Sustainability 11 06965 v2 PDF

sustainability
Article
Subtractive Building Massing for Performance-Based
Architectural Design Exploration: A Case Study of
Daylighting Optimization
Likai Wang 1 , Patrick Janssen 2 , Kian Wee Chen 3 , Ziyu Tong 1, * and Guohua Ji 1, *
1 School of Architecture and Urban Planning, Nanjing University, Nanjing 210093, China;
dg1436002@smail.nju.edu.cn
2 School of Design and Environment, National University of Singapore, Singapore 117566, Singapore;
patrick@janssen.name
3 Andlinger Center for Energy and the Environment, Princeton University, Princeton, NJ 08544, USA;
chenkianwee@gmail.com
* Correspondence: tzy@nju.edu.cn (Z.T.); jgh@nju.edu.cn (G.J.); Tel.: +86-13805178067 (Z.T.);
+86-13605192392 (G.J.)

Received: 15 November 2019; Accepted: 4 December 2019; Published: 6 December 2019
Abstract: For sustainable building design, performance-based optimization incorporating parametric

modelling and evolutionary optimization can allow architects to leverage building massing
design to improve energy performance. However, two key challenges make such applications
of performance-based optimization difficult in practice. First, due to the parametric modelling
approaches, the topological variability in the building massing variants is often very limited. This, in
turn, limits the scope for the optimization process to discover high-performing solutions. Second, for
architects, the process of creating parametric models capable of generating the necessary topological
variability is complex and time-consuming, thereby significantly disrupting the design processes.
To address these two challenges, this paper presents a parametric massing algorithm based on the
subtractive form generation principle. The algorithm can generate diverse building massings with
significant topological variability by removing different parts from a predefined volume. Additionally,
the algorithm can be applied to different building massing design scenarios without additional
parametric modelling being required. Hence, using the algorithm can help architects achieve an
explorative performance-based optimization for building massing design while streamlining the
overall design process. Two case studies of daylighting performance optimizations are presented,
which demonstrate that the algorithm can enhance the exploration of the potential in building massing
design for energy performance improvements.
Keywords: parametric massing algorithm; building massing design; performance-based optimization;

design exploration; subtractive form generation principle; passive energy savings; daylighting
1. Introduction
Recent decades have witnessed rapid urbanization in China, which has fed the demand for more
energy-efficient buildings moving towards a more sustainable future. While engineering methods,
such as system control, renewable energy generation, and high-performance materials, have been
widely applied in the building sector, the contribution made by architectural design remains limited.
This issue has raised widespread concerns in architecture and increased the awareness of embracing
performance-based design or performative design in architectural design [1,2]. In performance-based
design for energy sustainability, building massing design can play a role in energy performance
improvement [3–5]. A good building massing design can make the building adapt to the surrounding
Sustainability 2019, 11, 6965; doi:10.3390/su11246965 www.mdpi.com/journal/sustainability

Sustainability 2019, 11, 6965 2 of 20
urban environment and take advantage of climate resources, such as sun, light, and wind. As a result,
the building can benefit from passive energy savings by leveraging daylighting, passive heating or
cooling, and natural ventilation.
In sustainable architectural design, there are certain widespread building massing design strategies
for passive energy saving, such as atriums, courtyards, and stilts. With these strategies, it is possible
to achieve a moderate improvement in passive energy saving by following certain rules-of-thumb or
guidelines [6]. However, the full exploitation of applying these strategies for progressive energy saving
is not trivial. Because of the uncertainty resulting from the interaction between the climate conditions
and the specific surrounding urban environment, the application of these strategies in building massing
design for each project must be site-specific [7]. In this regard, a systematic exploration of a broad
range of building massing design alternatives can help to identify the appropriate application of these
strategies for the project.
In the past, building massing design exploration has mostly been partial and limited as
manually generating many different configurations and combinations of these strategies is tiring
and time-consuming. Recent progress in parametric modelling and evolutionary optimization
is considered a possible solution to these challenges. This allows architects to undertake
automated performance-based optimization as a means of building massing design exploration.
Parametric modelling allows for a design space (solution search space) to be defined, encompassing a
large number of design variants for evolutionary optimization to explore. Evolutionary optimization
evolves the population of building massing design and aims to identify the computed optimal solution
with the desired energy performance [8–10].
In the last decade, a growing body of research literature has developed around efforts to apply
performance-based optimization to maximize the potential of building massing design in passive energy
savings [11,12]. Despite many successful applications in the literature, applying performance-based
optimization for building massing design exploration is still challenging. While certain barriers
have been identified by other researchers, such as the lack of easy-to-use optimization tools and the
problematic integration of optimization with architectural design [13–15], another critical barrier is the
lack of topological variability among the variants offered by these techniques in the design space.
The lack of topological variability means that the building massing design variants generated by
the parametric model show little or no topological differentiation to the change of parameters. Thus,
despite a large number of design variants, the design space may actually only cover a small subset of all
possible design alternatives, which excludes many other competitive alternatives from being explored
by the optimization process. In order to overcome the lack of topological variability in the design
space, this research proposes a parametric algorithm to generate building massing designs based on a
subtractive form generation principle. The process starts with a solid building block defining the basic
volume of the building massing. By removing different parts from the building block, the generated
building massings show significant topological variability, which can imply different configurations
and combinations of various passive energy-saving strategies in the building massing design.
To place the research in context, we first discuss the progress that has been made in parametric
modelling for building massing design before going on to describe the details of the proposed algorithm.
We then present two case studies with associated results. We conclude the paper by discussing the
relative effectiveness of this algorithm and some limitations.
Parametric Modelling for Building Massing Design

Parametric models describe design by explicit rules and parameters [16]. Often, the rules encoded
are rigorous but inflexible. This often results in the design space defining a family of design variants that
are all very similar with limited topological variability [10,17–19]. The lack of topological variability is
often overlooked as architects may use parametric models to describe a specific design strategy for
each optimization. Hence, the fact that the design space only includes similar design variants is seen
as acceptable, since they are all derived from the same strategy. A case in point is that for building
massing design optimizations aimed at daylighting or passive solar energy, many researchers or
architects create parametric models to describe building blocks with central internal courtyards [20,21].
With such models, the shape of the courtyard can typically be changed by formal transformation
operations, such as rotation, twisting, and slanting [22,23]. The result of the optimization may
show remarkable improvements by identifying the fittest shape of the courtyards. However, such
optimization nonetheless leaves many other competitive solutions unexplored, for example, the one
characterized by features, such as stilts and solar envelopes.
In order to overcome the lack of topological variability, other researchers have proposed
sophisticated parametric modelling approaches to offering discrete formal variability of the generated
building massing. One commonly used approach is to include several massing algorithms, respectively
describing different building massing forms in one parametric model, and select one of these algorithms
to create the building massing each time [6,24,25]. This approach is often used in design problems at the
urban scale, and the desired urban form diversity can be simply achieved by varying permutations and
combinations of different forms over multiple buildings. Nault et al. [6] and Xiaodong et al. [25] used
this approach to create neighborhood-scale and urban block design based on a set of basic building
forms for energy-daylighting optimization and outdoor thermal comfort optimization. Another
approach is to create building massings composed of various small mass units. By arranging and
rearranging these mass units, the algorithm can generate different spatial configurations [26–29].
This approach is also widely used in conjunction with thermal-zone layout optimization to achieve
better energy efficiency [30,31]. Buildings generated by this approach typically have a cellular-look
massing. However, due to the parametric approach, the arbitrary arrangement of the mass units often
results in many chaotic design variants [28].
Applying the two approaches mentioned above is technically demanding and consumes a
great amount of time and effort. The former requires laborious developments of different massing
algorithms [19], while the latter requires complex constraint handling to ensure the designs generated
are meaningful [28,29,31].
Regardless of the difficulty in developing massing algorithms capable of offering desired
topological variability, another drawback shared by common performance-based optimization exercises
is that a large amount of parametric modelling is required for different projects. Hence, although
there are approaches to generating building designs with high topological variability, architects may
still struggle with such a complex parametric modelling task, which further results in significant
interruptions to the design process. This issue has been drawing concern recently [6,19,22,32].
The reusability of the parametric model, therefore, becomes a critical issue for minimizing the
interruption to the design process [6,33]. Ideally, it should be possible to reuse parametric models
in different design contexts [9,34]. However, this merit is often overshadowed as highly customized
parametric models are often built on architects’ preferences and biases. The designs generated by such
customized parametric models are often too specific or unique to be reused in other design contexts.
In contrast, this research explores the idea that it may be possible to reuse a parametric model
for different design contexts if generic design (building massing) strategies can be represented by the
parametric model [35]. Recently, a few studies have attempted to develop design support systems,
including predefined generic building forms, to reduce the effort and time invested in parametric
modelling, thereby allowing architects to carry out streamlined performance-based optimization
processes without the interruption of parametric modelling [6,25]. However, as pointed out in these
studies, offering sufficient topological variability in the design generated, again, presents a significant
challenge for developers.
In contrast to the existing approaches, the proposal of using the subtractive form generation
principle in parametric building massing algorithms may have some advantages in overcoming the
challenges related to the lack of topological variability and low reusability:
• First, the subtractive form generation principle is one of the most generic and widely applied
massing strategies in architecture. This strategy can fit many different common types of building
designs [4]. Using this principle ensures that, with simple constraints, most of the building
massings generated have acceptable and reasonable architectural features. In addition, by
removing different parts from a predefined building volume, the subtracted building massing can
show great topological variability. These two advantages can efficiently overcome the weaknesses
identified above.
• Second, many passive energy-saving strategies for building massing design, such as courtyards,
stilts, and solar envelopes, can be schematically described as removing (subtracting) parts from
a building block [3,35], which complies with the subtractive form generation principle. Thus,
with different parts removed, the subtracted building massing can be recognized as different
combinations and arrangements of these passive energy-saving strategies imposed in the building
massing design.
• Third, the subtractive form generation approach allows for easy customization of the types of
building design features that are generated. Architects can tune the types of features by adjusting
various global parameters associated with the subtractors. For example, such parameters may
include the size, position, and number of subtracting parts.
2. Methods
2.1. Overview of Proposed Algorithm

The proposed algorithm generates building massings according to a subtractive form generation
principle. The algorithm starts with a maximal building volume and then creates building massing
design variants by subtracting parts of this volume. By varying the number of parts to be removed as
well as their positions and dimensions, alternative massings with significant topological variability can
be generated. The core of the algorithm is to define the part to be removed, which is referred to as the
“subtractor” in this paper. The definition and formation of the subtractor consist of three basic steps:
(1) Initialization of a maximal volume and the subtractors, (2) constraining the position and size of
the subtractors, and (3) aligning the subtractors. In addition, there is one optional step to generate
building cores.
In order to be able to provide control over the features of the building massing after subtraction,
several initialization parameters are provided. These parameters are assigned before the optimization
process starts and remains fixed throughout the process. Other parameters, referred to as the
optimization parameters, are changeable and modified by the evolutionary algorithm during the
optimization processes. The initialization parameters include: (1) the dimension of the maximal
volume, (2) the number of subtractors, (3) the size constraint of the subtractor, and (4) the boundary
constraint regulating the relationship between subtractors and the maximal volume. The optimization
parameters define the dimension and spatial position of each subtractor. Moreover, two optional
initialization parameters are: (1) Specifying the footprint of the maximal volume, and (2) turning on or
off the generation of building cores.
Once the initialization parameters have been assigned, the algorithm can generate building
massing variants by varying the optimization parameters list, and these variants can then be sent to
various energy simulation tools for performance evaluation. An evolutionary algorithm can be used
to evolve building massing designs by iteratively changing optimization parameters and receiving
corresponding performance feedback from simulations. The overall building massing creation and
iterative evolutionary processes are illustrated in Figure 1.
2.2. Generative Steps

The generative process consists of three main generative steps related to the subtractors:
Initialization, constraints, and alignment. Finally, a fourth step can also optionally be added, to
insert building cores.
Sustainability 2019, 11, x FOR PEER REVIEW 5 of 21

Figure 1. Schematic diagram of the building massing creation and evolutionary optimization process.
2.2. Generative Steps

