Professional Documents
Culture Documents
JAN LUGOWSKI
EGI-2013-046MSC
Master of Science Thesis EGI-2013-046MSC
Jan Lugowski
Abstract
The purpose of this paper is to explore the ferrocement building technique for sustain-
able housing. Ferrocement involves the use of conventional cement with fine aggregate and
several layers of steel, with the advantage of higher strength than conventional reinforced
concrete, limited formwork and thinner sections. It is particularly suitable for thin shell
structures, where geometry minimizes bending loads. Architectural flexibility is one of the
main priorities considered in sustainable housing, along with energy efficiency, occupant
comfort, resistance to seismic and tornado events, affordability and durability. Ferroce-
ment’s historical and present applications are covered, along with other building techniques,
in order to establish best practices and possible improvements. Reducing construction labor
is a particular focus, which has limited ferrocement development in recent years. Computer
modeling of shell form finding is described, with three case studies created. A structural
analysis method is described and applied to each case study to verify general building code
safety. Energy modeling is performed in two climates for each case study in the United
States and compared to key PassivHaus energy demand limits. Net zero energy use is pos-
sible with on-site solar photovoltaic generation.
i
Acknowledgements
Thank you to my examiner, Jaime Arias, and advisor Peter Kjareboe for their support during
the project and allowing me full flexibility in exploring a topic which I deeply care about. The
project allowed me to combine my previous working experience with material learned during
the SEE program. I’m grateful for the holistic approach that my mentors embraced.
Jan Lugowski
Stockholm 2013
ii
Contents
1 INTRODUCTION 1
1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2 FERROCEMENT CONSTRUCTION 6
2.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2 Competing building systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.3 Opportunities for improving ferrocement . . . . . . . . . . . . . . . . . . . . . . . 21
2.4 Proposed building techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4 STRUCTURAL ANALYSIS 37
4.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.2 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.3 Material properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.4 Loads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.5 Evaluation strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
4.6 Case study 1: Shell Retrofit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
4.7 Case study 2: Calabaza . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.8 Case study 3: Ballena . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
6 CONCLUSIONS 56
BIBLIOGRAPHY 60
iii
A EXAMPLES OF FERROCEMENT CONSTRUCTION 61
C SOFTWARE 68
C.1 Modeling Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
C.2 FEA Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
C.3 Energy Analysis Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
iv
Chapter 1
INTRODUCTION
1.1 Motivation
The purpose of this thesis project is to explore and improve the ferrocement building technique
as applied to sustainable housing. Ferrocement construction refers to a specific style of steel and
concrete construction. It’s a method that involves the use of much more steel reinforcement
layers in the structure and a concrete mix that includes sand rather than coarse aggregate.
The resulting structure can be thinner than traditional reinforced concrete construction, while
retaining superior strength. Material costs are typically competitive with conventional building
techniques, but there is considerable manual labor involved in setting up the steel reinforcement
in the field. This is the reason why today ferrocement is used mostly in countries with cheap
labor.
A major appeal of ferrocement is that buildings can become very interesting in shape.
Famous examples include the works of Antoni Gaudı́, Felix Candela and even ship hulls. Taking
advantage of shell shapes that naturally place the concrete in compression is a key part of
designing with ferrocement. Shapes such as domes and catenary beams are especially well-
suited. However, more complex shells are very difficult to analyze by hand and therefore a
computer tool such as Finite Element Analysis (FEA) can enable complex designs.
The term “sustainable housing” has many aspects. For clarity, it is defined in this thesis as
incorporating the following features:
Aesthetics A beautiful building will be cherished and maintained for a long time, reducing
the need for new construction and increasing occupant comfort.
Strength The structure’s strength to withstand natural disasters such as hurricanes and earth-
quakes is key in many regions and avoids costly reconstruction.
Construction Minimal ecological footprint during construction, including the use of locally
sourced materials, renewable materials, minimizing energy use, minimizing waste, mini-
mizing disturbances to neighbors and avoiding the use of ecologically-sensitive land.
Fire resistance Use of materials that give occupants time to evacuate. Preferably the struc-
ture should not be flammable nor weaken considerably during a fire.
Energy efficiency Minimize the energy required to heat, cool and electrify the building. In-
cludes strategies such as passive solar heating, insulation, air heat recovery, heat pumps,
efficient appliances and smaller living spaces.
Energy production To minimize the impact of external power plants, the building should
generate as much of its energy needs as possible. Strategies include solar water heating,
solar photovoltaics, biomass combustion and small wind turbines.
1
CHAPTER 1. INTRODUCTION
Thermal comfort Stability of the interior temperature can be increased with thermal mass
to balance changes in the outdoor climate. Rounded building profiles reduce the surface
area for a given volume, which reduces thermal losses of the building envelope.
Air quality Fresh air should be ventilated constantly and preferably filtered to minimize CO2 ,
dust and other pollutants. Ventilation rates are compromised by maximizing energy ef-
ficiency. Air quality also means minimizing materials which give off Volatile Organic
Compounds (VOCs) and other harmful gases to the occupants.
Natural light Sunlight in the building increases occupant comfort, although maximizing glaz-
ing comes in conflict with energy efficiency.
Noise External noise should be attenuated in the structure. Internal noise from occupants and
equipment should also be minimized.
Building codes and practices typically cover only the bare minimum of sustainable building
ideology. For this reason, voluntary accreditation was created to help fill this gap. Examples
include the American LEED [63], British BREEAM [12] and Swedish Miljöbyggnad [56]. Each
takes a different approach, where LEED focuses on lowering energy costs and site selection,
BREEAM focuses on minimizing greenhouse gas emissions in construction and operation and
Miljöbyggnad focuses on occupant comfort.
The target is to reach net-zero energy use for a typical family in a given year for any climate
in the continental United States, while allowing for architectural flexibility and strength that’s
not available in conventional house building methods.
1.2 Objectives
The thesis is comprised of three main objectives:
Improved field construction Evaluate ferrocement construction of today and of the past.
Evaluate other building methods to find improvements that can be carried over. Attempt
to reduce field labor and the number of steel reinforcement layers. Explore novel concrete
techniques, such as textile forms, fiber reinforcement, prefabrication and others. Incorpo-
rate thick insulation, which historically hasn’t been a priority in ferrocement. The goal is
to present a viable building technique that incorporates the lessons learned. This work is
presented in Chapter 2.
Structural analysis Create geometry modeling and FEA framework for efficiently analyzing
the structure as exposed to loads, including deadload, wind and earthquake loads. The tool
should provide automated stress and deflection reports on the entire building. The goal
is to optimize the structure’s strength and provide support for building code compliance.
The FEA tool will be used on three house case studies. Modeling and design work is
presented in Chapter 3. FEA work is presented in Chapter 4.
