ctbuh.org/papers
Title:
A Design Optimization Workflow for Tall Buildings Using Parametric
Algorithm
Authors:
Mahjoub Elnimeiri, Professor, Illinois Institute of Technology
Mahsa Nicknam, PhD Candidate, Illinois Institute of Technology
Subjects:
Architectural/Design
IT/Computer Science/Software
Keywords:
Energy Efficiency
Environment
Optimization
Parametric Design
Performance Based Design
Structure
Publication Date:
2011
Original Publication:
CTBUH 2011 Seoul Conference
Paper Type:
1.
2.
3.
4.
5.
6.
Book chapter/Part chapter
Journal paper
Conference proceeding
Unpublished conference paper
Magazine article
Unpublished
© Council on Tall Buildings and Urban Habitat / Mahjoub Elnimeiri; Mahsa Nicknam
CTBUH 2011 World Conference
October 10-12, 2011, COEX, Seoul, KOREA
_______________________________________________________________________________________________________________________________________
MS01-01
A design optimization workflow for tall buildings using parametric
Algorithm
Mahsa Nicknam1, Mahjoub M.Elnimeiri2
Ph.D. Candidate, College of Architecture, Illinois Institute of Technology, Chicago, USA,
Email address: [email protected] 1
Professor, College of Architecture, Illinois Institute of Technology, Chicago, USA,
Email address: [email protected] 2
Presenter Photo
Mahjoub M. Elnimeiri
Biography
Dr. Mahjoub M. Elnimeiri holds a B. Sc. In Civil Engineering (with Honors) from the
University of Khartoum; a D.I.C. and M. Sc. in Structural Engineering from the
Imperial College, University of London, London, England and a Ph.D. in Structural
Engineering and Structural Mechanics from Northwestern University, Evanston,
Illinois, USA.
Since his educational completion in 1974, he has been involved in the practice of
structural and architectural work of buildings in Chicago, USA and overseas. His expertise is in the areas of
structural analysis, design and construction of buildings, and in particular tall buildings, and in the application
of computer technology. He is very active and well known in the professional and academic communities. He
has been speaker in many conferences and conventions. His publications are directly related to his areas of
study, and he is a recipient of the “State of The Art Award” of the American Society of Civil Engineers for the
year 1989.
Dr Elnimeiri is Chairman of CECI+ Engineers International, Chicago, Illinois, USA. CECI+ is a progressive
engineering firm using cutting edge of technology for service provision to its clients. The practice includes
structural analysis, design and overseeing construction of very special buildings, primarily tall buildings, at
different places in US and overseas. Few of these buildings are located in highly seismic areas. Before
beginning his own practice, Dr Elnimeiri worked with Skidmore Owings & Merrill, Chicago from 1979-1990.
During his experience at SOM, he worked closely with Architectural Partners in the design development of
many outstanding projects. Dr Elnimeiri left SOM as an Associate Partner and Senior Structural Engineer.
Dr Elnimeiri is also a full professor, Founder & Director of the PhD Program, at the College of Architecture at
Illinois Institute of Technology, Chicago, Illinois, USA.
______________________________________________________________________________________________________________________________________
561
CTBUH 2011 World Conference
October 10-12, 2011, COEX, Seoul, KOREA
_______________________________________________________________________________________________________________________________________
Abstract
arch paper is intended to incorporate structural pe
erformance and
a energy efficiency stra
ategies of
This resea
tall building
gs through an
a architecturral form gene
eration proce
ess. The maiin objective is to propose
e a basis
for a generative design
n methodolog
gy in order to
o develop arc
chitectural fo
orms of tall buildings that are
mance and energy
e
conse
ervation in arrchitecture.
optimized both in termss of the strucctural perform
e this goal, th
he research paper
p
proposses a digitally-based worrkflow to aid direct interac
ction
To achieve
among ene
ergy efficienccy, structurall performancce and forma
al design considerations w
within the sch
hematic
design pha
ase. By employing the pa
arametric mo
odeling platfo
orms, the stru
uctural and e
energy efficie
ency
implication
ns of changin
ng architectural forms of tall
t buildings would be un
nderstood byy the architec
ct. Some
architects may use the
e outputs as general
g
conccept for their designs and
d others mayy choose to use
u them
and how a se
ensitive tall building
b
form
m responds to
o certain perfformance me
etrics. The co
ore of this
to understa
research will
w be to define design drriven criteria to focus on the
t optimization process.