The generative process consists of three main generative steps related to the subtractors:
Initialization, constraints, and alignment. Finally, a fourth step can also optionally be added, to insert
building
Figurecores.
Figure 1.1.Schematic
Schematicdiagram
diagramofofthe
thebuilding
buildingmassing
massingcreation
creationand
andevolutionary
evolutionaryoptimization
optimizationprocess.
process.
2.2.1.
2.2.1.
2.2. Subtractor
Subtractor
Generative Initialization
Initialization
Steps
The
The first
The firststep
stepisisto
generative toinitialize
initialize
process the
themaximal
consistsmaximal
of threevolume
main and
volume and the
thesubtractors.
generative subtractors. The
The shape
steps related shape
to theof
of the
the maximal
maximal
subtractors:
Initialization, constraints, and alignment. Finally, a fourth step can also optionally be added, tocolumn
volume
volume is
is aa rectangular
rectangular block
block by
by default,
default, and
and its
its dimension
dimension is
is defined
defined by
by the
the number
number of
of column
insert
gridpoints
gridpoints
building in the x- and y-directions and the number of floors in the z-direction which
in the x- and y-directions and the number of floors in the z-direction, which are assigned
cores. are assigned by
architects in the(Figure
by architects initialization parametersthe
2-a). Changing (Figure 2a). Changing
dimensions allowsthe dimensions
different types allows different types
of buildings to be
of buildings
generated, to
e.g.,be generated,
high-rise
2.2.1. Subtractor Initialization or e.g., high-rise
low-rise or
buildings.low-rise
Two buildings.
types of For each building
subtractors—vertical massing
subtractorsdesign
and
variant, two types of subtractors—vertical subtractors and horizontal subtractors—are
horizontal subtractors—are created based on values in the optimization parameters lists. Vertical created based on
The first step is to initialize the maximal volume and the subtractors. The shape of the maximal
values in the have
subtractors optimization
an aimsparameters
to create lists. Vertical
features subtractors
related havestrategies,
to design an aims to create
such as features
atriumsrelated
and
volume is a rectangular block by default, and its dimension is defined by the number of column
to design strategies,
courtyards suchIn
(Figure 2-b). as contrast,
atriums and courtyards
horizontal (Figure aim
subtractors 2b). to
In create
contrast, horizontal
features, suchsubtractors
as stepped
gridpoints in the x- and y-directions and the number of floors in the z-direction, which are assigned
aim to create
(cascade) features,
roofs, emptysuch as stepped
floors, (cascade)
and stilts, roofs, empty
in the building floors,(Figure
massing and stilts,
2-c).inUsing
the building
these two massing
types
by architects (Figure 2-a). Changing the dimensions allows different types of buildings to be
(Figure 2c). Using
of subtractors, the these twomassing
building types ofissubtractors,
generated by theremoving
building the
massing
parts is generated
occupied by by removing
these the
subtractors
generated, e.g., high-rise or low-rise buildings. Two types of subtractors—vertical subtractors and
parts occupied by these subtractors
from the maximal volume (Figure 2-d). from the maximal volume (Figure 2d).
horizontal subtractors—are created based on values in the optimization parameters lists. Vertical
subtractors have an aims to create features related to design strategies, such as atriums and
courtyards (Figure 2-b). In contrast, horizontal subtractors aim to create features, such as stepped
(cascade) roofs, empty floors, and stilts, in the building massing (Figure 2-c). Using these two types
of subtractors, the building massing is generated by removing the parts occupied by these subtractors
from the maximal volume (Figure 2-d).
Figure 2.2. The

Figure The creation
creation of
ofthe
thebuilding
buildingmassing
massingbybytwo
twotypes
typesofofsubtractors:
subtractors: (a)
(a)generating
generating maximal
maximal
volume,
volume, (b) (b) vertical
vertical subtractors,
subtractors, (c)
(c) horizontal
horizontal subtractors,
subtractors, and
and (d)
(d) assembling
assembling of of vertical
vertical and
and
horizontal
horizontalsubtractors.
subtractors.
The number of subtractors, defined by architects, is an important factor affecting the topological
variability of the generated building massings. Figure 3 shows the generated building massings with
different
Figurenumbers of subtractors.
2. The creation It can
of the building be noticed
massing that
by two when
types the number
of subtractors: (a)of subtractors
generating increases,
maximal
subtractors
volume, become merged
(b) vertical and turn
subtractors, (c)the negativesubtractors,
horizontal volume subtracted from the building
and (d) assembling mass
of vertical andinto a
horizontal subtractors.
Sustainability
Sustainability2019,
2019,11,
11,xxFOR
FORPEER
PEERREVIEW
REVIEW 66of
of21
21
The number
The of subtractors, defined by architects, is an important factor affecting the topological
The number
numberof of subtractors,
subtractors, defined
definedby byarchitects,
architects,is
isan
an important
important factor
factor affecting
affecting the
the topological
topological
variability of
variability the generated building massings. Figure 3 shows the generated building massings with
variability of of the
the generated
generated building
building massings.
massings. Figure
Figure33 shows
shows the
the generated
generated building
building massings
massings with
with
different
Sustainability
different numbers
2019, 11, 6965of subtractors. It can be noticed that when the number of subtractors increases,
6 of 20
different numbers
numbers of of subtractors.
subtractors. ItIt can
can bebe noticed
noticed that
that when
when thethe number
number ofof subtractors
subtractors increases,
increases,
subtractors
subtractors become merged and turn the negative volume subtracted from the building mass into a
subtractors become
become mergedmerged and
and turn
turn thethe negative
negative volume
volume subtracted
subtracted from
from the
the building
building mass
mass into
into aa
complex
complex topological configuration.
topological By changing the number of subtractors, architects can control the
complex
complex topological configuration.
topological configuration. By
configuration. By
By changing
changingthe
changing thenumber
the numberof
number ofsubtractors,
of subtractors,
subtractors, architects
architects
architects can
cancontrol
can control the
control the
the
overall
overall topological variability and configurational complexity of the generated building massings.
overall topological
overall topological variability
topological variability and
variability and configurational
and configurational complexity
configurational complexity of
complexity of the
of the generated
the generated building
generated building massings.
building massings.
massings.
Figure
Figure 3.3.
3. Generated
Generated
Generated building
building
building massings
massings
massings based
based
based on
on on different
different
different numbers
numbers
numbers of
ofofsubtractors:
subtractors:VV indicates
indicates vertical
vertical
Figure
Figure 3. Generated building massings based on different numbers ofsubtractors:
subtractors:V Vindicates
indicates vertical
vertical
subtractors,
subtractors,
subtractors, H
H Hindicates
indicateshorizontal
horizontalsubtractors.
subtractors.
subtractors, H indicates
indicates horizontal
horizontal subtractors.
subtractors.
AsAsthetheaccumulated
the accumulatednegative
accumulated negativevolume
negative volumesubtracted
volume subtracted
subtracted byallall subtractors isisdynamically changed by
As
As the accumulated negative volume subtracted byby
by allall subtractors
subtractors
subtractors is dynamically
is dynamically
dynamically changedchanged
changed by
by
varying
by varying the optimization parameters, the gross area of the subtracted building massing varies
varying
varying thethe
the optimization
optimization
optimization parameters,
parameters,
parameters, thethe
the gross
gross
gross area areaof
area ofofthe
thesubtracted
the subtractedbuilding
subtracted building massing
building massing varies
massing varies
varies
accordingly.
accordingly.
accordingly. The
The gross
gross area
area isisa a critical
critical functional
functional requirement
requirement inin architectural
architectural design.
design. Therefore,
Therefore,
accordingly. The The gross
gross area
area is is aa critical
critical functional
functional requirement
requirement in in architectural
architectural design. design. Therefore,
Therefore,
althoughaa building
although a building massingmay mayhavehave excellent energy performance, it isititofitis isofoflittle
littlerelevance
relevance ifififitsits
although
although buildingmassing
a building massingmay
massing mayhave have excellent
excellent
excellentenergy
energy
energy performance,
performance,
performance, islittle relevance
of little if its gross
relevance its
gross area
area isarea
too is is
low too
orlowlow
too or or
high. too high. The algorithm, therefore, adjusts the maximal volume to create a
gross
gross area is too
too low tooThe
or too high.
high. algorithm,
The therefore,
The algorithm,
algorithm, adjusts adjusts
therefore,
therefore, the
adjustsmaximal
the volume volume
the maximal
maximal to createto
volume a building
to create
create aa
buildingclose
massing massing
to acloseclosetogross
target to a target
areagross
bygross area byincreasing
iteratively iterativelyincreasing
increasing ordecreasing
decreasing thenumber
number of
building
building massing
massing close to aa target
target gross area
area by iterativelyor
by iteratively decreasing
increasing orthe
or number the
decreasing of
thecolumn
number gridof
of
column
spans and grid spans
floors. With and floors. With the maximal volume adjusted, the appearance of subtractors also
column
column grid
grid spans
spans andthe
and maximal
floors.
floors. Withvolume
With the
the maximal adjusted,
maximal volume
volume the appearance
adjusted,
adjusted,the ofappearance
the subtractors of
appearance also varies as some
of subtractors
subtractors also
also
varies as
subtractors
varies some
may subtractors
be deactivated may be
due deactivated
to the size due to
constraint, the size
which constraint,
is described which in is
Sectiondescribed
2.2.2 in Section
(Figure 4).
variesas assome
somesubtractors
subtractorsmay maybe bedeactivated
deactivateddue dueto tothethesize
sizeconstraint,
constraint,which whichis isdescribed
describedin inSection
Section
2.2.2
The
2.2.2 (Figure
adjustment 4). The
of adjustment
building massingof building
is massing
executed during is executed
the during
building the
massing building
creation massing
process creation
and
2.2.2(Figure
(Figure4). 4).The
Theadjustment
adjustmentof ofbuilding
buildingmassing
massingis isexecuted
executedduring duringthe thebuilding
buildingmassing massingcreationcreation
isprocess
process andandis isindependent
independent independent
of the of ofthe
evolutionary theevolutionary
evolutionary
optimization optimization
process.
optimization At process.
the same
process. At At
time,the thesamesame
in order time,
time, toin inorder
give order
some to to
process and is independent of the evolutionary optimization process. At the same time, in order to
give somecontrol,
additional additional control,
architects architects are able to freeze certain (x, dimensions (x, y, or z).
give
give some
some additional
additional control,are
control, able to are
architects
architects freeze
are able
ablecertain
to
to freeze
freezedimensions
certain
certain dimensionsy, or z). (x,
dimensions (x, y,
y, or
or z).
z).
Figure4.4.
Figure
Figure Generatedbuilding
Generated buildingmassings
massingswith
withdifferent
differenttarget
targetgross
grossareas.
areas.
Figure 4.
4. Generated
Generated building
buildingmassings
massings with
withdifferent
different target
target gross
gross areas.
areas.
The
Themaximal
The maximalvolume
volumethat
thatisisaarectangular
rectangularblock
blockmay
maynot
notfitfit
into
intoirregular
irregularbuilding
buildingplots
plotsoror
satisfy
satisfy
Themaximal
maximalvolume
volumethat
thatis
isaarectangular
rectangularblock
blockmay
maynotnotfit
fitinto
intoirregular
irregularbuilding
buildingplots
plotsororsatisfy
satisfy
architectural intentions. Therefore,
architectural intentions.Therefore,
architectural the algorithm
Therefore, thealgorithm provides an initialization
algorithm providesananinitialization parameter to specify
specifythe
initialization parametertotospecify the
architectural intentions.
intentions. Therefore, thethe algorithm provides
provides an initialization parameter
parameter to specify the the
footprint
footprint
footprint ofofthe
themaximal
maximal volume,
volume, which
whichisisachieved
achieved by
by removing
removing a afixed
fixedpart
part from
from the
thebuilding
building
footprint of
of the
the maximal
maximal volume,
volume, which
which is
is achieved
achieved by by removing
removing aa fixed fixed part
part from
from the
the building
building
massing.
massing.As
massing. As shown
shownininFigure
Asshown Figure5,5,this
this
this operation
operation can
can create
create maximal
maximal volumes
volumes with
with anL-shaped
L-shaped
anL-shaped ororU-
massing. As shown in Figure 5, this operation can create maximal volumes with an L-shaped or
operation can create maximal volumes with an or U-
U-
U-shaped footprint.
shaped footprint.
shaped
shaped footprint.
footprint.
Figure Generated
5. 5.
Figure5. Generatedbuilding massings
building withwith
massings different footprints:
different (a) rectangular
footprints: footprint,
(a) rectangular (b) L-shaped
footprint, (b) L-
Figure
Figure 5. Generated
Generated building
building massings
massings with
with different
different footprints:
footprints: (a)
(a) rectangular
rectangular footprint,
footprint, (b)
(b) L-
L-
footprint, and (c)
shaped footprint, U-shaped
and (c) footprint.
U-shaped footprint.
shaped
shaped footprint,
footprint, and
and (c)
(c) U-shaped
U-shaped footprint.
footprint.
2.2.2. Subtractor Constraints
2.2.2. Subtractor
2.2.2. Constraints
2.2.2. Subtractor
Subtractor Constraints
Constraints
The second step is to constrain the size and position of subtractors according to the values in the
TheThe second step is to constrain the size and position of subtractors according to the values in the
The second
optimization step
step is
is to
parameters
second to constrain
list. the
the size
In order
constrain to and
and position
prevent
size of
of subtractors
over-large
position according
or over-small
subtractors voidsto
according the
from
to the values
valuesin
in the
appearing in
the
optimization
optimization parameters
parameters list. In order to prevent over-large or over-small voids from appearing in
the building massing,
optimization parameters thelist. In
In order
horizontal
list. to
to prevent
preventofover-large
orderdimension subtractors
over-large or
orisover-small
restricted voids
over-small by thefrom
voids appearing
size constraint,
from appearing byin
in
which architects specify the upper and lower size limit of subtractors in the unit of column-grid span
numbers. When the size of a subtractor does not satisfy the size constraints, a number of operations are
the building massing, the horizontal dimension of subtractors is restricted by the size constraint, by
the building
Sustainability massing,
2019, 11, 6965 the horizontal dimension of subtractors is restricted by the size constraint, 7 ofby
20
numbers. When the size of a subtractor does not satisfy the size constraints, a number of operations
numbers. When the size of a subtractor does not satisfy the size constraints, a number of operations
are to
used used to automatically modify the subtractors. When the size is above thetheupper limit, ofthe
are usedautomatically
to automaticallymodifymodify
the subtractors. When the
the subtractors. size isthe
When above
sizethe
is upper
abovelimit,
the upper dimension
limit, the
dimension
the subtractor of the subtractor
is decreased is decreased to the upper limit. When the size is below the lower limit,
dimension of the subtractortoisthe upper limit.
decreased to theWhen
upperthe sizeWhen
limit. is below
thethe
sizelower limit,
is below the
the subtractor
lower limit,
the subtractor
is deactivated, is deactivated, and the corresponding part is not subtracted from the building massing
the subtractor isand the corresponding
deactivated, part is not subtracted
and the corresponding part is notfrom the building
subtracted from the massing
building(Figure
massing 6).
(Figure
Note that 6).
theNote
size that the sizedoes
constraint constraint
not does notthat
guarantee guarantee
all that all subtracting
subtracting parts parts appearing
appearing in the in the
building
(Figure 6). Note that the size constraint does not guarantee that all subtracting parts appearing in the
building
mass are massmass are satisfactory.
satisfactory. When twoWhen When
or moretwo or more subtractors arecan
merged, it can turn void
into a large
building are satisfactory. twosubtractors are merged,
or more subtractors areitmerged,
turnitinto
cana turn
large into a in the
large
void
maximal in the maximal
volume volume
thatvolume that
could exceed could exceed
the upper the upper boundary.
void in the maximal that could exceedboundary.
the upper boundary.
Figure 6. The size constraint in the horizontal direction: assuming the size constraint is 2 to 5, (a)
Figure 6. The size
Figure sizeconstraint
constraintininthe horizontal
the horizontaldirection: assuming
direction: the size
assuming constraint
the size is 2 tois5,2(a)
constraint to before
5, (a)
before size being constrained, and (b) after size being constrained.
size being
before size constrained, and (b)and
being constrained, after
(b)size being
after size constrained.
being constrained.
Apartthan
Other from theuser-defined
user-defined sizeconstraint
constraintininthe thehorizontal
horizontaldirection,
direction,another
anotherconstraint,
constraint,which
which
Apart from the
the user-defined size size constraint in the horizontal direction, another constraint, which
is hard-coded
is hard-coded
hard-coded in in
in thethe algorithm,
the algorithm, restricts
algorithm, restricts
restricts the the
the size size
size of of subtractors
of subtractors
subtractors in in
in thethe vertical
the vertical direction.
direction. In In order
In order
order tototo
is vertical direction.
differentiatevertical
differentiate verticalandandhorizontal
horizontalsubtractors,
subtractors,two twodifferent
differentconstraints
constraintsare aresubject
subjecttotothethetwo
twotypes
types
differentiate vertical and horizontal subtractors, two different constraints are subject to the two types
of subtractors.
of subtractors.
subtractors. For For
For thethe vertical
the vertical subtractors,
vertical subtractors,
subtractors, the the top
the top
top and and bottom
and bottom
bottom facesfaces
faces are are automatically
are automatically
automatically aligned aligned
aligned to to the
to the
the
of
maximalvolume
maximal volume ifthe thedisplacement
displacementbetween
betweenthe theface
faceandandthethemaximal
maximalvolumevolumeisisless lessthan
than30%30%ofof
maximal volume ifif the displacement between the face and the maximal volume is less than 30% of
the total
the total height
total height
height of of
of thethe building
the building (Figure
building(Figure
(Figure7-a).7-a).
7a). At At
At thethe
the samesame
same time,time,
time, for for each
for each vertical
each vertical subtractor,
vertical subtractor,
subtractor, at at least
at least
least oneone
one
the
face has
face has
has to to satisfy
to satisfy
satisfy thisthis constraint,
this constraint,
constraint, or or else
or else
else the the subtractor
the subtractor
subtractor is is deactivated.
is deactivated.
deactivated. For For horizontal
For horizontal subtractors,
horizontal subtractors,
subtractors, the the
the
face
verticalsize
vertical sizeofofevery
everysubtractor
subtractorhashastotobebeless
lessthan
than30% 30%ofofthe thetotal
totalheight
heightofofthethebuilding
buildingbut butshould
should
vertical size of every subtractor has to be less than 30% of the total height of the building but should
also
also notnot be smaller than one floor. For any horizontal subtractors violating this condition, the
also notbe besmaller
smaller than oneone
than floor. For any
floor. Forhorizontal
any horizontal subtractors violating
subtractors this condition,
violating the algorithm
this condition, the
algorithm
automatically automatically
changes the changes the
size to the
make size
it to to make
satisfactory it satisfactory (Figure
(Figure 7b).(Figure 7-b). 7-b).
algorithm automatically changes size make it satisfactory
Figure 7. The size constraint in the vertical direction.

