
4h
5h
6h
7h
8h
9h
10h
M
step 1 : sun forms geometry
analysis results
annual sunlight average
4h
5h
6h
7h
8h
9h
10h
M
step 1 : sun forms geometry sun optimized geoemtry
final geometry with the passage
k e e r C e Than n o i t a t s train
M
step 2 : structural optimization
Morphogenesis of flux structure with Topology Optimization
M 1914 LeCorbusier - architectural perspective
Opposite approaches. Le Corbusier’s Mansion Domino 1914 (left) structural sub-rationalism
M 1914 LeCorbusier - architectural perspective
Opposite approaches. Le Corbusier’s Mansion Domino 1914 (left) structural sub-rationalism
vs 1904 George Mitchell - mathematical perspective
M 1914 LeCorbusier - architectural perspective
vs 2011 A. Isozaki, M. Sasaki, Convention Center, Qatar- architectural perspective
Opposite approaches. Le Corbusier’s Mansion Domino 1914 (left) structural sub-rationalism, Topology Optimized structure by Sebastian Białkowski 2018 (right) structural rationalism
M 1914 LeCorbusier - architectural perspective
n o i
l a
t a r ri
vs 2011 A. Isozaki, M. Sasaki, Convention Center, Qatar- architectural perspective
l a
t a r
n o i
Opposite approaches. Le Corbusier’s Mansion Domino 1914 (left) structural sub-rationalism, Topology Optimized structure by Sebastian Białkowski 2018 (right) structural rationalism
Example of a TO implementation: Initial design configuration of the L-bracket optimization case and the solution by R. Picelli et al
step 1 : TO algorithm initial conditions
step 2 : calculations
material: stainless steel
load:
algorithm setup:
Density Value [%] 3.5 Young Modulus [GPa] 210.0 Poisson Factor [v] 0.3
class C5 regarding norm PN-EN 1991-1-1
resolution [pts/m2] 100.0 iterations [int] 60
volume for structure
Load Value 7.5 [kN/m2]
initial conditions
boundin g box
step 3 : results
run iterative optimization process topology optimized geometry [mesh]
optimized structure
diagonal prestressing rod ring steel pipe
loads
main steel pipe
aluminum coating
suppor ts locat ion
TOPOLOGY OPTIMIZATION
axonometry
northern view
eastern view
southern view
western view
eastern view
sounthern view
western view
M
ground floor
SOLAR OPTIMIZATION
northern view
roof top
annual average of sun radiation 10 9 8 7 6 5 4 3 2 1 0
[hours]
M
step 3 : programmatic optimization
functions distribution in the given volume
M functions:
parameters for the functions distribution optimization:
1. residential 3. services 2. offices 4. urban farms
1. visual comfort 3. cross ventilation abilities 2. sun radiation analysis 4. privacy rate
analysis of visual comfort and cross-ventilation abilities
M
sun radiation simulation
M
privacy rate simulation
M
M
programme distribution
farms
75 modules = 8.8% of all modules
housing
363 modules = 42.9% of all modules
services
222 modules = 26.2% of all modules
offices
187 modules = 22.1% of all modules
section variation
M
ground floor variation
2 045m² of monsoon - proof public space
15x17 26.5
15x17 26.5
15x17 26.5
15x17 26.5
M
M
9th floor variation D 10x17 27.5
15x17 26.5
7x17 27.5
15x17 26.5
15x17 26.5
15x17 28
15x17 26.5
15x17 26.5
C
15x17 26.5
15x17 26.5
15x17 26.5
15x17 26.5
15x17 26.5
D
C
monsoon-proof public space
M
view from the bay
M
general performance
L scale
M scale
urban scale:
architecture:
73 000 people to inhabit 57m to sustainable transport
up to 40% less construction material most exposed to sun areas eliminated
25 000 people fed from roofs
36.5% of cross ventilated space
290 000 powered from roofs
2 045m² monsoon-proof public space
flood areas eliminated
21 475m² for rent
49.97% covered with greenery
1900m² of low sun exposed terraces
20 480 kg less CO2 annually monsoon-proof public spaces
790MW of power generated per year 1 875m² or 9 375m³ of urban farms
fresh air with urban ventilation
160people feed from the solar farms
Algorithmic Mumbai thank you for your attention