Keywords
s:
Multi objecctive Optimizzation, Enviro
onmental dessign, Genera
ative system,, Performancce-based des
sign,
Parametricc design, Ene
ergy efficienccy, Tall buildiings, Genetic
c algorithms,, Structural P
Performance
Introduction
The depletin
ng energy re
esources on earth and th
he rising costt of energy are
a the most real problem
ms in today’ss
world. In su
uch a conte
ext, energy efficient
e
desiign is becom
ming importa
ant as we th
hink more about energyy
conservatio
on and the environment. Meanwhile, structure ha
as to assume
e its place a
and directly contribute
c
to
o
the building
g’s overall en
nergy conserrvation strate
egy (Elnimeirri and Almusharaf, 2010)). The interes
st in efficientt
structure within a tall bu
uilding highlig
ghted by the fact that the
e structural co
osts constitu
ute up to 30%
% of the totall
n cost of the building, (Fig
gure 1) (Elnimeiri and Alm
musharaf, 20
010).
construction
5
30
40
A
Architecture
M
MEP
S
Structure
25
E
Elevator
Sys..
Figure 1. Sttructural costt in relation to
o the total bu
uilding construction cost (source: Elnimeiri and Allmusharaf,
2010)
______________________________________________________________________________________________________________________________________
562
CTBUH 2011 World Conference
October 10-12, 2011, COEX, Seoul, KOREA
_______________________________________________________________________________________________________________________________________
Today architectural dessign is drive
en by the se
earch for ex
xotic forms that more th
hat often lea
ad to costlyy
n. In the current
c
tall building dessign practice
e, issues pertaining
p
to
o structure and energyy
construction
conservatio
on are typicallly left to be dealt
d
with aftter the archittectural form is well articu
ulated. Hence as a resultt,
such an approach may enable a bu
uilding to sta
and upright, and
a may also reduce the
e energy con
nsumption in
n
g. It will not yield
y
solution
ns that will pe
erform well with
w referencce to structurre and energ
gy efficiency,,
the building
(Figure 2).
Figure 2. Tall building design
d
processs
The
e process sh
hows the resu
ults are not necessarily
n
efficient
e
neith
her structurallly nor energy
y wise
s
in
n the traditio
onal enginee
ering analysis
s and design
n process, th
he geometry
y is imported
d
Moreover, sometimes
manually in
nto Analysis Tools and the
t
pre-processing is pe
erformed by an engineer via the gra
aphical userr
interface. This
T
approach takes a lo
ot of time wh
hen multiple options are
e proposed b
by the archittect. In such
h
approach, the engineer will often fin
nish a preliminary analysis and begin to communicate the new
w scheme to
o
he design has been radic
cally changed
d, (Chok and
d Donofrio, 2010)(Figure
2
e
the architecct, only to find out that th
3).
Architecttural
design
n
maybe
e
subjected
d to
significa
ant
change
es
Figure
e 3. Tradition
nal Architectu
ural design workflow
w
whic
ch maybe yie
elds a differe
ent architectu
ural form
t modern era
e of archite
ectural desig
gn of tall build
dings, in Chiccago, betwee
en 1960 and
d
It is evident that during the
______________________________________________________________________________________________________________________________________
563
CTBUH 2011 World Conference
October 10-12, 2011, COEX, Seoul, KOREA
_______________________________________________________________________________________________________________________________________
1975, there was a meaningful collaboration between the architect and the structural engineer. A good
example for that is John Hancock Building in Chicago, 1996, (Figure 4). However energy efficiency strategies
were never considered in this building.
Figure 4. John Hancock Center, Chicago, USA. (Courtesy of Skidmore
Owings and Merrill)
As mentioned above, the development of tall buildings had been traditionally a product of a close
collaboration between the architect and the engineer. Since in the design process, the most important design
decisions, those which have the most significant cost impacts, are made at the concept stage of a project,
the collaboration between different disciplines within the design team in this stage become very important.
Barriers to this collaboration often have resulted in that small and large architectural firms either left the
sophisticated modeling of environmental metrics until too late in the design process, or they neglect it
altogether (Galasiu and Reinhart, 2007), so as a result they would come up with inefficient architectural form.
To address such disconnect, this research paper propose a digitally-based workflow to aid direct interaction
among energy efficiency, structural performance and formal design considerations to assist early design
activities.