Apart from the size constraint, users also define the boundary constraint, which determines
Apart from the size constraint, users also define the boundary constraint, which determines
Apart
whether from subtractors
vertical the size constraint,
can breakusers alsothe
through define
outerthe
faceboundary constraint,
of the maximal volume which
in thedetermines
horizontal
whether vertical subtractors can break through the outer face of the maximal volume in the horizontal
whether
direction.vertical
When subtractors canisbreak
the constraint throughany
deactivated, the outer face
face of of the maximal
a vertical subtractorvolume
close into the horizontal
face of the
direction. When the constraint is deactivated, any face of a vertical subtractor close to the face of the
direction. When the
maximal volume is constraint
aligned to is
onedeactivated,
of the facesany
of face of a vertical
the maximal subtractor
volume so thatclose to theaface
it creates of the
vertically
maximal volume is aligned to one of the faces of the maximal volume so that it creates a vertically
maximal
open voidvolume
runningisthrough
aligned the
to one of thevolume
maximal faces of(Figure
the maximal
8a). In volume
contrast,so
thethat it creates
boundary a vertically
remains intact
open void running through the maximal volume (Figure 8-a). In contrast, the boundary remains
open
when void running constraint
the boundary through the maximal and
is activated, volume (Figure subtractor
any vertical 8-a). In contrast,
close to the face
boundary remains
of the maximal
intact when the boundary constraint is activated, and any vertical subtractor close to the face of the
intact
volume when the boundary
is moved away fromconstraint
the face isofactivated,
the maximal andvolume
any vertical subtractor
(Figure 8b). close to the face of the
maximal volume is moved away from the face of the maximal volume (Figure 8-b).
maximal volume isthe
Deactivating moved awayconstraint
boundary from the face of the
permits maximal
radical volume
changes (Figure
in the 8-b).massing. Figure 9
building
presents generated building massings with the boundary constraint deactivated. As it can be
seen, massing design variants consisting of two or more completely separate buildings can emerge.
These types of massing design variants allow the optimization process to explore beyond solutions
characterized by one building massing. Such an expansion to the solution space may benefit the
building energy performance, for example, by allowing for inter-block shading among buildings [3].
Figure2019,
Sustainability 8. Generated
11, 6965 building massings with or without the boundary constraint activated: (a) the8 of 20
Figure
Sustainability
boundary 8.2019,
Generated
11, x FORbuilding
constraint massings
PEER REVIEW
disabled, with or constraint
(b) the boundary without the boundary constraint activated: (a) the 8 of 21
enabled.
boundary constraint disabled, (b) the boundary constraint enabled.
Deactivating the boundary constraint permits radical changes in the building massing. Figure 9
Deactivating
presents generated thebuilding
boundary constraint
massings permits
with radical changes
the boundary in the
constraint building massing.
deactivated. As it canFigure 9
be seen,
presents
massinggenerated building
design variants massings
consisting withorthe
of two boundary
more constraint
completely deactivated.
separate buildings As
canitemerge.
can be These
seen,
massing design variants consisting of two or more completely separate buildings
types of massing design variants allow the optimization process to explore beyond solutions can emerge. These
types of massing
characterized design
by one variants
building allow Such
massing. the optimization
an expansion process to explore
to the solution spacebeyond solutions
may benefit the
characterized
building energyby performance,
one building formassing. Such
example, by an expansion
allowing for to the solution
inter-block space
shading maybuildings
among benefit the [3].
Figure8.8. Generated
Figure building
buildingmassings with ororwithout
withoutthe boundary
boundaryconstraint
constraintactivated:
activated: (a)
(a)the
building energyGenerated
performance, for massings
example,with
by allowing forthe
inter-block shading among buildings the
[3].
boundary
boundaryconstraint
constraintdisabled,
disabled,(b)
(b)the
theboundary
boundaryconstraint
constraintenabled.
enabled.
Deactivating the boundary constraint permits radical changes in the building massing. Figure 9
presents generated building massings with the boundary constraint deactivated. As it can be seen,
massing design variants consisting of two or more completely separate buildings can emerge. These
types of massing design variants allow the optimization process to explore beyond solutions
characterized by one building massing. Such an expansion to the solution space may benefit the
building energy performance, for example, by allowing for inter-block shading among buildings [3].
Figure 9. Examples of massing designs generated with the boundary constraint deactivated.
Examplesofofmassing
Figure9.9.Examples
Figure massingdesigns
designsgenerated
generatedwith
withthe
theboundary
boundaryconstraint
constraintdeactivated.
deactivated.
2.2.3. Subtractor
2.2.3. Subtractor Alignment
Alignment
2.2.3. Subtractor Alignment
Alignmentisisanan
Alignment important
important approach
approach by which
by which architects
architects regulate regulate the massing
the massing elementselements
to maintain to
Alignment
maintain is
geometric an important
order [35]. approach
However, by which
arbitrary architects
placements
geometric order [35]. However, arbitrary placements of subtractors can result in poor alignment. regulate
of the
subtractors massing
can elements
result in poor to
maintain
alignment.
For example,geometric
Fortwo
example, ordertwo
subtractors [35]. However,
subtractors
may mayarbitrary
be positioned placements
be positioned
close close of
together, subtractors
together,
creating acreating
space canbetween
aresult
space in thepoor
betweentwo
alignment.
the two For example,
subtractors that two
may subtractors
be too may
narrow be
to positioned
be of any close
use.
subtractors that may be too narrow to be of any use. Thus, in the horizontal direction, the parallel together,
Thus, in the creating
horizontal a space between
direction, the
the two
parallel subtractors
faces from that may
different be too
subtractorsnarrow are to be
aligned of any
to be use. Thus,
co-planar
faces from different subtractors are aligned to be co-planar when the distance between the two faces is in the
when horizontal
the distance direction,
between the
the
parallel
two than
less faces
faces Figurefrom
aishalf
lessspan different
9.than a half
Examples
of oneof subtractors
span of one
massing
column are
column
designs
grid aligned
(Figure grid
generated
10a). to(Figure
be co-planar
with the10-a).
Similarly, when
theSimilarly,
boundary faces thetwo
of distance
the
constraint faces of between
deactivated.
partly two partly
overlapping the
two faces is less
overlapping
subtractors are thanaligned
subtractors
also a halfarespan
also
in theofaligned
one column
same in the
way to grid
same
avoid (Figure
way to
small 10-a).
avoid
jagged Similarly,
small
faces thevoid
in jagged
the faces
faces of two
merged in the partly
by void
two
overlapping
2.2.3.
merged
or subtractors
Subtractor
by two or more
more subtractors Alignment are
(Figure also
subtractors aligned
10b). Lastly, in
(Figure
when the same
10-b). way
Lastly, when
a subtractor to avoid small
a subtractor
has a face jagged
close to the faces
has maximal in
a face close the void
to the
volume’s
merged
maximal byvolume’s
twoface
or more subtractors (Figure 10-b). Lastly, whenthat a subtractor has aNoteface close to only
the
boundary, the
Alignment isisboundary,
aligned
an to the
important face
the approach is aligned
boundary (Figure
by which to thearchitects
10c). boundary
Note (Figure
the 10-c).
last alignment
regulate the massing that
operation the
elements lastto
maximal
alignment
applies volume’s
operation
to vertical boundary,
only
subtractors the
applies
if the face
to is
verticalaligned to
subtractors
boundaryarbitrary constraint the boundary
if the (Figure
boundary
is deactivated. 10-c).
constraint Note
is that the
deactivated. last
maintain geometric order [35]. However, placements of subtractors can result in poor
alignment operation only applies to vertical subtractors if the boundary constraint is deactivated.
alignment. For example, two subtractors may be positioned close together, creating a space between
the two subtractors that may be too narrow to be of any use. Thus, in the horizontal direction, the
parallel faces from different subtractors are aligned to be co-planar when the distance between the
two faces is less than a half span of one column grid (Figure 10-a). Similarly, the faces of two partly
overlapping
Figure 10.subtractors
Figure 10. The procedures
The procedures areof
ofalso aligned
alignment:
alignment: (a)in
(a) the same
aligning
aligning way to avoid
two separated
two separated small(b)
subtractors,
subtractors, jagged
(b) aligning
aligning faces
twoin
two the void
partly
partly
merged
Figure by10.two
overlapping
overlapping or more subtractors
Thesubtractors,
procedures
subtractors, and (c)
(c)aligning
of alignment:
and (Figure
aligning 10-b).two
(a)aaaligning
subtractor
subtractor Lastly,
to
tothe when
themaximal
separated a volume.
subtractor
subtractors,
maximal has a face
volume.(b) aligning close to the
two partly
maximal volume’s
overlapping boundary,
subtractors, and (c) thealigning
face isa aligned
subtractor totothetheboundary (Figure 10-c). Note that the last
maximal volume.
In architectural
alignment operationdesign, apart from
only applies the alignment
to vertical subtractors amongif themassing
boundary elements,
constraint the isplacement
deactivated. of the
In architectural design, apart from the alignment among massing elements, the placement of the
face Inof architectural
massing elements (walls
design, (walls or
apart fromfacades) typically
the alignment also aligns
among with grid lines defined by the modulus
face of massing elements or facades) typically alsomassing
aligns withelements, the placement
grid lines defined by of thethe
related
face of to the column
massing elements grid. (walls
Hence, or thefacades)
faces of all subtractors
typically also are aligned
aligns withto its
gridnearest
lines n/4 (n could
defined by be
the 0)
modulus related to the column grid. Hence, the faces of all subtractors are aligned to its nearest n/4
position in
modulus between twocolumn
adjacent columns (Figure 11). ofWith the constraints are and operations mentioned
(n couldrelated to the
be 0) position grid.
in between Hence, the faces
two adjacent columnsall subtractors
(Figure 11). aligned
With thetoconstraints
its nearest n/4
and
above,
(n many
could bementioned commonly
0) positionabove, unwanted
in between features
two adjacent that may appear
columnsfeatures in the
(Figurethat building
11).may massing
Withappear when
the constraints using
and
operations many commonly unwanted in the building
Sustainability 2019, 11,approaches
parametric x FOR PEER REVIEW are excluded. 9 of 21
operations
massing when mentioned above,
using parametric many commonly
approaches areunwanted
excluded.features that may appear in the building
Figure 10. The procedures of alignment: (a) aligning two separated subtractors, (b) aligning two partly
massing when using parametric approaches are excluded.
overlapping subtractors, and (c) aligning a subtractor to the maximal volume.
In architectural design, apart from the alignment among massing elements, the placement of the
face of massing elements (walls or facades) typically also aligns with grid lines defined by the
modulus related to the column grid. Hence, the faces of all subtractors are aligned to its nearest n/4
Figure Figure
(n could be Alignment
11.0)
11. Alignmentposition operation
in of
operation theof
between thetwo
face offace of subtractors
adjacent
subtractors thetoaxes:
tocolumnsthe axes: assuming
(Figure
assuming 11). the constraint
With
the size size
theconstraint
is is 2 and
constraints
to 5,
operations (a) before being
mentioned aligned
above, to axes,
many and (b)
commonly after being
unwanted aligned to axes.
features
2 to 5, (a) before being aligned to axes, and (b) after being aligned to axes. that may appear in the building
massing when using parametric approaches are excluded.
Despite the fact that the number of subtractors is predefined as an initialization parameter, the
various operations applied during the building massing generation process can cause the actual
various operations applied during the building massing generation process can cause the actual
appearance of the number of parts being subtracted from the building massing to vary significantly.
As a result, building massing with different topological complexity can be generated, from the one
without any part removed to that with the largest number of parts permitted removed. As such, the
generated building massings may present various combinations and arrangements of different
Figure 11. Alignment operation of the face of subtractors to the axes: assuming the size constraint is
2 to 5,Figure 11. being
(a) before Alignment operation
aligned to axes, of
andthe(b)face of being
after subtractors toto
aligned the axes: assuming the size constraint is
axes.
2 to 5, (a) before being aligned to axes, and (b) after being aligned to axes.
Despite the
various operations fact that
applied the number
during of subtractors
the building massingisgeneration
predefinedprocess
as an initialization
can cause the parameter,
actual the
variousofoperations
appearance the numberapplied
of partsduring the building
being subtracted frommassing generation
the building process
massing cansignificantly.
to vary cause the actual
appearance
As a appearance
result, of
building the number
massing of
with parts being
different subtracted
topological from the
complexity building
can be massing
generated,to vary
from
of the number of parts being subtracted from the building massing to vary significantly.significantly.
the one
Asaany
without
As aresult,
result, building
partbuilding massing
removedmassing withthe
to that with
with different topological
largesttopological
different complexity
number of complexity
parts canbe
permitted
can begenerated,
generated,
removed. fromthe
As such,
from theone
the one
without
generated any
building part removed
massings to
may that with
present the largest
various number of
combinations parts
and permitted removed.
arrangements
without any part removed to that with the largest number of parts permitted removed. As such, the of As such,
different the
generated
passive
generated building
design building
strategies, massings
such as may
massings may
several present
smaller
present various
courtyards
various combinations
or one large
combinations and arrangements
courtyard (Figure
and arrangements of different
12). passive
of different
passivestrategies,
design design strategies, such assmaller
such as several severalcourtyards
smaller courtyards or one
or one large large courtyard
courtyard (Figure 12).
(Figure 12).
Figure 12. Examples of massing designs generated with the same number of subtractors.
2.2.4.2.2.4.
Building CoresCores
Building
2.2.4. Building Cores
As widely
As widely applied
applied totomodern building design,
modern building design,building
building cores
cores can serve
can serve as circulation,
as circulation, evacuation,
evacuation, As andwidely applied
structural to
supports.modernIn building
this design,
consideration, building
the
and structural supports. In this consideration, the algorithm provides an optional operation cores
algorithm can
provides servean as circulation,
optional
to create
evacuation,
operation
buildingto create
cores. andThestructural
building
principle supports.
cores.forThe Inthese
principle
creating this consideration,
forcores
creating these
is based the
oncores algorithm
is based of
the footprint onprovides
thebuilding
the an massing
footprint optional
of
operation
the building to
massingcreateand building
complies cores.
with The
the principle
firefighting for creating
evacuation these cores
regulation
and complies with the firefighting evacuation regulation in China. In China, the door-to-door distance is based
in on
China. the
In footprint
China, of
the building
the door-to-door
from a room distance massing and complies
fromevacuation
to its closest a room to stairwith the
its closest firefighting
and the evacuation evacuation
distance stair
of anyand regulation
the in
point distance
the room in China.
of any
to itspointIn
closest China,
in exit
the door-to-door
the room
door to corridors
to its closest distance befrom
exit door
should acorridors
notomore roomthanto22 its closest
should
m. beevacuation
Since no proposed
the more thanstair andm.
22
algorithm theSince
distance
doesthe of any point
notproposed
subdivide in
the
the
algorithm room
does to
not its closest
subdivide exit
the door
floor to
plancorridors
into should
rooms, the be no more
calculation than
of
floor plan into rooms, the calculation of the number and position of the building cores is simplified tothe 22 m.
number Since
and the proposed
position
of the algorithm
building
fulfill does not
cores
a requirement subdivide
is simplified
that everyto the floor
fulfill
face aplan
of the into rooms,
requirement
maximal thatthe
volume’s calculation
every of
face oftothe
footprint itsthe number
maximal
closest and
volume’s
cores is noposition
more
of the
footprint
than 35building
to its Whencores
m.closest cores isissimplified
the positionno more
of the tocores
than fulfill
35 is aWhen
requirement
m.determined, thethat
the position every
of the
column face
gridcores ofisthe
cell maximalorvolume’s
determined,
containing the
close to
column footprint
grid
(equal orcellto containing
less its
thanclosest
half ofcores
orthe is notomore
close (equal
column-grid than 35the
or less
span) m. When
than
center halfthe
of position
point the of the
column-grid
of any core cores
span)
is turned isinto
determined,
thethecenter
buildingthe
column
pointcore.
of any
After grid
coretheiscell containing
turned
building into
coresthe orbuilding
close
have been togenerated,
(equalAfter
core. or less
allthe than half of
building
subtractors the
cores
are also column-grid
have span)with
been generated,
required to align theallcenter
these
point
subtractors of
are any
also
cores (Figure 13). core is
requiredturnedto into
align the
with building
these core.
cores After
(Figure the
13). building cores have been generated, all
subtractors are also required to align with these cores (Figure 13).
Figure 13. Generated