Energy analysis Evaluate the annual energy needs of three house case studies. Find the
appropriate energy system technologies and perform rough system sizing. This work is
presented in Chapter 5.
The project’s scope and focus is visualized in a more general context in Fig. 1.1.
2
CHAPTER 1. INTRODUCTION
Figure 1.1: Scope and points of focus for the thesis. Remodeling and demolition are not in the
scope.
1.3 Methodology
A strategy for achieving the objectives above is described here, specifying the owner profile,
building purpose and applicable building codes.
3
CHAPTER 1. INTRODUCTION
4
CHAPTER 1. INTRODUCTION
Supporting building codes are also part of the design process and are introduced in the
relevant sections of the report.
5
Chapter 2
FERROCEMENT
CONSTRUCTION
2.1 Overview
Ferrocement is a type of steel-reinforced concrete. The basic mixture is Portland cement and
sand (three parts sand to one part cement by volume) and no larger aggregate stones are
added [19]. Water is carefully controlled and the mixture has a plaster-like consistency. A
mixture that’s too wet is difficult to apply and forms weaker bonds with the steel. Chemi-
cal admixtures may be added to change workability, accelerate or retard set time, reduce the
water required (plasticizers), increase strength, increase adhesion, add color and enhance wa-
ter resistance. For example, acrylic is useful plasticizer to increase strength, water resistance
and improve adhesion to existing concrete (such as in cold joints). Even the basic cement-
sand mixture exhibits less cracking, higher strength, better adhesion and smoother surface than
traditional concrete mixtures.
Reinforcement includes more steel layers in the form of reinforcing bars, welded wire and
mesh, together known as the armature. The layers are close to each other, allowing for thin
shells that carry loads uniformly. The concrete mixture is worked into the armature by hand
like plaster. Significant labor is required to setup the armature skeleton, attach the mesh layers
and pack the concrete. There is more steel and less concrete used as compared to traditional
reinforced concrete. Steel can be replaced by natural fibers such as bamboo and hemp. Glass
and fiber reinforcement is explored later in this chapter.
The most common causes of failure of ferrocement (and reinforced concrete in general) are
steel corrosion and incomplete mortar penetration. As reinforcement rusts, it expands and
cracks the concrete. Cracked concrete allows more air and water to reach the steel, leading to a
destructive circle. The most common causes of corrosion are internal air pockets and exposed
steel. Air pockets are solved by vibrating the concrete right after pouring and/or manually
pushing the mixture as is done with troweling. Concrete cover is always required for any steel
such that it’s not exposed to the atmosphere. In ferrocement, since the outer-most mesh is close
to the surface, it’s best to use galvanized steel.
The failure mode of ferrocement is different from standard reinforced concrete. Given a
large impact or point load, a thin shell will deform (sometimes buckle) and result in cracks over
a wide area. The structure will however remain intact. In contrast, reinforced concrete will
tend to break, crumble and delaminate, causing the structure to fail.
One of the first documented uses was in 1847, when Frenchman Joseph-Luis Lambot filed a
patent for a wire-reinforced boat. It wasn’t until the 1950s that ferrocement was used in large
public structures as well as houses. For example, there was a community of builders around
Santa Barbara, CA at this time who found the low cost as well as earthquake and fire resistance
6
CHAPTER 2. FERROCEMENT CONSTRUCTION
of ferrocement houses to be perfect for the region’s needs. More examples are described in the
following sections. Today, ferrocement is used mostly in countries with low labor costs, usually
for water tanks and houses.
1. Reinforcing bars spaced in 40–50 cm squares. Rebar thickness is #3 (9.5 mm) or #4 (12.5 mm).
2. Welded wire attached on both sides of rebar. The wire is spaced in 15 cm squares and
5 mm thick.
3. Steel lath, or thin gauge expanded steel, is attached with hog rings to the armature’s inner
surface. Two layers of chicken wire can also be used.
4. Plaster sand mixed with cement and worked by hand and trowels into the armature
simultaneously from the inside and outside.
Figure 2.1: Wall section for ferrocement tanks and finished wall armature [18]
There are nuances to be aware of when creating the floor-wall and wall-roof joints, temporary
supports, penetrations, access hatch, etc. but in general construction requires little skilled labor.
It is a technique still used in countries where low material costs are important and labor is
inexpensive. Fiberglass Reinforced Plastic (FRP) has gained in popularity for small tanks and
steel plate for large tanks due to lower labor costs. Where the water is corrosive, an alternate
strategy is to cover a plastic tank with a thin ferrocement cover to protect from the sun and
extend lifetime considerably.
2.1.2 Boats
Boats were one of the original applications of ferrocement, extending into the 20th century. Dur-
ing both World Wars, ship builders were encouraged to use concrete to save steel. If constructed
properly, ferrocement boats are strong, low maintenance and of similar weight to composite and
wooden ones [27]. An amateur rush to DIY boat building in the 1970s and 80s, pushed by the
promise of cheap projects, led to many poorly-built boats and ferrocement’s poor reputation.
7
CHAPTER 2. FERROCEMENT CONSTRUCTION
This is one of the reasons why the technique is rarely in use today. Experienced builders and
quality inspectors have not been replaced over time and even a well-built boat is more difficult
to sell. Another perceived drawback is the uniqueness of each boat, as mechanized production
hasn’t advanced as compared to wood and composites. Nevertheless, it remains the cheapest
building technique for amateurs for boats longer than 8 m [26]. A few companies like Hartley
Boat Plans still sell DIY manuals and full-size patterns, advocating ferrocement.
The building technique is interesting to study because it is perhaps the strongest and highest-
quality ferrocement shell possible. It’s the most labor-intensive and contains the most reinforce-
ment layers. A documented example of an 8.5 m boat built in South Carolina in the 1970s that
is still in service is presented here [1]:
1. Rebar rod trusses are laid out, welded and covered in mesh. Truss skeleton hangs from
ceiling of a building
2. Trusses are connected with spring steel wire, 50 mm spacing top to bottom, 100 mm side
to side. Welded at front and back of boat, otherwise wire tied. This creates a fairly dense
mesh
5. Mortar with polyester resin is pushed through armature and troweled smooth
There are problems which arise with building boats this way. There can be eight or more
layers of mesh and the intent is to overlap layers such that largest hole is no bigger than a pencil
to improve strength and impact resistance. This dense armature is difficult to pack the mortar
into and air voids remain. Dry packing can also occur, where the mortar is dewatered through
pushing and doesn’t cure properly. It’s therefore not surprising that amateur boats tended to
be of poor quality.
An alternate technique called Laminated Ferrocement (LFC) was developed and patented in
the 1960s by Martin Iorns. Rather than the mortar packed into the armature, the idea behind
LFC is that the armature is packed into the mortar. First, a thin (3 mm) layer of mortar is
sprayed onto a mold of the hull. Expanded metal lath is placed on top and lightly pushed in.