Parametric Design
Within the last years digital tools have extensively progressed so as to allow the architects to increasingly
become involved in an integrative collaborative architectural building design approaches. In these years,
digital media is not used as a representational tool for visualization but as a generative tool for the derivation
of form and its transformation.
Figure 5. Tall Building Generation Process by Parametric Design (Courtesy of Park, 2005)
______________________________________________________________________________________________________________________________________
564
CTBUH 2011 World Conference
October 10-12, 2011, COEX, Seoul, KOREA
_______________________________________________________________________________________________________________________________________
In the last few decades computer simulations have proven to be a powerful digital tool for studying the
environmental performance of buildings. By using simulation and analysis tools, it is possible to engage in an
architectural design practice that is significantly enhanced by iterative process that depends on feedback
from the evaluation of the environmental impact into the architectural design decision making. (shown in
Figure 6).
Among various digital design approaches, parametric design has become a strong trend in contemporary
architectural design practice and education (Day, 2010). The term refers to a practice of digitally modeling a
series of design variants whose relationships to each other are defined through one or several mathematical
relationship (parameters), then form a parametric space that may comprise of tens or hundreds of related but
distinct forms. By relying on parametric values, it becomes possible to associate variables, relations, and
dependencies to geometry and structure in the design process (Dominik, Jiwu, Mike & Mark, 2004). This
facilitates rapid editing, re-adaptability, and enables useful libraries of scripts and reusable parts for future
use. Hence these parametric modeling systems help the architect to rapidly generate and test a large
number of design options and variation in the early stage of design by linking the generative tools to
structural, energy efficiency analysis and evaluation tools. This approach minimizes performance-based
study time and aids the project team in making better informed decisions earlier in the design process.
Extending the potential of parametric/associative models to respond to both structural and energy efficiency
inputs and performance criteria is the main focus of this research.
Figure 6. Parametric modeling processes and solar
incidence angle analysis for the Lotte Tower project
in Seol, Korea (Courtesy of Skidmore, Owings &
Proposed Workflow
In this work, we emphasize mainly on a general trend to architectural form design approaches defining
mostly the basics of performance strategies of tall buildings. The advantage of parametric design in research
application is to plan and synthesize the overall requirements and relationships of many design elements into
one form. The convergence of generative tools, analytical solutions and moving away from traditional formdriven approaches are the main achievements. Various parametric modeling platforms have been employed
to simply evaluate how changes in architectural forms of a tall building would affect its structural performance
as well as its energy efficiency.
______________________________________________________________________________________________________________________________________
565
CTBUH 2011 World Conference
October 10-12, 2011, COEX, Seoul, KOREA
_______________________________________________________________________________________________________________________________________
Figure 7. Conceptual Design approach
This research paper demonstrates how a fixable 3D model can be parametrically altered towards targeted
solutions with the help of a near real time feedback generated by structural/energy analysis and optimization.
The main focus in this work is defining the architectural model in a way to be parametrically reconfigurable
while at the same time remains dynamically responsive to performance-based engineering input. The
architectural model will be kept in a fluid state allowing for intuitive variations according to the feedback from
performance-based data. All different solutions can be analyzed and evaluated according to ones objective
and preference.
Energy performance analysis, in parallel with structural analysis, is performed on the parametric models
through an optimization framework leading to a considerably faster performance-based study. It is shown
that, the proposed workflow gives the architect the opportunity to select among different design options
based on ones objectives. Using this workflow to simulate environmental and structural performance
provides a visualization of results with color mapping, enabling performance to be easily understood by all
members of the design team.
Figure 8. Proposed Workflow
______________________________________________________________________________________________________________________________________
566
CTBUH 2011 World Conference
October 10-12, 2011, COEX, Seoul, KOREA
_______________________________________________________________________________________________________________________________________
Multi-obje
ective optim
mization
This research paper will focus on Multi-objective
M
e optimizatio
on techniquess in lie single
e optimizatio
on approach..
ptimization allows direct communicatio
c
on between multi-disciplinary simulattion software
e,
Since Multi--objective op
it is becomin
ng a key facttor in the dessign processs.