Figure building
13. Generated massings
building with or
massings without
with building
or without cores:cores:
building (a) building core generation
(a) building core generation
Figure
disabled, (b) 13. Generated
building core building
generationmassings
enabled. with
disabled, (b) building core generation enabled. or without building cores: (a) building core generation
disabled, (b) building core generation enabled.
2.3. Implementation
2.3. Implementation
2.3. Implementation
In order
In order to facilitate
to facilitate the ease
the ease of use,
of use, the algorithm
the algorithm was was implemented
implemented as a as a plug-in
plug-in component
component
(FigureIn order
14a) in to facilitate the
Rhino-Grasshopper,ease of use,
which the
is algorithm
one of the was
most implemented
popular
(Figure 14-a) in Rhino-Grasshopper, which is one of the most popular 3D and parametric modelling 3D and as a plug-in
parametric component
modelling
(Figure
toolstools 14-a) in [36].
in architecture
in architecture Rhino-Grasshopper,
[36].
A userA user which
interface
interface is14-b)
one of
(Figure
(Figure thewas
14b)
was most
also popular
also 3D
integrated
integrated and
into theparametric
into modelling
the component
component for for
toolstoinspecify
architects
architects architecture
to specify [36].
the A userparameters
interface
initialization
the initialization (Figure
parameters and14-b)
and other otherwas also integrated
attributes
attributes to thetointo
related
related thethe component
building
building massing
massing for
design,architects
design,
suchsuch to specify
as floor
as floor the initialization
height,
height, spanspan parameters
size (width),
size (width), and other
and facade
and facade types. attributes
types.
As toAs related
thetofacade to
the facade the building
type,type,
only only massing
curtain
curtain
design,
wallswalls
andand such
strip
strip as floor height,
windows
windows span size
areprovided
are provided in(width),
the andimplementation,
thecurrent
current facade types. As
implementation, towhich
which thecould
facade type,
be further
could only curtain
expanded
be further
walls
in the and
future. strip
Withwindows
the are
component
expanded in the future. With the component and provided
and thein the
user current
interface, implementation,
architects can which
interact could
with
user interface, architects can interact with the thebe further
building
expanded
massing in the process
creation future. With the component
and observe the changeand in
thethe
user interface,
renewed architects
building massingcan interact with
instantly the
in the
Rhino-Grasshopper environment. The timely visual feedback of the generated building massings
allows architects to undertake parameter tuning to exclude unwanted or unsuitable features in the
building massing and ensure that these features do not appear in evolutionary optimization.
The implementation in the Rhino-Grasshopper environment takes advantage of the Visual
Programming Language, where no coding is required. It allows architects, even computational design
novices, to use the algorithm and establish a performance-based optimization system for building
massing design in a plug-and-play fashion. The implementation of the proposed algorithm is part
of an integrated evolutionary design toolkit, which also embeds a diversity-driven evolutionary

building massing creation process and observe the change in the renewed building massing instantly
in the Rhino-Grasshopper environment. The timely visual feedback of the generated building
algorithmallows
massings [37]. As such, ato
architects performance-based optimization
undertake parameter tuning tosystem canunwanted
exclude be established simply by
or unsuitable
connecting
features thebuilding
in the component of this and
massing building massing
ensure algorithm
that these anddo
features thenot
evolutionary
appear in algorithm with
evolutionary
various building performance simulation tools, such as DIVA, Ladybug, Honeybee, Archsim, etc. [38].
optimization.
14.(a)(a)
Figure14.
Figure Component
Component implemented
implemented inin Rhino-Grasshopper,(b)
Rhino-Grasshopper, (b)the
theuser
userinterface.
interface.
2.4. Case Studies