Another thin mortar layer is sprayed on and the next layer of mesh set, rotated 90 degrees
8
CHAPTER 2. FERROCEMENT CONSTRUCTION
against the first. The process that Iorns used had five layers in total and a hull thickness of
19 mm. Air pockets and dry packing were eliminated. Labor was significantly reduced by not
building a skeleton frame or tying mesh together. This system underwent extensive testing for
commercial sea-worthiness certification by the U.S. Coast Guard in 1976, including tension,
compression, bending, shear, impact and fatigue [58].
LFC was the industry’s best chance to build with concrete. However, boatyards have moved
on to fiberglass and other materials, pushed by easier quality control, mechanization of produc-
tion, public perception and insurance rates.
2.1.3 Houses
There are many inspiring ferrocement houses in existence. Architects like Antoni Gaudı́ have
left their mark with imaginative and playful creations. Ferrocement’s plasticity and strength
has been used for roofs covering large open areas without columns, furniture built into the
structure and sculptures, among many others that are difficult with other building methods.
Illustrated examples of what is possible are presented in Appendix A.
Contemporary ferrocement house builders are rare in the U.S. One example is Steve Kornher,
who builds mostly in Central Mexico and documents his projects under his company Flying
Concrete [35]. Steve’s architectural creativity is exactly the aim of this thesis. He prefers to
use light-weight concrete, where locally-available pumice is mixed with cement in a 8:1 ratio
by volume for walls and 5:1 for roofs. He offers a 400 year warranty on roofs and prefers a
vault design for strength. The typical roof section is shown in Fig. 2.3. The other advantages
of vaulted roofs are little formwork needed, the inherent compression-only load distribution,
simple geometry and pleasant interior space.
Figure 2.3: Roof section used by Flying Concrete (adapted from [35])
An interesting point is that thick rigid insulation can be incorporated into this design,
although it is not typically used in Mexican houses due to the mild climate. Another point
is that reinforcing bars are not used at all for smaller vaults. The armature is temporarily
supported by wood and shaped welded wire. The supports are re-used between vaults. There
are some nuances to consider when tying into the wall and incorporating a second floor, such as
casting an edge beam and nailing the mesh into the wall. The thickness of the strong concrete
increases to 100 mm if light-weight concrete is used. The bottom-most thin shell is placed first
so that the concrete above doesn’t sag the mesh under its weight. The advantage of light-weight
concrete is that it’s easier to work with and can even be nailed into.
Walls are typically constructed out of concrete blocks and reinforced. Arches over doors and
windows are also done with blocks. If a more elaborate window shape is needed, a ferrocement
mesh is used. Windows and doors are often custom-made; it’s important to deter theft with
9
CHAPTER 2. FERROCEMENT CONSTRUCTION
steel bars so it’s an additional challenge to create appealing designs that don’t look like jail bars
and fit into the architecture. The climate allows for single-pane windows and doors. Examples
of construction by Steve’s crew are shown in Fig. 2.4.
Javier Senosiain of Arquitectura Orgánica also builds beautiful houses out of ferrocement (see
Appendix A for examples). Other builders have variations of these methods, by for example
building walls in ferrocement, not using light-weight concrete and installing standard windows.
However, the basic approach remains similar and is considered the “base” procedure to build
upon later in this chapter.
10
CHAPTER 2. FERROCEMENT CONSTRUCTION
There are many other notable engineers of the 20th century. What makes large shell struc-
tures so interesting to study is each project’s unique approach and the improvements in builders’
techniques. Lessons such as strength in double curvature, prefabricated patterns and ribbed
shells carry over to ferrocement on a house scale.
2.1.5 Sculpture
Sculptures have been a popular application of ferrocement and perhaps the most experimental in
terms of techniques. Commercial examples range from large donuts to two story cowboy boots
and other such gimmicks to attract clients. Backyard artists create garden bridges, statues,
gazebos, benches, fake wood and many others. Since structural strength is not as critical as
buildings or boats, sculptures are the natural choice for exploring new ways of using ferrocement.
Notable examples include:
Fiberglass, hemp and other flexible meshes soaked in mortar and built up in small sheets
like paper maché
Complements to mesh reinforcement for difficult curves such as wire weaving and fibers
Foam cores carved by hand or hot wires
Foamed concrete used as a lightweight core, reinforced on surface
Various molding techniques such as painted silicone (for copying), earth and sand casting
as well as textiles
Coloring with acrylic or latex paint, acid staining and pigments
Many of these techniques carry over to structural work and are explored in more detail later
in the report.
11
CHAPTER 2. FERROCEMENT CONSTRUCTION
Unfortunately, stick building has many serious disadvantages that motivate a different ap-
proach. The most significant is the low strength against high winds. Every year the hurricane
and tornado seasons destroy houses in the U.S. despite building code provisions, sometimes even
entire towns [11]. Many more are heavily damaged by flying debris and fallen trees. Another
disadvantage is vulnerability to moisture and significant damage from flooding as well as the
danger of mold growth. Insects such as termites are a nuisance in some areas and increase the
cost of maintenance. Fire resistance is relatively poor even with fire retardants impregnated
into the wood, requiring expensive sprinkler systems in public buildings and more expensive
insurance. Noise from footsteps and floor creaking is common.
Although not necessarily inherent to wooden houses, thermal performance of the average
house is poor. Air tightness relies on a vapor barrier liner underneath the exterior siding and
can easily be installed incorrectly or damaged. None of the other wall components provide air
tightness unless there is a brick or plaster facade. Insulation is typically fiberglass or mineral
wool in-between studs. By design this space is usually limited to 100 mm and thermal bridging
of the studs is not accounted for. External rigid insulation can correct the problem but is
relatively rare. Insulation of the roof is usually done by actually insulating the top-most ceiling
12
CHAPTER 2. FERROCEMENT CONSTRUCTION
and leaving the attic space uninsulated. This can also be corrected by also adding insulation
to the roof framing.
This proven building technique enjoys commercial success in the U.S., especially for stores,
restaurants and similar applications. The foam structure is quickly erected and when combined
with an efficient shotcrete contractor the building is quickly completed. Advantages include fire
resistance, strength against earthquakes and strong winds, energy efficiency through insulation
and envelope tightness, low maintenance, flexibility in interior/exterior surface finish, noise
absorption and favorable total lifetime cost. All parts of the structure use the same system -
floors, wall, roofs, columns and arches. There is flexibility in decorative features as the foam
can be cut to any (flat) shape easily. Building code conformity is long-established. Interior
concrete serves as thermal mass.