The propossed workflow
w, (Figure 8),, allows integrating the Structural
S
an
nd Energy S
Simulation so
oftwares into
o
one design environmentt. The proposed automatted process, as shown in
n Figure 9, will allow the achievement
a
t
orm satisfying
g the specifie
ed optimization objectives.
of a final arcchitectural fo
Figure 9. Conceptual Framework
k for the Prop
posed Processs
The optimizzation part of
o the propossed workflow
w will encoun
nter more tha
an one objective includin
ng structurall
efficiency and Energy efficiency
e
objjectives. So in some cas
ses we might have conflicting objectives. One off
a
o using Gen
of
netic Algorith
hm (GAs) is that they alllow true mullti-objective optimization.
o
.
the major advantages
Multi-objecttive optimization techniqu
ues keep the
e objectives separate du
uring the opttimization pro
ocess ratherr
than being collapsed into
i
a single
e, weighted objective function from
m the beginning. So in cases with
h
o
the
ere will be no
o single optim
mum result1.
conflicting objectives
Figure 10. Tall building form ge
eneration using Genetic A
Algorithm
Genetic Alg
gorithm manipulates in pa
arallel a pop
pulation of so
olutions, so in
n the processs of the optimization the
e
1
http://www
w. esteco.com
m
______________________________________________________________________________________________________________________________________
567
CTBUH 2011 World Conference
October 10-12, 2011, COEX, Seoul, KOREA
_______________________________________________________________________________________________________________________________________
optimal trad
de-offs amon
ng multiple crriteria can be
e found. In th
his process the
t output is not a single solution butt
a Pareto Frrontier (or tra
ade-off curve
e) of high-perrformance so
olutions, which will allow
w the architec
ct to pick the
e
point at which he would
d like to ope
erate or he can
c
choose to consider a new criterria (Caldas and
a
Norford,,
2001) (Figure 11).
Figure 11. Example
E
of a Pareto frontier. The boxxed points re
epresent feassible choices, and smaller values are
preferred to
t larger one
es. Point C iss not on the Pareto
P
Frontier because it is dominated by both point
p
A and
point B. Points A and B are not stricctly dominate
ed by any oth
her, and hencce do lie on tthe frontier.(Courtesy of
Wikipedia)
ons
Conclusio
This researrch paper introduced a workflow
w
that encompass
ses a loop off four main p
phases of the
e procedure,,
namely form
m generation
n, structural performance
p
evaluation, energy
e
perfo
ormance eva
aluation, and optimization
n
that are to be
b conducted
d under pred
defined criterria. Parametrically archite
ectural form generation proved
p
to be
e
flexible eno
ough that alllows architect to manipulate certain
n architecturral design in
ntentions to improve the
e
performance of the tall building witth respect to
o structure and
a
energy efficiency strategies. The
T
workflow
w
e architect to
o incorporatte structure and energyy
developed in this research paper is intended to help the
onsiderationss into the arcchitectural fo
orm generatio
on process. This
T
will help
p the architec
ct to achieve
e
efficiency co
a well-round
ded performa
ance-based architectural design of ta
all buildings.
In such casse of Multi-o
objective op
ptimization, structure
s
and
d energy co
onsiderationss might have
e conflicting
g
objectives; hence exterm
mination of one
o goal can
n only be ach
hieved at the
e expense off another. In this processs
o
se
eparate. The
e optimizatio
on process will
w then lead
d to a Paretto frontier off
it is better to keep all objectives
T
Ge
enetic Algoriithm was cho
osen as an optimization
o
technique in
n the propos
sed workflow
w
solutions. Therefore,
that manipu
ulates in para
allel a popula
ation of high--performance
e solutions.
Reference
es
[1] http://ww
ww. esteco.ccom
[2] Bazjana
ac V. (2001) Acquisition of Building Geometry
G
in the Simulatio
on of Energyy, Presented
d at the 2001
Building
g Simulation Conference in Rio De Ja
aneiro, Brazil.
[3] Besseru
ud K., Archite
ectural Geno
omics, Skidm
more, Owings
s, and Merrilll LLP (BlackB
Box Studio).
[4] Bourgeo
ois D., Reinh
hart C. F. an
nd Ward G. (2008) A Standard
S
Dayylight Coefficcient Model for Dynamicc
Daylighting Simulatiions , Buildin
ng Research & Informatio
on 36 (1) pp.6
68 - 82.
______________________________________________________________________________________________________________________________________
568
CTBUH 2011 World Conference
October 10-12, 2011, COEX, Seoul, KOREA
_______________________________________________________________________________________________________________________________________
[5] Caldas L. G. and Norford L. K. (2001) A Design Optimization Tool Based on a Genetic Algorithm,
Automation Construction, pp. 173 – 184.