The implementation in the Rhino-Grasshopper environment takes advantage of the Visual
Programming Language, where no coding is required. It allows architects, even computational
2.4.1. Design Setting
design novices, to use the algorithm and establish a performance-based optimization system for
buildingTo massing
demonstratedesign the
in efficacy of the proposed
a plug-and-play algorithm,
fashion. The two case-study
implementation performance-based
of the proposed algorithm
is part of an integrated evolutionary design toolkit, which also embeds a studies
optimization processes for building massing design were carried out. The two case respectively
diversity-driven
describe a high-rise
evolutionary algorithm slab-type
[37]. Asbuilding
such, adesign and a middle-rise
performance-based deep plansystem
optimization building design
can (Figure 15).
be established
The difference between these two building design objects is to clarify the
simply by connecting the component of this building massing algorithm and the evolutionarycapability of the proposed
algorithm to handle different settings of building massing design. For both case
algorithm with various building performance simulation tools, such as DIVA, Ladybug, Honeybee, studies, the energy
performance
Sustainability
Archsim, objective
etc.2019, 11, x FORwas
[38]. PEERtoREVIEW
maximize daylighting for passive energy savings. 11 of 21
2.4. Case Studies
2.4.1. Design Setting

To demonstrate the efficacy of the proposed algorithm, two case-study performance-based
optimization processes for building massing design were carried out. The two case studies
respectively describe a high-rise slab-type building design and a middle-rise deep plan building
design (Figure 15). The difference between these two building design objects is to clarify the
capability of the proposed algorithm to handle different settings of building massing design. For both
case studies, the energy performance objective was to maximize daylighting for passive energy
savings.
Figure 15. Plan

Planand aerial
and view
aerial of of
view thethe
building plot:plot:
building the orange blockblock
the orange indicates the building
indicates being
the building
designed.
being designed.
The building plot for the two case studies is located at the campus of Nanjing University in
Nanjing City, Jiangsu Province, China. Several high- and middle-rise buildings surround the building
plot and cast shadows on the plot (Figure 16), thereby presenting a significant challenge in achieving
desired daylighting. Without the assistance of performance-based optimization, architects may
conceive high-performing solutions by manual trial and error. However, this is not only time-
consuming and laborious, but the solution is likely to be prejudiced due to architects’ cognitive biases.
Figure 15. Plan and aerial view of the building plot: the orange block indicates the building being
designed.
The
The building
building plotplot for two
for the the case
two studies
case studies is located
is located at the campus
at the campus of Nanjingof University
Nanjing University
in Nanjingin
Nanjing
City, City,
Jiangsu JiangsuChina.
Province, Province, China.
Several Several
high- high- and middle-rise
and middle-rise buildingsthe
buildings surround surround the
building building
plot and
plot and cast shadows on the plot (Figure 16), thereby presenting a significant challenge
cast shadows on the plot (Figure 16), thereby presenting a significant challenge in achieving desired in achieving
desired daylighting.
daylighting. Without the Without the assistance
assistance of performance-based
of performance-based optimization,optimization, architects
architects may may
conceive
conceive high-performing
high-performing solutions bysolutions by and
manual trial manualerror.trial and error.
However, However,
this is this is not only and
not only time-consuming time-
consuming
laborious, butand
the laborious,
solution isbut the to
likely solution is likely due
be prejudiced to betoprejudiced
architects’due to architects’
cognitive biases.cognitive biases.
Figure Sun
16.16.
Figure path
Sun ofof
path the building
the plot
building (winter
plot solstice).
(winter solstice).
For the optimization objective, the fitness evaluation considers both the energy reduction resulting
For the optimization objective, the fitness evaluation considers both the energy reduction
from daylighting and the gross area of the building massing. For the energy reduction from daylighting,
resulting from daylighting and the gross area of the building massing. For the energy reduction from
annual lighting energy (LE) consumption is taken as the performance indicator, and the general
daylighting, annual lighting energy (LE) consumption is taken as the performance indicator, and the
objective of the evolutionary optimization is to minimize this value. At the same time, the gross area
general objective of the evolutionary optimization is to minimize this value. At the same time, the
of the building massing considers a penalty function in the fitness calculation. While the proposed
gross area of the building massing considers a penalty function in the fitness calculation. While the
algorithm can automatically adjust the dimension and floor number of the building massing to make
proposed algorithm can automatically adjust the dimension and floor number of the building
the gross area close to the required target value, there could be an unacceptable gross area difference if
massing to make the gross area close to the required target value, there could be an unacceptable
the original difference is too large. Hence, the difference is considered a penalty function to punish
gross area difference if the original difference is too large. Hence, the difference is considered a
unsatisfactory design variants. In this regard, the value of lighting energy increases proportionally with
penalty function to punish unsatisfactory design variants. In this regard, the value of lighting energy
the increase in the difference between the actual gross area and the required gross area. Therefore, even
increases proportionally with the increase in the difference between the actual gross area and the
with low lighting energy, the building massing still receives an unfavorable fitness value when the gap
required gross area. Therefore, even with low lighting energy, the building massing still receives an
between its gross area and the target value is too large. The fitness function can be described below:
Aactual − Atarget
!
f itness = LE × 1 + | | , (1)
Atarget
where LE indicates the lighting energy consumption, and Aactual and ATarget indicate the actual gross
area of the building massing and the required target gross area.
The annual lighting energy consumption is calculated by a Radiance-based simulation tool, called
DIVA [39], in Rhino-Grasshopper. In order to ensure each simulation can be finished in a reasonable
timeframe, the lowest simulation quality is adopted, which means that the Radiance engine calculates
a lower number of bounces of rays for the simulation. Using a higher quality makes the simulation
too time-consuming, since there might be thousands of simulations involved in one evolutionary
optimization process. Moreover, although the lowest quality degrades the accuracy of the simulation,
the bias to the overall optimization result is not significant. This is because the fitness ranking among
high-performing and low-performing design variants calculated by the lowest simulation quality
corresponds to that calculated by the higher quality according to our observation. With the lowest
simulation quality applied, each simulation lasts from 1 to 2 min (HP Z440 Workstation with a Xeon
4-core CPU). The typical time for one evolutionary optimization process reaching the required number
of design generations and performance evaluations (4980, to be exact, in the case study) is around 4 to
5 days. While the time consumed by the optimization process cannot satisfy the time constraint set
by practice, this can be addressed with an increase in computational power and incorporating other
approaches, such as parallel computing and cloud computing.
With regard to the optimization algorithm, the case study uses a newly developed hybrid
evolutionary algorithm designed to support explorative design optimization, called SSIEA (steady-state
island evolutionary algorithm) [37]. This algorithm can yield several high-performing solutions, while
the solutions also have enhanced design differentiations. The algorithm integrates an island-model
approach and a steady-state replacement strategy into a standard evolutionary algorithm to increase
diversity in the design population and improve the search efficiency. The island model allows the
algorithm to launch multiple parallel search processes to evolve several “niching” subpopulations.
For each subpopulation, it is guided to focus on a different region in the design solution space.
The separation and multitude of subpopulations counteract the exploitative natural inherited in the
standard evolutionary algorithm and prevents the optimization from producing a family of very similar
design solutions due to the genotypic similarity. At the same, the steady-state replacement strategy
speeds up the optimization process by increasing the evolutionary pressure on each subpopulation. We
examined the algorithm in a previous study [37], which indicates that this algorithm can significantly
outperform the standard genetic algorithm. Incorporating this algorithm with the proposed building
massing generation algorithm can retrieve several high-performing solutions from the optimization
process. Furthermore, these solutions have significant topological differentiation due to the genotypic
diversity in the individuals from different subpopulations. The diversity in the result enhances the
feedback from optimization, which helps architects discover underlying compromises and trade-offs
characterizing the design problem.
For the setting of SSIEA, six subpopulations (islands) are set for the optimization process in
order to achieve a relatively high diversity in the optimization result. Each subpopulation has 50
individuals as the initial size, and, thereby, the overall initial population size is 300. In each iteration,
10 individuals are randomly selected from each subpopulation, and six higher-ranking individuals are
selected as parents by tournament selection to reproduce an equal number of offspring individuals by
crossover and mutations. With all offspring individuals evaluated by simulations, the 10 originally
selected individuals compete against the offspring individuals, and inferior ones are then replaced
by higher-ranking offspring individuals. Lastly, 130 iterations are set as the termination criteria, so
that there are 4980 design generations and evaluations in each optimization process as a result, which
ensures that the design can be sufficiently evolved.
2.4.2. Initialization Parameters

With the proposed algorithm, the first step is to specify the initialization parameters. In order to
widen the scope of exploration, the boundary constraint is deactivated to achieve a relatively radical
change in the generated building massings. Then, we specify the parameters related to the attribute of
two building massing designs, including total floor numbers, numbers of column grids in the x- and
y- directions, floor heights, and column-grid span sizes. For the second case study, the building plot
shown in Figure 15 is not suitable for a rectangular footprint, and, therefore, an L-shaped footprint
is set for the maximal volume and divides the building massing into a west wing and a south wing.
This part of the setting will be presented at the end of this section together with other initialization
parameters in the user interface.
Apart from the parameters mentioned specifying the attributes of the building massing design,
we also compare three initialization parameter setups varying the size constraints and the number of
subtractors (Table 1). For the first case study, Figure 17 illustrates the generated building massings
for the three initialization modeling parameter setups. From the first row in Figure 17, it can be seen
that if the number of subtractors and the size constraint are set too small, the building massings will
not embody significant topological variability. This is despite the fact that the boundary constraint
is deactivated, which will improve topological variability. For the other two setups, the features
appearing in the generated building massings are mostly similar as shown in the second and third rows
in Figure 17. Considering that a larger number of subtractors applied to the third setup expands the
search space and increases the search difficulty for evolutionary optimization, the second initialization
parameter setup is selected for the subsequent execution of the optimization process.
deactivated, which will improve topological variability. For the other two setups, the features
appearing in the generated building massings are mostly similar as shown in the second and third
rows in Figure 17. Considering that a larger number of subtractors applied to the third setup expands
the search space and increases the search difficulty for evolutionary optimization, the second
initialization parameter
Sustainability 2019, 11, 6965 setup is selected for the subsequent execution of the optimization
13 process.
of 20
Table 1. Initialization parameters.