This system can be used by the owner-builder without even the need to purchase specialty
panels. Rigid foam boards and steel mesh are readily accessible. Foam can be cut to the desired
design. Specialty mesh ties and spacers are available (and even patented) from companies like
SIPcrete, but these also be made by hand from plastic or metal rods. Finally, shotcrete does
not require specialty contractors. One can spray a concrete mix with a tool like the Stucco
Sprayer (see Fig. 2.8) offered by MortarSprayer.com [43] connected to an air compressor. It’s
possible to apply up to 5 m3 /hr of concrete this way.
The key disadvantage of this building system is that non-flat surfaces require special foam
molding. Architectural freedom is therefore limited without changing the structure of the
insulation. Such changes are explored later in this chapter.
Concrete isn’t the only type of skin commercially available for SIPs. Oriented strand
boards (OSB) are made of fast-growing trees and replace concrete on both sides of the foam.
13
CHAPTER 2. FERROCEMENT CONSTRUCTION
They are purchased as a complete panel. The advantage of this type of SIP is less labor: after
erecting the panels and sealing the joints, external siding can be directly attached and the in-
terior surface finished. Companies like Thermocore [57] offer custom-cut panels with window,
door and conduit provisions. They claim that a house shell can be erected in 2 days. The
wood’s softness is an advantage when attaching to it. The disadvantages of using wood are
reduced fire resistance, reduced strength and vulnerability to moisture and insects.
ICFs are most appropriate for basement walls, where the outside foam can be covered in
plastic and the system saves on formwork. It is also apparently a better system with respect to
sound attenuation inside the building than SIP.
14
CHAPTER 2. FERROCEMENT CONSTRUCTION
A disadvantage of this system is the cost of ordering a concrete pump mixer and pump trucks
multiple times during construction. Great care must be taken to avoid weaknesses in the foam
blocks because it’s possible to blow out a wall entirely during a pour. Another disadvantage
when compared to SIPs is that the concrete isn’t as effective as a thermal mass. Also, there
is more work to finish the foam surfaces as compared to the SIP concrete surfaces ready for
plaster or stucco. Construction in general is more complex than SIP.
The roof cannot use this system, unless it is a flat roof. Architectural details also require
another technique such as SIP. Overall, it is difficult to recommend the ICF system over SIP.
It’s noteworthy that Am-cor’s chief architect, Angus W. Macdonald, has experimented with
a wide variety of meshes and types of construction in general. He notes that expanded metal
laths have the most favorable cement interlocking and provide a large surface area for the cement
to interact with. This is critical for a one layer system, but this experience is also useful for
multi-layer systems where the type of mesh often isn’t seen as important.
Disadvantages of this system include thermal bridging of the steel frame, limited customiza-
tion and low internal thermal mass. Insulation in-between wall beams must be supplemented
by, for example, rigid insulation on the inside to completely cover the steel’s conduction path.
Thermal mass inside must be done separately, adding cost. The lack of architectural freedom is
difficult to overcome, as the prefabricated shapes are not intended to be curved or customized
15
CHAPTER 2. FERROCEMENT CONSTRUCTION
to a great extent by the manufacturer. The structural strength of the structure appears to be
engineered into the steel beams to conform to IRC and other codes.
2.2.5 Earthbags
The earthbag system involves building walls and sometimes roof using polypropylene bags or
tubes filled with local soil [34]. Bags and tubes with misprints are purchased inexpensively. The
bags are stacked like bricks on top of a concrete or stone foundation. Each bag is tampered
after being placed. Barbed wire strands are placed horizontally in-between each course to bond
the bags together and add tensile strength. Vertical reinforcing bars are spaced every meter.
A reinforced concrete ringbeam is poured at the top of the wall if a wooden framed roof is
installed. Otherwise, the bags are shaped into a dome. Wooden frames are used for doors
and windows. The interior and exterior are plastered with a mix of straw, clay and sand and
painted with a lime mixture to resist water. Examples of houses under construction are shown
in Fig. 2.11.
The entire process relies on few external materials and very inexpensive, ecologically-friendly
buildings can be constructed this way. Since rounded shapes are more stable, earthbag buildings
heavily incorporate them. In some cases, building code compliance is achieved by incorporating
conventional wooden/concrete posts and beams. Thermal mass is very high, which makes for
comfortable buildings in hot climates. The earthbag system is very similar to traditional adobe.
In climates where insulation is necessary, an additional external layer of bags can be incor-
porated filled with insulation. Several projects have used insulation fill directly, such as volcanic
rock, which solves the problem with one layer of bags. Since material weight is reduced, the
building process is also quicker. The projects built this way have been in areas not regulated by
building codes, however. The concept of insulation-filled bags is explored later in this chapter.
The disadvantage is that manual labor is extensive with earthbags, with relatively few
opportunities for automation. Moisture must be carefully controlled by roof overhangs and
good drainage.
2.2.6 Earthships
Earthship is a building system popularized by Michael Reynolds that strives for completely
off-grid houses [51]. The defining characteristic is the use of automotive tires recovered from
landfills to build the walls. Tires are stacked like bricks and each tire is filled with soil from
the site and compacted. Non load-bearing walls are often built with bottles and soil mortar,
allowing for curved and colorful designs. The walls are backfilled with soil from the north,
making the roof easily accessible. Strength of the walls is enormous, as is the thermal mass.
16
CHAPTER 2. FERROCEMENT CONSTRUCTION
Walls are plastered and painted on the interior. Flooring is placed directly on compacted soil
after good drainage is ensured. Roofing consists of tree log beams and boards, plastered on the
outside. In desert climates of the South-Western U.S. where Earthships are most commonly
found, insulation is not even necessary as thermal mass evens out the hot days and cold nights.
The earth’s naturally stable temperature is also conducted through the thick walls.
More than just an ecological structure, the house as a complete system is covered by the
Earthship system as shown in Fig. 2.12. Building floorplans are based on passive solar concepts.
The buildings stretch East-West, are covered completely in the North and have all windows
facing the South. This allows for copious amounts of natural light and heating. The roof and
window shutters are designed to control the amount of sunshine reaching the interior, based
on the local climate. An internal greenhouse is right behind the large windows, resulting in a
source of food and cooling from the plants. Usually a windowed wall separates the greenhouse
corridor space from the interior to mitigate solar heating and serve as a thermal barrier.
Rainwater is collected into a cistern and filtered for consumption. Water is recycled several
times: gray water from sinks and showers feeds botanical cells and finds re-use in toilets. Black
water is run through another set of botanical cells. The result is that the greenhouse and outside
plants are nourished, while eliminating the need for a sewage connection.
Electricity is generated using solar PV panels facing the south and stored in batteries.
Additional generators such as wind turbines are sometimes added in remote areas. Direct
current appliances reduce the total electricity use. HVAC energy needs are low.
Earthships have been built in a variety of climates. Wet regions require additional drainage
provisions, more roof cover and water-resistant plaster. Cold climates incorporate insulation
behind the tire wall. Architectural flexibility is fairly extensive due to the use of soil and/or
concrete shaped by hand. Built projects exhibit curved walls, colored glass embedded in walls,
arches and other creative features.