[6] Caldas L. (2008) Generation of Energy-Efficient Architecture Solutions Applying GENE_ARCH: An
Evolution-Based Generative Design System”, Advanced Engineering Informatics 22, pp.59– 70.
[7] Caldas L. G. (2001) An Evolution-Based Generative Design System: Improving the Environmental
Performance of Buildings. Ph.D. Thesis, Building Technology Program, Dept. Of Architecture, MIT,
Cambridge, MA.
[8] Caldas L. G. and Norford L. K. (2003) Genetic Algorithms for Optimization of Building Envelopes and the
Design and Control of HVAC Systems, Journal of Solar Energy Engineering, Vol. 125, pp. 343-351.
[9] Caldas L., and Norford L. K. (2001) Architectural Constraints in a Generative Design System:
Interpreting Energy Consumption Levels, Proc. of Building Simulation, Int., Building Performance
Simulation Association.
[10]
Caldas L., Norford L., and Rocha, J. (2001) An Evolutionary Model for Sustainable Design, Proc.
Passive and Low Energy Architecture (PLEA).
[11]
Chok K. and Donofrio M. (2003) Abstractions for Information Based Design, International Journal of
Architectural Computing, Vol. 08, pp. 223-255.
[12]
Chok K. and Donofrio M. (2010) Structure at the Velocity of Architecture, ACADIA.
[13]
Chun Huang Y., Lam K.P., and Dobbs G. M. (2008) A Scalable Lighting Simulation Tool for
Integrated Building Design, Third National Conference of IBPSA-USA, Berkeley, California.
[14]
Elnimeiri M. and Almusharaf A. Structure and Architectural Form of Tall Buildings. Proceeding of the
International Conference on Sustainable Building Asia, Seoul, Korea, pp. 54-60.
[15]
Galasiu A. and Reinhart C. F. (2007) Current Daylighting Design Practice: A Survey. Building
Research and Information, 36(2), pp.159-174.
[16]
Gehry Technologies (2004) Digital ProjectTM, 12541-A Beatrice Street, Los Angeles, California,
USA, http://www.gehrytechnologies.com.
[17]
Krawczyk R. J., Park S. M., and Sharpe D.(2005), Innovative Tall Building Form Development ,
Council of Tall Buildings and Urban Habitat, CTBUH Conference, New York, NY.
[18]
Lagios K., Niemasz J. and Reinhart C. F. (2010) Animated BuildingPerformance Simulation (ABPS)
– Linking Rhinoceros/Grasshopper with Radiance/Daysim, To be published in the conference
proceedings of SimBuild 2010 in August 2010 in New York City.
[19]
Lam K. P., Chun Huang Y. and Dobbs G. M., Development of a Lighting Simulation Tool for
Integrated Building Design, Center for Building Performance and Diagnostics, Carnegie Mellon
University, Pittsburgh, Pennsylvania, USA.
[20]
Lam K. P. and Yeang K. (2009) Computational Building Performance Modeling and Ecodesign,
Architectural Design, Volume 79, Issue 5, pp. 126–129, 2009.
[21]
Pan Y., Huang Z, Wu Z. ( 2007)Calibrated Building Energy Simulation and Its Application in a HighRise Commercial Building in Shanghai, Energy and Building 39, pp. 651-657.
[22]
Park S. M., Elnimeiri M., Sharpe D.C., and Krawczyk R. J. (2004), “Tall Building Form Generation
by Parametric Design Process”, CTBUH, Seoul, Korea, pp, 56- 62.
[23]
[24]
Reinhartand C.F. and Wienold J. (August 2010 ) The Delighting Dashboard - A Simulation-Based
Design Analysis for Daylit Spaces, Accepted for publication in Building and Environment.
[25]
Schinable M. A. (2007) Parametric Designing in Architecture, University of Sydney, Australia, A
Dong, A Vande Moere & JS Gero (eds), CAAD Futures, pp.237-250.
[26]
Schlueter A. and Thesseling F. (2009) Building Information Model Based Energy/Exergy
Performance Assessment in Early Design Stages”, Automation in Construction 18, 2009, pp. 153– 163.
[27]
Yi Y. K. and Malkawi A. M. (2009) Optimizing Building Form for Energy Performance Based on
Hierarchical Geometry Relation”, Automation in Construction 18, 2009, pp. 825–833.
______________________________________________________________________________________________________________________________________
569