Table 1. Initialization parameters.
Number of Subtractors Size Constraint
Number of Subtractors Size Constraint
Vertical subtractors Horizontal subtractors Vertical subtractors Horizontal subtractors
Vertical Subtractors Horizontal Subtractors Vertical Subtractors Horizontal Subtractors
Setup 1 2 3 [2, 3] [2, 3]
Setup 1 2 3 [2, 3] [2, 3]
Setup Setup
2 2 3 3 44 [3,
[3, 4] 4] [3, 5] [3, 5]
Setup Setup
3 3 4 4 55 [3, 5]
[3, 5] [3, 5] [3, 5]
Figure17.
Figure Randomsampling
17.Random sampling generated
generatedbuilding
buildingmassings
massingsby the
by initialization parameter.
the initialization parameter.
For the second case study, Figure 18 shows the generated building massings, also based on the
Forinitialization
three the secondparameter
case study, Figure
setups. 18 shows
In this the
case, the generated
second setup inbuilding
Table 1 ismassings,
selected asalso based
it results in on the
three initialization
reasonable parameter
topological setups.
variability. In this the
In contrast, case, the second
generated setup
building in Tableby1the
massings is selected
first setupas(the
it results
in reasonable topological
first line in Figure 18) are variability. In contrast,
relatively similar. thegenerated
For those generated bybuilding
the third massings
setup (in thebythird
the first
line setup
(the first line in Figure 18) are relatively similar. For those generated by the third
in Figure 18), the possibility of creating two separate building massing entities may be not suitable forsetup (in the third
Sustainability
line
theinfunctional
Figure 2019, 11,the
18), x FOR PEER REVIEW
possibility
requirement in this of due
case creating
to thetwo separate
lack of building
accessibility massing entities may be 14
and connectivity. of 21
not
suitable for the functional requirement in this case due to the lack of accessibility and connectivity.
Figure 18.Random
Figure18. Randomsampling generated
sampling building
generated massings
building by theby
massings initialization parameter.
the initialization parameter.
The process of setting up the optimization algorithm is straightforward for the user and does
The process of setting up the optimization algorithm is straightforward for the user and does
not require any parametric modelling to be performed. Instead, the user only needs to fill in certain
not require any
initialization parametric
parameter modelling
settings to be user
in a graphical performed.
interfaceInstead, the user
(GUI). Based only
on the needs
above to fill in certain
parameter
initialization
setup, the userparameter
interface forsettings in astudies
the two case graphical user
is filled outinterface
as below (GUI).
(Figure Based
19). Theon the above
difference parameter
in the
setup, the user interface for the two case studies is filled out as below (Figure 19). The
settings on the user interface between the two case studies is highlighted. The basic attributes specify difference in
the
the settings
parametersonrelated
the user interface
to floor between
numbers, span the two case
numbers, etc. studies is highlighted.
For the second case study,The basic
a fixed cutattributes
specify
volume the parameters
in the relatedistoset
building massing floor numbers,
to create span numbers,
the L-shaped footprint.etc. For the second case study, a fixed
cut volume in the building massing is set to create the L-shaped footprint.
not require any parametric modelling to be performed. Instead, the user only needs to fill in certain
initialization parameter settings in a graphical user interface (GUI). Based on the above parameter
setup, the user interface for the two case studies is filled out as below (Figure 19). The difference in
the settings on the user interface between the two case studies is highlighted. The basic attributes
specify the parameters related to floor numbers, span numbers, etc. For the second case study, a fixed
cut volume in the building massing is set to create the L-shaped footprint.
Figure 19. The

The setting
setting on
on the
the user
user interface.
interface.
3. Results
3. Results
3.1. Case Study 1

3.1. Case Study 1
The first case study describes a slab-type high-rise office building with a target gross area of
The 2first case study describes a slab-type high-rise office building with a target gross area of
45,000 m . We take the building massing shown in Figure 15 without any part removed as the
45,000 m2. We take the building massing shown in Figure 15 without 2any part removed as the
benchmark. The gross area of the benchmark building massing is 43,740 m . The LE of the benchmark
benchmark. The gross area of the benchmark building massing is 43,740 m2. The LE of the benchmark
building massing is 813,811 kWh, and the fitness after being penalized by the gross area difference is
building massing is 813,811 kWh, and the fitness after being penalized by the gross area difference is
836,597.
836,597.
Figure 20 shows the elite individual in each subpopulation found by the optimization process.
Figure 20 shows the elite individual in each subpopulation found by the optimization process.
In this case study, the surrounding buildings prevent a large amount of daylight from reaching the
In this case study, the surrounding buildings prevent a large amount of daylight from reaching the
building massing (Figure 16) and cast shadows on the facade of the building, which can undermine
building massing (Figure 16) and cast shadows on the facade of the building, which can undermine
the daylighting quality on the floor with windows frequently shaded by the shadow. In order to
the daylighting quality on the floor with windows frequently shaded by the shadow. In order to
improve daylighting, three dominant features can be found among elite individuals. First, stilts appear
improve daylighting, three dominant features can be found among elite individuals. First, stilts
in the building massing of Elite1, Elite4, and Elite5, which raise the overall building mass. Raising the
appear in the building massing of Elite1, Elite4, and Elite5, which raise the overall building mass.
building mass reduces the space close to the ground, which naturally has worse daylighting quality
Raising the building mass reduces the space close to the ground, which naturally has worse
due to heavy
Sustainability daylight
2019, 11, x FORobstructions
PEER REVIEWby the surrounding buildings. 15 of 21
daylighting quality due to heavy daylight obstructions by the surrounding buildings.
Figure 20. Elite individuals of natural daylighting (south-west aerial view).

view).
Second, cases, such

Second, cases, suchasasElite2,
Elite2,Elite5,
Elite5,andand Elite6,
Elite6, appear
appear as aastower-type
a tower-type building
building massing.
massing. The
The reason for this is that tower-type building massing increases the indoor floor
reason for this is that tower-type building massing increases the indoor floor area at a higher altitude area at a higher
altitude where
where there arethere aredaylight
fewer fewer daylight obstructions
obstructions due todue to surrounding
surrounding buildings.
buildings. Figure
Figure 21 illustrates
21 illustrates the
the natural lit space in a cross-sectional direction. Note that the natural-lit space
natural lit space in a cross-sectional direction. Note that the natural-lit space (warmer red-colored(warmer red-colored
areas)
areas) increases
increases as as the
the floor
floor level
level is
is raised. The first
raised. The first and
and second
second features
features suggest
suggest that
that the
the overall
overall
performance
performance of of daylighting
daylighting can can bebe improved
improved by by manipulating
manipulating the the building
building massing
massing in in the
the vertical
vertical
direction—either
direction—eitherreducingreducingthethe proportion
proportionof lower floor levels
of lower floor as in Elite1
levels as inor increasing the proportion
Elite1 or increasing the
of higher floor levels as in Elite6. In contrast, the third feature manipulates
proportion of higher floor levels as in Elite6. In contrast, the third feature manipulates the the building massing
buildingin
the horizontal
massing in thedirection.
horizontalAll elite individuals
direction. tend to havetend
All elite individuals a stepped
to havefootprint,
a steppedwhich can enlarge
footprint, which canthe
spacing with the south high-rise building as well as increase the facade area exposed
enlarge the spacing with the south high-rise building as well as increase the facade area exposed to to daylight [40].
The [40].
daylight effect of the optimization on daylighting is significant. The elites achieve an average 96%
improvement compared with the benchmark building massing. It means that, for this case study,
the performance of daylighting is sensitive to the change in the building massing. This finding also
suggests that while there is great potential in building massing design to achieve better daylighting,
where there are fewer daylight obstructions due to surrounding buildings. Figure 21 illustrates the
natural lit space in a cross-sectional direction. Note that the natural-lit space (warmer red-colored
areas) increases as the floor level is raised. The first and second features suggest that the overall
performance of daylighting can be improved by manipulating the building massing in the
Sustainability 2019, 11, 6965
vertical
15 of 20
direction—either reducing the proportion of lower floor levels as in Elite1 or increasing the
proportion of higher floor levels as in Elite6. In contrast, the third feature manipulates the building
inappropriate
massing building massing
in the horizontal design
direction. All can
elitealso lead to much
individuals intensive
tend to havelighting energy
a stepped consumption.
footprint, which can
Lastly, the elites with a tower-like massing may be inappropriate when a rigorous urban planning
enlarge the spacing with the south high-rise building as well as increase the facade area exposed to
code regulates the height of the building. However, this helps to enhance the tendency of raising the
daylight [40].
building massing for better daylighting for this case study.
Figure 21.
Figure Daylight autonomy
21. Daylight autonomyanalysis by by
analysis floors.
floors.
3.2. Case Study 2
The effect of the optimization on daylighting is significant. The elites achieve an average 96%
The second
improvement case study
compared withdescribes a middle-rise
the2 benchmark deep-plan
building massing. building
It meansfor multi-purpose use study,
that, for this case with the
a target gross area of 100,000 m . Similar to the first case study, the building massing without any
performance of daylighting is sensitive to the change in the building massing. This finding also
part removed is taken as the benchmark, as shown in Figure 15. The gross area of the benchmark
suggests that while there is great potential in building massing design to achieve better daylighting,
building massing is 102,332 m2 . The LE of the benchmark building massing is 1.9334 × 106 kWh, and
inappropriate building
the fitness after being massing
penalizeddesign can also
by the gross arealead toismuch
factor 1.9785intensive lighting with
× 106 . Compared energy
the consumption.
first case
Lastly, the elites with a tower-like massing may be inappropriate when a rigorous
study, the change in the building plot allows the south wing of the building to be exposed urban planning
to direct
code regulates
sunlight the height
without of thebybuilding.
obstructions However,
other buildings this16).
(Figure helps tothe
Thus, enhance
buildingthemassing
tendency of raising
should react the
building massing
differently for better
compared daylighting
with the for this case study.
first case study.
Figure 22 shows the elite individual in each subpopulation. As the deep-plan building is
3.2. unfavorable
Case Study 2to daylighting, a common approach reflected in the elite building massing to overcome
this drawback is to reduce the depth of the floor plan by inserting subtractors. This approach can result
The second case study describes a middle-rise deep-plan building for multi-purpose use with a
in two dominant features found among these elite building massings. First, light wells appear in Elite1
target
andgross
Elite5,area
whichof 100,000 m2. Similar
allow daylight to the
to reach theinner
first case study,
part of the building
the building. Second,massing
there iswithout
a setbackany
in part
removed is taken as the benchmark, as shown in Figure 15. The gross area of the
the south wing of the building massing (the blue-colored volume shown in Figure 23), which allows benchmark building
massing is 102332
the floor m2. Thethinner
plan to become LE of the
and benchmark
escape from thebuilding
shadowmassing is 1.9334
cast by the × 10high-rise
south-west 6 kWh, and the fitness
building.
after being
Apart frompenalized
these twoby the gross
features, area factor
horizontal is 1.9785
carve-outs × 10 .inCompared
appearing 6 withofthe
the south wing first
Elite1 andcase study, the
stepped
roofs in Elite6 also help to reduce the depth of the floor plan and increases the
change in the building plot allows the south wing of the building to be exposed to direct sunlightfacade area exposed to
daylight.
without Similar to the
obstructions by first casebuildings
other study, stilts(Figure
also appear
16). inThus,
Elite1 the
and Elite4 due to
building the sameshould
massing reason react
explained in the first case study.
differently compared with the first case study.
In summary, the elite building massing achieves an average 95.6% improvement in the overall
Figure 22 shows the elite individual in each subpopulation. As the deep-plan building is
performance compared with the benchmark building massing. Concerning the features, thin plans
unfavorable to daylighting, a common approach reflected in the elite building massing to overcome
stand out in this case study as deep-plan buildings are naturally inferior in daylighting performance.
thisIndrawback is to reduce the depth of the floor plan by inserting subtractors. This approach can
contrast, courtyards, as a commonly adopted strategy for daylighting, do not appear in this case
study, which is mostly because a courtyard often occupies a large amount of building floor area, and
its contribution is not sufficiently significant to offset its disadvantage of a loss of floor area. Last, but
not least, Elite1 shows a mixture of different passive energy-saving strategies, including self-shading,
stilts, and light wells. It suggests that synergizing these strategies allows for significant progressive
improvement. Such mixtures of strategies are challenging to conceive by human architects and unlikely
to be found when the optimization is confined to building massing with limited topological variability.
setback in the south wing of the building massing (the blue-colored volume shown in Figure 23),
which allows the floor plan to become thinner and escape from the shadow cast by the south-west
which allows the floor plan to become thinner and escape from the shadow cast by the south-west
high-rise building. Apart from these two features, horizontal carve-outs appearing in the south wing
high-rise building. Apart from these two features, horizontal carve-outs appearing in the south wing
of Elite1 and stepped roofs in Elite6 also help to reduce the depth of the floor plan and increases the
of Elite1 and stepped roofs in Elite6 also help to reduce the depth of the floor plan and increases the
facade area exposed to daylight. Similar to the first case study, stilts also appear in Elite1 and Elite4
facade area2019,
Sustainability
exposed to daylight. Similar to the first case study, stilts also appear in Elite1 and 16
Elite4
due to the same 11, 6965 explained in the first case study.
reason of 20
due to the same reason explained in the first case study.
Figure 22. Elite individuals of daylighting (south-east aerial view).