Manual labor required to construct the structure is extensive, although relatively unskilled.
Filling tires with soil and compacting with sledge hammers is perhaps the single most time-
consuming part of construction. The structure is most often built by the owners and the hired
17
CHAPTER 2. FERROCEMENT CONSTRUCTION
1. A reinforced concrete ringbeam foundation is built. The ringbeam transfers loads to the
ground, provides stiffness and holds the airform during construction.
2. The airform membrane is attached to the ringbeam, sealed and inflated with fans.
3. Polyurethane foam insulation is sprayed onto the membrane from inside the airform. The
insulation has the added function of stiffening and stabilizing the airform. It is 50–100 mm
thick.
4. Reinforcement bars are attached to the inside of the insulated shell. Windows and door
forms are attached to the polyurethane shell to block off the shotcrete.
5. Shotcrete is sprayed from the bottom up in rings inside the structure in layers, achieving
approximately 75 mm thickness.
6. Air pressure is removed once shotcrete cures. Windows and doors are cut out. Overhangs
are attached to the outside.
The finished structure is shown in Fig. 2.13. Hiring specialty contractors to spray foam
and shotcrete is the most expensive part of the construction process, but the contractors can
be efficient due to the simple geometry. The owner-builder can complete all other tasks with
unskilled labor. Various cost estimates place concrete domes as comparable to wooden stick-
built construction or a bit more expensive. The total cost is considered much lower due to low
maintenance, long lifetime, energy efficiency and resistance to natural disasters.
Disadvantages of concrete domes built this way include architectural limitations, echo prop-
agation inside and sub-optimal use of interior volume. A large dome will have a two story
18
CHAPTER 2. FERROCEMENT CONSTRUCTION
arrangement of rooms around the perimeter and the tall center of the dome can be tricky to
utilize.
There are several lessons learned which may be applied to ferrocement. The concept of
inflatable formwork needn’t be restricted to dome shapes. The membrane can be made cheaply
to custom shapes. The option of creating smaller inflatable forms can be considered, where the
outside of the form is used instead. Rigid insulation that’s sprayed on will be revisited later in
Section 2.3.1. The shotcrete shell’s strength, durability and versatility are proven by decades
of dome building, even without the benefit of mesh reinforcement. Dome architecture has been
made more interesting in projects where large sections are cut out, blurring the transition with
outdoor space and a tarp-like roof effect.
Attempts to contact the company to find out if the system is available outside of Japan
were unsuccessful. This highlights a key disadvantage of dependence on the manufacturer. It’s
likely that the system would need to be structurally reinforced to meet IRC, which is why
it isn’t available in North America or Europe. However, there is an exciting opportunity if
a concrete shell is added. The foam’s inherent weaknesses in terms of debris protection and
fire resistance can be corrected. By applying a layer of shotcrete on the interior, this system
effectively becomes similar to monolithic domes described in §2.2.7 with the key advantage of
no inflation equipment or foam spraying contractors. Shotcrete application would also become
simpler, as the foam is much more rigid. A stucco or thin concrete shell exterior would further
protect against debris in strong winds.
19
CHAPTER 2. FERROCEMENT CONSTRUCTION
on site and an entire building printed as shown in Fig. 2.15b. A high level of automation is the
goal.
The technology is still in its infancy, but has already been used to produce intricate sculp-
tures and building wall concepts. It is not yet commercialized and the first printed building is
yet to be completed. The high level of automation is attractive, but the technique is not suitable
for ferrocement. Objects can be embedded in the wall, though a thin shell with multiple layers
of steel does not appear to be feasible to print. Reinforcement will need to be improved to
provide safe housing and allow for thinner concrete sections. Using this technology to prefabri-
cate complex wall and roof panels may well become an attractive alternative to on-site gantry
printing.
20
CHAPTER 2. FERROCEMENT CONSTRUCTION
Concrete cloth has been used on housing projects, mainly as a skin on an existing building
for the purpose of hurricane protection. It protects the building very effectively against flying
debris. Another housing application has been emergency shelters and military outposts, where
the cloth is sewn at the factory to be an inflatable form. Installation on-site only requires two
hours, where the fabric is rolled out, door installed, fabric anchored, inflated and soaked with
water. It becomes a usable shelter after one day (see Fig. 2.16b). The 54 m2 version of the
shelter costs about $35,000.
The clear advantage of this fabric is how quickly it can be deployed and adapted to any shape.
In ferrocement, it could replace the outer-most steel mesh layers and also serve as formwork. For
example, a shell could be built by attaching the fabric to both sides of a reinforcing bar skeleton
and fiber-reinforced concrete used as the middle binder. Another attractive application is for
smaller details and joints where bending steel mesh is arduous, such as column-roof junctions,
furniture and artistic features.
The material is fairly new to market, having been patented in 2010. Direct competition
is non-existent, which is reflected in the price. There is considerable engineering involved in
its manufacturing, including fabric fiber spacing, binding of the three layers and selection of
cement additives. Not surprisingly, it hasn’t been reproduced by amateurs at this point.
Rigid foam
Closed-cell rigid foam is the most effective type of insulation. It is readily available in 1.2 x 2.4 m
boards and various thicknesses. Polyisocyanurate is more expensive than EPS or XPS, but its
performance and lower required thickness can justify the extra cost. The problems in using
these boards with ferrocement include the limited flexibility of the boards to adjust to curved
surfaces, gluing consecutive layers to the concrete and other foam before the final lath can be
21
CHAPTER 2. FERROCEMENT CONSTRUCTION
attached as well as difficult penetration of lath ties through the foam. All of these challenges
can be solved, but the installation process becomes a labor-intensive combination of cutting,
gluing, hand-fitting and tape sealing each layer. The performance may in the end be similar to
simply using loose fill insulation.
Sprayed foam
Sprayed polyurethane and icynene have excellent insulating performance, adhere well to con-
crete, are airtight, are quickly applied and form arbitrary shapes. However, it is the most
expensive solution. DIY kits such as Foam It Green cost $445 per m3 of finished foam [24].
With a desired thickness of 20 cm, a modest family house would cost $20–30,000 to cover.
Contractors cost more for insulating existing homes, but may be cheaper in spraying one thick
continuous shell. For comparison, XPS boards cost $250–350 per m3 but require more labor to
install in this case.
Light-weight concrete
Light-weight aggregates can be mixed with cement to provide insulating concrete. Internal
thermal bridging guarantees that the U-value will never match pure insulation and must be
thicker. However, locally available materials such as volcanic rock (scoria), perlite, pumice,
vermiculite or recycled EPS make this a potentially inexpensive option. In ferrocement, it
could be used as an external shell with metal lath attached to the structure in stages to provide
formwork. Rastra’s EPS blocks, for example, are made of 85 percent recycled EPS [49].
Papercrete is composed of waste paper shredded and mixed with Portland cement. A mix-
ture with paper and cement in equal weight shrinks by 5 percent [39]. Cement can be avoided
altogether by substituting clay soil and lime, but the increased shrinkage and drying time makes
it difficult to work with. There is a range of materials that can be mixed with, including ground
polystyrene. The technique is attractive because of its very low cost, including DIY mixers that
are slowly towed trailers with a lawnmower blade powered by the wheels. Papercrete is not
recognized as a structural material in the IRC despite its wood-like strength. A thin, sealed
protective concrete shell is needed to protect from moisture.
Aircrete or cellular concrete is another type of light-weight concrete that’s mixed with foam-
ing agents. As it cures, the foam creates small air pockets that are 50–85 percent by volume
in the finished product. The mixture is very fluid before it cures and requires enclosed form-
work. Commercially available aircrete comes in factory-cast blocks, as spray foam installed by
contractors or can be made on-site by the builder. Contractors cost significantly more than
spray foam. Foam agents can be purchased to use on-site but require special foam generator
equipment. If the volume of cast parts is high on a project, the expense may be justified because
casts are easy to handle and do not develop shrinkage cracks.
Bagged insulation
Earthbag builders have experimented the most with plastic bags filled with materials such as
shredded EPS, perlite and scoria. The bags (or tubes) can be stacked like bricks and lightly
tampered, forming a thick shell next to the ferrocement structure. String ties attached to the
ferrocement can hold a metal lath on the outside of these bags for a final thin concrete layer.
The advantage of this approach is that recycled or low-cost local materials can be used and the
bags easily follow the curves of the structure. Labor in filling and stacking the bags is quick. Air
voids need to be avoided, especially between bags. It may be necessary to use a small amount
of household spray foam.
22
CHAPTER 2. FERROCEMENT CONSTRUCTION
It should be expected that a material like shredded EPS will not perform as well as a rigid
board. However, with tight stacking, there’s no reason why it shouldn’t perform well. The
advantages in cost and simplicity make this a very attractive solution.
Fabric forming lends itself to prefabrication as well as laminating with several layers of
non-metallic mesh. Forms for roofs should be cast upside-down so that they naturally take
on a catenary shape. Field casting would be most suitable for tree-like columns. An entire
building could be cast in sections, although it becomes apparent that matching edge geometry
on complex curves is difficult. The only reasonable solution would be to design with straight
joints in mind that are consistent for the entire shell. It would not be practical to create more
1
Centre for Architectural Structures and Technology
23
CHAPTER 2. FERROCEMENT CONSTRUCTION
than a few curved beams for the cast’s frame. Accurate modeling of the final shape is tricky,
especially local buckling. These difficulties are worth solving, however, because of the low cost
of this type of formwork.
Cheap raw material should lead to lower cost with more mass production2
Basalt production is still low-volume, but the material holds advantages over other FRP
types for its resistance to alkalinity and lower cost. Research has been limited, although basalt
beams were found to perform favorably according to ACI 440 [46]. ACI 440 is a companion
code to ACI 318 which deals with FRP reinforcement [6].
Thin basalt rebar is as flexible as a rope, which creates intriguing possibilities such as
overlaying on top of an inflated form. Laminating complex curves in a mold also becomes much
easier with the mesh. Since concrete cover for corrosion protection isn’t necessary, shells can be
built with the LFC technique to be very thin and still include the necessary reinforcement to
comply with ACI 318 and ACI 440.
24
CHAPTER 2. FERROCEMENT CONSTRUCTION
Figure 2.18: Deflated and inflated form of the Mexican Whale House. Yellow coat on inflated
form is the start of polyurethane foam [52]
25
CHAPTER 2. FERROCEMENT CONSTRUCTION
is air-tight. Once cured, laminate layers would be built up on the outside of this layer. Bagged
insulation would be placed before a final exterior LFC shell is formed. This approach reduces
costs from spray foam, although it complicates fabricating the tough fabric to be inflatable and
incorporate penetrations without cutting the cured layer.
A technique being developed by Monolithic Constructors is the uninsulated EcoShell [42],
which is built by inflating an air form and spraying shotcrete directly on the outside. A thin
layer is set and basalt ropes are wrapped around the structure at 200 mm spacing. Once a
reinforcing grid is finished, another layer of shotcrete is sprayed. The final thickness is about
38 mm. This provides the same advantages as using Concrete Cloth at a lower cost. The single
grid of ropes can be complemented by more mesh layers to create a laminate.
2.4.1 Foundation
The only type of foundation considered here is the on-grade insulated slab with spread footings
under the walls. If not insulated, the footings are required by code to be under the local soil frost
line to avoid frost heave damage. Starter bars are embedded in the footings to attach the wall
armature and concrete later. The footings are reinforced with bars and the slab is reinforced
with welded wire. All utilities, such as water pipes and conduits must be placed beforehand.
There are several alternatives to concrete footings, such as using stones or earthbags. These
options are not explored here since the pouring of concrete for the slab is still required.
Slab foundations are very common in house construction. The only challenge lies in insu-
lation: typically rigid insulation is placed under the slab but the footings create a significant
thermal bridge. A suitable commercial solution is found in Irish company’s KORE system,
which claims to have the lowest losses on the market and has been approved by the PassivHaus
Institute [2]. The system includes EPS blocks under the entire foundation as shown in Fig. 2.19.
Floor heating tubes can be set into the slab.
The slab can be floating type, or monolithic. For a double shell wall, double footing at
external walls is necessary to completely break the thermal bridge. EPS-100 and EPS-300
denote the load bearing capacity of the insulation. According to an engineering report done for
the foundations of PassivHaus using this system, the bearing capacity of EPS-100 is 48 kPa3 and
EPS-300 is 120 kPa [32]. Standard EPS boards rated according to ASTM C578 are available
3
The maximum bearing load is based on 1 percent compression of the insulation. ASTM-C578 ratings are
based on 10 percent.
26
CHAPTER 2. FERROCEMENT CONSTRUCTION
Figure 2.19: Slab insulation detail with molded EPS footing forms
in the American market, with ratings of 35–276 kPa. Loads on the insulation must be checked
during design.
KORE literature claims a U-value of 0.08 W/m2 -K for this floor system. They typically
install a Radon barrier below the concrete. EPS does not transmit moisture, so there’s no need
for an additional external barrier.
This design qualifies as frost-protected shallow foundation, which allows building above
the frost line in northern climates. That reduces the amount of concrete that needs to be
used. The insulation is also used as integrated formwork, which reduces labor and waste as
compared to a wooden framework. The labor in attaching insulation later is also eliminated.
Shallow foundations are commonplace in Scandinavia [44]. They are addressed specifically in
IRC §R403.3, which deals completely with frost protection rather than energy efficiency.