Figure 22. Elite individuals of daylighting (south-east aerial view).
Figure23.
Figure Set-backininthe
23.Set-back thesouth
southwing
wingofofthe
thebuilding
buildingmassing
massing(south-east
(south-eastaerial
aerialview).
view).
Figure 23. Set-back in the south wing of the building massing (south-east aerial view).
4. Discussion
In summary, the elite building massing achieves an average 95.6% improvement in the overall
In summary,
The case the
twocompared elite allowed
studies buildingthe massing
efficacy achieves an averagealgorithm
of the massing.
proposed 95.6% improvement in the overall
performance with the benchmark building Concerning to thebefeatures,
investigated under
thin plans
performance
different design compared
settings, with
in termsthe ofbenchmark
surrounding building
buildings massing.
and Concerning
building design the features,
objects. thin plans
Strengthened by
Inthe explorative
contrast, characterasofa the
courtyards, proposedadopted
commonly evolutionary algorithm,
strategy the optimization
for daylighting, processin
do not appear successfully
this case
In contrast,
explored and courtyards, as a commonly adopted strategy foranddaylighting, do not appear in this case
study, which isevaluated
mostly becausedifferent building
a courtyard massing
often occupiesvariants a large identified
amount ofmultiple
buildinghigh-performing
floor area, and
study, which
solutions is mostly
for each design because
scenario. a courtyard
Due to the often occupies a large amount of building floor massings
area, and
its contribution is not sufficiently significant to extensive topological
offset its disadvantage variability,
of a loss ofthe building
floor area. Last, but
its contribution
retrieved from showsis not
the sufficiently
optimization significant
present to offset its disadvantage
significant of aThe
lossdifference
of floor area. Last, but
not least, Elite1 a mixture of different passive design differentiation.
energy-saving strategies, among
including self-shading, these
not least,
elite and Elite1
variants shows a mixture of different passive energy-saving strategies, including self-shading,
stilts, lighthelps
wells.toItreveal
suggests performance trade-offs
that synergizing theseand compromises.
strategies allows for At significant
the same time, without
progressive
stilts,
the andtolight
need performwells.anyIt suggests
actual that synergizing
parametric modelling, these
the strategies
process allows
of carryingfor out
significant progressive
performance-based
improvement. Such mixtures of strategies are challenging to conceive by human architects and
improvement.is streamlined.
optimization Such mixtures of strategies are challenging thetoinitialization
conceive byparameters
human architects restrictand
unlikely to be found when theArchitects optimization onlyisneed to specify
confined to building massing with limitedto topological the
unlikely
overall to be
geometricfound when
feature inthe optimization
the generated is confined
building to building
massings. In massing
this regard, with
the limited
parametric topological
model
variability.
variability.
based on the algorithm can be viewed as a “meta-model” of the subtractive form generation principle,
4.from which various task-specific versioning models for different projects can be readily derived.
Discussion
4. Discussion
While the proposed algorithm addresses the two challenges identified in the first section, there are
The two case studies allowed the efficacy of the proposed algorithm to be investigated under
threeThekey two
issues case
andstudies
limitationsallowed
to bethe efficacy ofFirst,
highlighted. the proposed
the generated algorithm
building tomassings
be investigated under
are all created
different design settings, in terms of surrounding buildings and building design objects.
different
from design geometries.
orthogonal settings, inIn terms of surrounding
this regard, buildings and
the optimization-based building
exploration withdesign objects.
the proposed
Strengthened by the explorative character of the proposed evolutionary algorithm, the optimization
Strengthened
algorithm canby betheseenexplorative
as an initial character of theof
exploration proposed evolutionary
diverse building algorithm,
massings the optimization
but without in-depth
process successfully explored and evaluated different building massing variants and identified
process successfully
exploitation explored
of the accurate and evaluated
solution under each different
specificbuilding
building massing
massing variants and identified
configuration. In other
multiple high-performing solutions for each design scenario. Due to the extensive topological
multiple
words, thehigh-performing
result of the optimization solutionscan forbeeach
further design scenario.
improved Duegeometric
if higher to the extensive
freedom is topological
provided
variability, the building massings retrieved from the optimization present significant design
variability,
by incorporatingthe building
transformation massings retrieved
approaches, suchfrom the rotation,
as twist, optimization
and slantpresent significant
[22,23,41]. design
Incorporating
differentiation. The difference among these elite variants helps to reveal performance trade-offs and
differentiation.
these The difference
in the algorithm requiresamong thesevariables
additional elite variants helpstransformation
to define to reveal performance operations.trade-offs
Sinceand
the
compromises. At the same time, without the need to perform any actual parametric modelling, the
compromises.
algorithm At the
already hassame time, without
a relatively the need
large number to perform any
of optimization actual parametric
parameters, adding extra modelling,
parameters the
process of carrying out performance-based optimization is streamlined. Architects only need to
process
may of carrying
be ineffective outthe
since performance-based
increase in the size of optimization is streamlined.
the design solution space often Architects
impedes only need to
the discovery
specify the initialization parameters to restrict the overall geometric feature in the generated building
specify
of the initialization
high-performing parameters
solutions to restrict
[28]. Instead, it isthe overall
more geometric
advisable feature in
to separate thethe generated building
optimization process
massings. In this regard, the parametric model based on the algorithm can be viewed as a “meta-
massings.
into In thisInregard,
two stages. the first the parametric
stage, the coarsemodel basedmassing
building on the algorithm
can be exploredcan be while,
viewedinas thea second
“meta-
model” of the subtractive form generation principle, from which various task-specific versioning
model”
stage, of thenumber
a small subtractive form generation
of massings reflecting the principle,
specific from which various
architectural task-specific
implications versioning
can be selected, and
models for different projects can be readily derived.
models for different
transformation projects
operations cancan
thenbe be
readily
appliedderived.
to evolve more in-depth solutions.
Second, in the context of evolutionary optimization, the arbitrary placement of subtractors creates
a many-to-one relationship between the optimization parameters lists and the corresponding output.
This relationship defines the genotype–phenotype mapping. For example, although two design
variants (phenotypes) have similar topological configurations, the two variants could be generated
from two entirely different optimization parameter lists (genotypes). This is because the two different
parameter lists contain the segment (chromosomes) at different positions with similar or identical
values, and this segment results in a subtractor in similar spatial positions and with similar dimensions.
creates a many-to-one relationship between the optimization parameters lists and the corresponding
output. This relationship defines the genotype–phenotype mapping. For example, although two
design variants (phenotypes) have similar topological configurations, the two variants could be
generated from two entirely different optimization parameter lists (genotypes). This is because the
two different parameter lists contain the segment (chromosomes) at different positions with similar
or identical values, and this segment results in a subtractor in similar spatial positions and with
similar dimensions. Figure 24 shows the parallel coordinate visualizing the parameter lists of the elite
design 24
Figure variants.
shows the Compared with the similarity
parallel coordinate visualizingamong the formal
the parameter listsrepresentations (phenotypic
of the elite design variants.
representation),
Compared with theas similarity
shown in among
Figure 20
theand Figure
formal 22, the parameter
representations lists (genotypic
(phenotypic representation)
representation), as shown
doFigures
in not show 20 equivalent
and 22, the similarity.
parameter This
lists many-to-one mapping could
(genotypic representation) dobe problematic
not for evolutionary
show equivalent similarity.
optimization
This many-to-one as itmapping
may result in early
could entrapmentfor
be problematic into a local optimum
evolutionary [28]. However,
optimization as it mayinresult
this research,
in early
the many-to-one
entrapment into a mapping may [28].
local optimum help However,
to clarify in
dominant features
this research, when these features
the many-to-one mappingrepeatedly
may help
appear
to clarifyindominant
many subpopulations.
features when these features repeatedly appear in many subpopulations.
Figure 24. Parallel