2.4.2 Columns
Columns follow the slab in construction sequence. A hybrid column construction technique
is considered here that merges a steel armature with textile forming. A steel rebar skeleton
is erected in accordance to standard details and connected to the starter bars in the slab
foundation. Standard here refers to vertical bars with circumferential rings around them. Layers
of mesh are wrapped around and tied to the skeleton to finish the armature.
Given a tight seam, an inner liner of stretch fabric shouldn’t be necessary. Two sheets of
polypropylene fabric are cut oversize to match the shape of the armature. Clamps are used
to tightly close the form on both sides. The pour is best done in three stages to control the
pressure: a minimal initial pour is done to minimize leaking at the bottom, then the form is
clamped and poured to half way up and finally the top half finished. The fabric needs to be
supported vertically from the armature to prevent horizontal creases. Local tension can be
applied to, for example, create a root-like pattern at the bottom of the column. The form needs
to be vibrated from the outside right after each pour. Some of the mesh will not be covered by
the pour, so plastering with concrete is necessary after the fabric is removed.
An alternate solution is to apply concrete directly to the armature without fabric using a
drier mix. In this case, it may be best to wrap the mesh around the skeleton in stages so that
the column can be filled quicker from the top and troweled at the mesh. Adding fibers to the
concrete is a good idea in either case to limit surface cracking.
27
CHAPTER 2. FERROCEMENT CONSTRUCTION
opposed to forcing it to an existing rebar skeleton. Using only mesh is more difficult to justify
to building codes. To make field erection efficient, the following needs to be done:
Design Computer design contains every bar’s shape, location and overlapping bars.
Shop preparation Bending to be done in a shop - best by an automated bender and cutter.
Otherwise, a template system like an overhead projector can be used, projecting on a
floor and bars shaped with power tools. Projected image on ground is actual scale, so the
projector effectively makes custom templates and saves considerable work in measuring.
Each rebar piece is labeled. Bar overlaps are also marked.
Beam welding Beams of rebar are welded in the shop. Arches, columns and other small pieces
can be wire-tied in the shop as well.
Organization Bent rebar is stacked in groups so in the field they’re sequentially placed. For
example, vertical bars are attached to the slab from left to right in a given room and
horizontal bars from the bottom up.
Rebar attachment A quick method of attaching bars together is needed. Automatic wire
tying tools are available which take only one second to wrap and twist wire at a joint.
Welding Critical joints and beams need to be welded to further stiffen and strengthen the
structure once the entire skeleton is assembled. A small MIG welder is sufficient and can
run on a normal generator on-site.
28
CHAPTER 2. FERROCEMENT CONSTRUCTION
29
Chapter 3
3.1 Overview
This chapter presents the methodology and implementation of structural shell modeling. The
three house case studies are built using this system and also presented. The goal of the system is
to make the design and structural analysis as quick and easy as possible. To this extend, various
software solutions are explored. It’s unavoidable that some complexity exists, for example in
scripting, but the overhead of learning and using the tools should be as small as possible. The
designer should be able to convert hand sketches, rough dimensions and ideas into a functional
model with a few hours worth of work.
3.2 Functionality
The scope of features desired in the modeling system is:
3D geometry The entire house shell must be modeled as a solid, with discrete reinforcing
bars. Other reinforcement, such as welded wire and meshes are best modeled as shells
offset from the rebars. Ease of use of the CAD software is important, as is affordability
to the designer/builder.
2D drawings The 3D model must be able to export 2D drawings of any desired parts of the
geometry. This includes cross sections with dimensions. Every rebar’s dimensions and
shape can be exported separately to aid in shop bending.
FEA link The relevant geometry must be cleanly exported to the FEA software such that
additional tweaking isn’t necessary. This includes assemblies and material types. Relevant
geometry is the structural framework and doesn’t need to include non-structural details.
Scripting and automation The entire design and analysis process must ultimately be rep-
resented in macros and scripts. This is critical for automating routine tasks in creating
geometry and performing the structural analysis. There needs to be a template for any
30
CHAPTER 3. MODELING AND DESIGN OF CASE STUDIES
project, where the user concentrates only on the structure’s features and has the remain-
ing workflow automated. Common text files should include all inputs such as material
properties and loads. Text files should also be used where possible for transferring infor-
mation between different software packages and debugging. The scripting language should
be easy for the user to learn.
Post-processing All tasks such as generating standard views, key drawings, analysis results
and reports need to be automatically done for the user. Material quantity estimates for
steel, concrete and insulation need to be available.
It’s also important to clarify the limits to the scope. Building Information Modeling (BIM),
which involves all building elements is not the focus of this project even if the 3D model is
a significant contribution in that direction. Likewise, utilities such as electrical conduits and
wiring, water pipes, HVAC and lighting are not considered1 .
31
CHAPTER 3. MODELING AND DESIGN OF CASE STUDIES
L
S = 2a sinh
2a
L
d = a cosh −1
2a
where d is the dip of the suspended cable from points at the same level, L is the span, S is
the cable length and a is the horizontal portion of the tension divided by the cable’s weight per
unit length (see Fig. 3.2). The equations are solved by trial and error.
Circular (Roman) arches are the strongest shape to resist a uniform pressure load. They
are also the best choice where supporting a flat surface and a non-uniform vertical load which
increases down the arch, such as a bridge or aqueduct. A semicircular dome roof has the
advantage of no thrust forces when loaded by a uniform pressure, unlike a catenary roof. It is
the first choice in design of a vaulted dome ceiling that will support another story.
Pointed (Gothic) arches are slender and provide the greatest strength resisting a point load
in the middle, up to the extreme case of becoming triangular.
32
CHAPTER 3. MODELING AND DESIGN OF CASE STUDIES
(a) Internal pressure (b) Fabric weight (c) Point load and fabric weight
3.4.3 Ribs
Ribs are beams connected to the shell which serve to increase load-carrying capacity. They
are especially needed to reinforce large penetrations and resist concentrated loads, where shells
are at their weakest. Concentrated loads include columns, upper story connections, external
attachments and others. The addition of rib beams is one way to strengthen the structure
without changing its shape.
33
CHAPTER 3. MODELING AND DESIGN OF CASE STUDIES
exactly this, as shown by an example in Fig. 3.4 of a 2D wall profile. Macros were created to
automate the analysis for exploring such designs.
Figure 3.4: Topological optimization of a plane. Circle holes are unmeshed to force their shape.
Seven loads are on top edge and five supports on bottom. The red shape is the optimized shape
for 50 percent material reduction.
34
CHAPTER 3. MODELING AND DESIGN OF CASE STUDIES
its highest point. A large, tapering center column is necessary to support the roof. There are
12 sections. Windows are 1.6 m in diameter. The “eyes” are 2.5 m at their largest dimension.