Figure 24. Parallel coordinate
coordinate visualizing
visualizing the
the parameters
parameters of
of elite
elite design
design variants.
variants.
Third,
Third, instead
instead ofof considering
consideringone oneobjective
objectiveasasinin the
the case
case studies,
studies, architects
architects cancan alternatively
alternatively use
use multi-objective approaches, such as Pareto optimization, to include multiple
multi-objective approaches, such as Pareto optimization, to include multiple objectives. Using multi- objectives.
Using multi-objective
objective optimization optimization can show trade-offs
can show trade-offs and compromises
and compromises among conflicting
among conflicting objectives.
objectives. While
While using multi-objective optimization has become popular in the literature,
using multi-objective optimization has become popular in the literature, it should be noted that itit should be noted
that
mayitresult
may result in a reduction
in a reduction in theinsearch
the search efficiency
efficiency of the
of the evolutionary
evolutionary optimization
optimization process
process [42].
[42]. In
In addition,
addition, thethe resultofofmulti-objective
result multi-objectiveoptimization
optimizationblends
blendsfeatures
featuresrelated
related toto conflicting
conflicting objectives,
objectives,
which
which could
could make
make itit difficult
difficult to
to identify
identify the
the dominant
dominant feature
feature related
related to
to each
each objective
objective asas clearly
clearly as
as
shown
shown inin the
the case
casestudies
studiesof ofthis
thisresearch.
research.
Last,
Last,apart
apartfrom
fromthetheissues andand
issues limitations mentioned
limitations mentionedabove, there there
above, are two areother
twodirections related
other directions
to the algorithm itself that can be considered in future research. First, when considering
related to the algorithm itself that can be considered in future research. First, when considering the the energy
consumption of air conditioning, the partition on the floor plan of rooms is critical to producing accurate
simulation results. While there is a technique to conduct simple thermal zoning by a straight-skeleton
subdivision [43], grammar approaches [44] allow for a more realistic approximation of the subdivision
of rooms. These techniques can be integrated into the algorithm. Second, increasing the degree of
customization can allow the algorithm to satisfy more specific design settings. For example:
• More complex shapes for the maximal volume beyond box-like masses to make possible
combinations, such as towers and podiums and non-orthogonal geometries.
• Provide more practical user-defined constraints to regulate the overall building massing, such as
maximum height, width, and length of the building massing.
• Provide additional placement strategies of building cores, such as placing the cores at corners
rather than in the center of the building.
5. Conclusions
The focus of this research was to develop a reusable algorithm that generates building massing
with extensive topological variability to help architects conduct explorative and non-disruptive
performance-based optimization. The underlying intention of the research was to use optimization
as a means of exploiting the potential in building massing design for passive energy savings at the
outset of design processes. This process can narrow down the scope of consideration for architects,
thereby facilitating knowledge extraction for performance-aware design processes. At the same time,
what is indispensable is that such optimization-based exploration should be well integrated into the
architectural design process.
For the algorithm focusing on design exploration rather than providing direct solutions, we
acknowledge that the building massing found by the case-study optimization may not satisfy all
aspects of architectural design [16,45]. Instead, in the data-poor situation of conceptual design stages,
the solution retrieved from the optimization plays a role as the “medium of reflection” in design
processes and allows architects to draw inspiration from its result [13]. In this regard, the use of the
subtractive form generation principle has the advantage that the architectural implication revealed by
the optimization result is intuitive and cognizable, which facilitates architects’ gaining insight from the
optimization results.
Driven by the optimization process, subtractors are manipulated only to remove volume from the
building massing that contributes to the most dominant and profitable effect on the improvement in
the building energy performance. It is evident that compared with the random sampling alternatives
in Figures 17 and 18, irrelevant and unnecessary voids appearing in those sampling alternatives
were mostly discarded and excluded across the optimization process. With minimal redundant
information, architects’ focus can remain on the dominant building massing characters and the
associative architectural implications related to building energy performance. In contrast, ambiguity and
distractions may happen if the geometry of the building massing is too complex. Desired architectural
implications could be misled by other less relevant issues.
In order to maximize the utility of performance-based optimization to assist architects in achieving
compelling energy sustainability, the proposed algorithm provides an unconventional approach to
using such optimization in conceptual design processes. Without parametric modelling, using the
algorithm encourages architects to leverage optimization processes to explore unknown solutions
and collect feedback related to various building energy performance factors at the outset of the
design process. In contrast, conventional performance-based optimization is only carried out after
the stage of design ideation and concept development. However, when the design concept is already
predetermined in building massing design, the room for performance improvements is limited. More
importantly, for those concepts developed in the data-poor situations of early design stages, these
concepts could be problematic and may result in flaws, which “can rarely be compensated at later
design stages and incurs a great redesign expense” [46]. In contrast, the proposed algorithm allows for
early interventions of performance-based optimization before and during concept development stages.
Such interventions increase architects’ awareness of building energy performance, thereby reducing
the possibility of a poor decision being made during the building massing design stages. The increased
awareness also allows architects to extrapolate the trajectory beyond the possibilities explored by the
optimization process, thereby driving the design towards a desired environmental-friendly solution.
Author Contributions: Conceptualization, L.W., P.J. and K.W.C.; methodology, L.W.; software, L.W.; resources,
G.J.; writing—original draft preparation, L.W.; writing—review and editing, K.W.C., P.J. and Z.T.; supervision,
G.J.; funding acquisition, Z.T. and G.J.
Funding: The work described in this study was sponsored by the projects of the National Natural Science
Foundation of China (NSFC#51578277) and (NSFC#51378248).
Acknowledgments: The authors would like to thank anonymous reviewers, editors, Zhonggao Chen, and
Hyoung-June Park for their detailed and valued comments and suggestions. A warm acknowledgment is given to
Derek Pung Shuai Shi for his kind help to this research.
Conflicts of Interest: The authors declare no conflict of interest.
References
1. Shi, X. Performance-Based and Performance-Driven Architectural Design and Optimization. Front. Archit.
Civ. Eng. China 2010, 4, 512–518. [CrossRef]
2. Kolarevic, B.B.; Malkawi, A. Performative Architecture: Beyond Instrumentality; Spon Press: New York, NY,
USA, 2005.
3. DeKay, M.; Brown, G.Z. Sun, Wind, and Light: Architectural Design Strategies, 2nd ed.; John Wiley & Sons:
Hoboken, NJ, USA, 2014.
4. Simitch, A.; Warke, V. The Language of Architecture: 26 Principles Every Architect Should Know; Rockport Pub:
Beverly, MA, USA, 2014.
5. McMullan, R. Environmental Science in Building, 4th ed.; Palgrave Macmillan Education: London, UK, 1998.
6. Nault, E.; Waibel, C.; Carmeliet, J.; Andersen, M. Development and Test Application of the UrbanSOLve
Decision-Support Prototype for Early-Stage Neighborhood Design. Build. Environ. 2018, 137, 58–72.
[CrossRef]
7. Yi, Y.K.; Malkawi, A.M. Site-Specific Optimal Energy Form Generation Based on Hierarchical Geometry
Relation. Autom. Constr. 2012, 26, 77–91. [CrossRef]
8. Turrin, M.; Von Buelow, P.; Stouffs, R. Design Explorations of Performance Driven Geometry in Architectural
Design Using Parametric Modeling and Genetic Algorithms. Adv. Eng. Inform. 2011, 25, 656–675. [CrossRef]
9. Woodbury, R.F.; Burrow, A.L. Whither Design Space? AIE EDAM Artif. Intell. Eng. Des. Anal. Manuf. 2006,
20, 63–82. [CrossRef]
10. Dino, I.G. Creative Design Exploration by Parametric Generative Systems in Architecture. Metu J. Fac. Arch.
2012, 29, 207–224. [CrossRef]
11. Touloupaki, E.; Theodosiou, T. Performance Simulation Integrated in Parametric 3D Modeling as a Method
for Early Stage Design Optimization—A Review. Energies 2017, 10, 637. [CrossRef]
12. Shi, X.; Tian, Z.; Chen, W.; Si, B.; Jin, X. A Review on Building Energy Efficient Design Optimization from the
Perspective of Architects. Renew. Sustain. Energy Rev. 2016, 65, 872–884. [CrossRef]
13. Wortmann, T. Efficient, Visual, and Interactive Architectural Design Optimization with Model-Based Methods.
PhD Thesis, Singagore University of Technology and Design, Singapore, 2018. [CrossRef]
14. Janssen, P. Dexen: A Scalable and Extensible Platform for Experimenting with Population-Based Design
Exploration Algorithms. AI EDAM 2015, 29, 443–455. [CrossRef]
15. Kyropoulou, M.; Ferrer, P.; Subramaniam, S. Optimization of Intensive Daylight Simulations: A Cloud-Based
Methodology Using HPC (High Performance Computing). In Proceedings of the PLEA 2018 HONG KONG
Smart and Healthy within the 2-Degree Limit, Hong Kong, China, 10–12 December 2018; pp. 150–155.
16. Wortmann, T.; Nannicini, G. Introduction to Architectural Design Optimization. In City Networks: Collaboration
and Planning for Health and Sustainability; Karakitsiou, A., Migdalas, A., Rassia, S.T., Pardalos, P.M., Eds.;
Springer International Publishing: Cham, Switzerland, 2017; pp. 259–278. [CrossRef]
17. Burry, M.; Murray, Z. Architectural Design Based on Parametric Variation and Associative Geometry.
In Proceedings of the 15th eCAADe Conference Proceedings, Vienna, Austria, 17–20 September 1997;
ISBN 0-9523687-3-0.
18. Oxman, R. Thinking Difference: Theories and Models of Parametric Design Thinking. Des. Stud. 2017,
52, 4–39. [CrossRef]
19. Harding, J.E.; Shepherd, P. Meta-Parametric Design. Des. Stud. 2017, 52, 73–95. [CrossRef]
20. Caldas, L.; Santos, L. Generation of Energy-Efficient Patio Houses With GENE_ARCH. eCAADe 30 2012,
1, 459–470.
21. Chen, K.W.; Janssen, P.; Schlueter, A. Multi-Objective Optimization of Building Form, Envelope and Cooling
System for Improved Building Energy Performance. Autom. Constr. 2018, 94, 449–457. [CrossRef]
22. Chen, K.W.; Choo, T.S.; Norford, L. Enabling Algorithm-Assisted Architectural Design Exploration for
Computational Design Novices. Comput. Des. Appl. 2019, 16, 269–288. [CrossRef]
23. Wang, L.; Janssen, P.; Ji, G. Progressive Modelling for Parametric Design Optimization: An Example of How
Parametric Design Optimization Can Support Reflection. In Proceedings of the 24th CAADRIA Conference,
Wellington, New Zealand, 15–18 April 2019; pp. 383–392.
24. Chen, K.W. Architectural Design Exploration of Low-Exergy (LowEx) Buildings in the Tropics. Ph.D. Thesis,
ETH Zurich, Switzerland, 2015.
25. Xu, X.; Yin, C.; Wang, W.; Xu, N.; Hong, T.; Li, Q. Revealing Urban Morphology and Outdoor Comfort
through Genetic Algorithm-Driven Urban Block Design in Dry and Hot Regions of China. Sustainability
2019, 11, 3683. [CrossRef]
26. Si, F.; Wang, T. Building Massing Optimization in Early de Sign Stage. In Proceedings of the 20th International
Conference of the Association for Computer-Aided Architectural Design Research in Asia, Daegu, Korea,
20–22 May 2015; pp. 583–592.
27. Janssen, P. A Generative Evolutionary Design Method. Digit. Creat. 2006, 17, 49–63. [CrossRef]
28. Wang, L.; Janssen, P.; Ji, G. Utility of Evolutionary Design in Architectural Form Finding: An Investigation
into Constraint Handling Strategies. In International Conference on-Design Computing and Cognition; Springer
International Publishing: Cham, Switzerland, 2019; pp. 177–194. [CrossRef]
29. Stouffs, R.; Janssen, P.; Roudavski, S. An Evolutionary Design Process: Adaptive-Iterative Explorations
in Computational Embryogenesis Vignesh. In Proceedings of the 18th International Conference on
Computer-Aided Architectural Design Research in Asia (CAADRIA 2013), Singapore, 15–18 May 2013;
pp. 137–146.
30. Yi, H. User-Driven Automation for Optimal Thermal-Zone Layout during Space Programming Phases. Arch.
Sci. Rev. 2016, 59, 279–306. [CrossRef]
31. Yi, H.; Yi, Y.K. Performance Based Architectural Design Optimization: Automated 3D Space Layout Using
Simulated Annealing. In Proceedings of the 2014 ASHRAE/IBPSA-USA Building Simulation Conference,
Atlanta, GA, USA, 10–12 September 2014; pp. 292–299.
32. Cichocka, J.; Browne, W.N.; Ramirez, E.R. Optimization in the Architectural Practice An International Survey.
In Proceedings of the 22nd CAADRIA Conference, Suzhou, China, 5–8 April 2017; pp. 387–397.
33. Janssen, P. A Design Method and Computational Architecture for Generating and Evolving Building Designs.
PhD Thesis, The Hong Kong Polytechnic University, Hong Kong, China, 2005.
34. Woodbury, R. Elements of Parametric Design; Routledge: Abingdon, UK, 2010.
35. Akin, O.; Moustapha, H. Strategic Use of Representation in Architectural Massing. Des. Stud. 2004, 25, 31–50.
[CrossRef]
36. Wortmann, T. Architectural Design Optimization-Results from a User Survey. In Proceedings of the
Architecture Across Boundaries 2019, Suzhou, China, 19–21 June 2019; CKnE Social Sciences: Dubai, UAE.
[CrossRef]
37. Wang, L.; Janssen, P.; Ji, G. Diversity and Efficiency: A Hybrid Evolutionary Algorithm Combining the
Island Model with a Steady-State Replacement Strategy. In Proceedings of the 24th CAADRIA Conference,
Wellington, New Zealand, 15–18 April 2019; pp. 593–602.
38. Sun, Y.; Turrin, M.; Sariyildiz, S.; Ren, S.; Yang, D. Multi-Disciplinary and Multi-Objective Optimization
Problem Re-Formulation in Computational Design Exploration: A Case of Conceptual Sports Building
Design. Autom. Constr. 2018, 92, 242–269. [CrossRef]
39. Jakubiec, J.A.; Reinhart, C.F. DIVA 2.0: Integrating Daylight and Thermal Simulations Using Rhinoceros 3D,
Daysim and EnergyPlus. Proc. Build. Simul. 2011, 20, 2202–2209.
40. Raji, B.; Tenpierik, M.J.; van den Dobbelsteen, A. Early-Stage Design Considerations for the Energy-Efficiency
of High-Rise Office Buildings. Sustainability 2017, 9, 623. [CrossRef]
41. Yi, Y.K.; Malkawi, A.M. Optimizing Building Form for Energy Performance Based on Hierarchical Geometry
Relation. Autom. Constr. 2009, 18, 825–833. [CrossRef]
42. Cao, K.; Huang, B.; Wang, S.; Lin, H. Sustainable Land Use Optimization Using Boundary-Based Fast Genetic
Algorithm. Comput. Environ. Urban Syst. 2012, 36, 257–269. [CrossRef]
43. Dogan, T.; Reinhart, C.; Michalatos, P. Autozoner: An Algorithm for Automatic Thermal Zoning of Buildings
with Unknown Interior Space Definitions. J. Build. Perform. Simul. 2016, 9, 176–189. [CrossRef]
44. Wang, X.Y.; Yang, Y.; Zhang, K. Customization and Generation of Floor Plans Based on Graph Transformations.
Autom. Constr. 2018, 94, 405–416. [CrossRef]
45. Bradner, E.; Iorio, F.; Davis, M. Parameters Tell the Design Story: Ideation and Abstraction in Design
Optimization. Simul. Ser. 2014, 46, 172–197.
46. Okudan, G.E.; Tauhid, S. Concept Selection Methods—A Literature Review from 1980 to 2008. Int. J. Des.
Eng. 2008, 1, 243. [CrossRef]
© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

Sustainability 11 06965 v2 PDF

Uploaded by

Copyright:

Available Formats

Sustainability 11 06965 v2 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Sustainability 11 06965 v2 PDF

Uploaded by

Copyright:

Available Formats

sustainability

Abstract: For sustainable building design, performance-based optimization incorporating parametric

Keywords: parametric massing algorithm; building massing design; performance-based optimization;

Sustainability 2019, 11, 6965; doi:10.3390/su11246965 www.mdpi.com/journal/sustainability

Parametric Modelling for Building Massing Design

2.1. Overview of Proposed Algorithm

2.2. Generative Steps

Sustainability 2019, 11, 6965 5 of 20

2.2. Generative Steps

Figure 2.2. The

Figure 7. The size constraint in the vertical direction.

Figure 13. Generated

Sustainability 2019, 11, 6965 10 of 20

2.4. Case Studies

2.4. Case Studies

2.4.1. Design Setting

Figure 15. Plan

2.4.2. Initialization Parameters

Table 1. Initialization parameters.

Figure 19. The

3.1. Case Study 1

Figure 20. Elite individuals of natural daylighting (south-west aerial view).

Second, cases, such

Figure 22. Elite individuals of daylighting (south-east aerial view).

Figure 24. Parallel

You might also like