The shape is created by defining a rib frame and inflating the surface. Continuity of curvature
is very important between the shell and column to minimize stress concentrations, which is why
inflation is done on the entire structure. The bulging shell is not an ideal shape in terms of
distributing dead load in that bending moments are created along the wall. For this reason,
stiffening ribs are added inside the “creases” in the shell. The 0.3 m ribs run continuously from
the wall, to the roof and down the column.
The column is hollow so that it may be used to channel rain water from the roof, either for
storage or discharge through a pipe. It’s also a convenient structure to wrap a spiral staircase
around.
Internal walls and the second story floor are not modeled even though they connect directly
to the outer shell. The reason is to ensure that the outer shell is structurally sound and can
withstand all loads on its own.
35
CHAPTER 3. MODELING AND DESIGN OF CASE STUDIES
36
Chapter 4
STRUCTURAL ANALYSIS
4.1 Overview
The intent of the structural analysis presented here is to generally establish a building’s safety.
The analysis methodology and the process of selecting software are presented. Several building
codes are referred to for establishing loads and load combinations. Simplified acceptance criteria
are presented for tension, compression and bending. Each of the three case studies is analyzed
for its ability to resist dead and live loads.
4.2 Methodology
It is not practical nor accurate to calculate each member of a shell structure according to
code formulas which assume rectilinear geometry. Fortunately, an elastic analysis of the entire
structure is allowed in ACI 318. Shell theory may be used, which simplifies the analysis geometry
and reduces calculation time as compared to a fully solid model. The model’s concrete structure
is assumed to be uncracked, linearly elastic, homogeneous and isotropic. In effect, only concrete
is analyzed and the reinforcement is designed for the structure based on tensile forces that have
to be supported.
The codes (ACI 318 and ASCE 7) require designing for service and ultimate strengths.
Service loads are what the building is expected to experience in its lifetime. As such, the building
is expected to be repairable after loading events (i.e. no plastic deformation in reinforcement,
limited crack widths). Ultimate loads are extreme cases, where only protection against collapse
is to be designed for. In this project, the evaluation is based on ultimate strength only.
Elastic, linear analysis is computationally efficient in that a given load can be solved for only
once and combined with others in post-processing. For example, dead load can be solved with
a factor of 1.0, as can each of the wind and seismic directions. In post-processing, the result
components (stresses and displacements) are scaled and added for any desired combination.
ANSYS does the result manipulation for each node on a component basis, so that derived
results such as principal stresses are calculated only at the end. For example, a given node’s
stress in the x direction for load combination 1.2D +Evertical +Ehorizontal (dead load and seismic
factors) would be:
37
CHAPTER 4. STRUCTURAL ANALYSIS
if tension occurs. This approach was not used here because it is much more computationally
intensive and requires more detailed modeling.
To evaluate results, traditionally the structure is divided into sections and reinforcement
is designed based on stress results. The approach taken here is to evaluate each element of
the structure to ensure that a given ferrocement technique is adequate. Any areas that need
additional material can also be identified, namely, where tension forces are beyond the rein-
forcement’s strength or where concrete crushing occurs.
The amount of steel in the total composite is described by the volume ratio, Vr :
Vreinf orcement
Vr =
Vtotal
IFS 10-01 recommends a minimum Vr of 1.8% to obtain ferrocement characteristics. For a
welded wire mesh of N layers, wire thickness dw , concrete thickness h and wire spacing D, Vr
becomes [31]:
πN d2w
Vr =
2hD
The Young’s Modulus and material density used in the analysis are the combination of steel
and concrete:
4.4 Loads
4.4.1 Load combinations
ACI 318 requires the following load case combinations to be evaluated:
U = 1.4D
U = 1.2D + 1.6L + 0.5(Lr or S or R)
38
CHAPTER 4. STRUCTURAL ANALYSIS
U = 1.4D
U = 1.2D + 1.0WN,S,E,W
U = 1.2D + 1.0EN,S,E,W
U = 0.9D + 1.0WN,S,E,W
U = 0.9D + 1.0EN,S,E,W
E = Eh ± Ev
Eh = 0.385D
Ev = 0.28D
The vertical component is added in load combinations with 1.2D and subtracted in load
combinations with 0.9D.
39
CHAPTER 4. STRUCTURAL ANALYSIS
First principal stress, σ1 , for shell inner and outer surface. The maximum is reported.
Third principal stress, σ3 , for shell inner and outer surface. The minimum is reported.
Crushing is checked by comparing the largest compressive stress, σ1 , against the con-
crete compressive strength. It’s conservatively assumed that only concrete resists compression.
IFS 10-01 recommends that compression be limited to 0.45fc . Since each node is reported in a
generated table, it’s straightforward to find where crushing occurs and how large the region is.
The acceptance criterion is therefore:
40
CHAPTER 4. STRUCTURAL ANALYSIS
shear and tension using various adjustment factors. Crack widths would also be calculated.
The analysis done here provides a good estimate of stresses and displacements as well as regions
that would need additional strengthening in a detailed design.
4.6.3 Results
Representative displacement plots for two load cases are shown in Fig. 4.1, showing a roof sag
in the 1.4D load combination and a wall bulge in the 1.2D + Enorth combination. Specific notes
for each load combination are discussed here:
1.4D Roof sags by 5 mm. Tension due to bending is highest on the roof’s inside surface and
requires another layer of mesh reinforcement.
Seismic The walls bulge out a little due to horizontal acceleration. The maximum of 2 mm
occurs on the southern wall during the 1.2D + Enorth load combination. Generally the
1.2D combinations yield higher displacements and stresses than the 0.9D combinations.
Roof sag increases to 6 mm for Ewest and Eeast . The largest tension occurs on the roof
ridge, in the middle of each roof section and in the corners of window and door openings.
No crushing issues.
Wind Wind loads stress the structure less than seismic. Maximum wall bulge is 4 mm on
the north and south walls. The north and south walls exhibit more bulging than the
41
CHAPTER 4. STRUCTURAL ANALYSIS
east and west walls, even in cross winds. Roof uplift is maximum for 0.9D + Weast/west
combinations, but is under 2 mm. No crushing issues.
(a) Vertical displacement for 1.4D (b) Horizontal Y displacement for 1.2D + Enorth
One strategy to improve the building’s performance would be to introduce curvature to espe-
cially the roof. This can be accomplished by changing the flat surface to a mini barrel vault that
rests on ribs. Another strategy would be to curve the entire roof into a half-cylinder. Both of
these options would reduce the concrete required and resulting loads. However, constructability
would be much more difficult in creating the armature and especially placing concrete from only
one side.
42
CHAPTER 4. STRUCTURAL ANALYSIS
Creating the armature and concreting the lower half of the shell first, including the second
story floor, would reduce scaffolding for completing the upper half.
Insulation is attached to the outside of the shell. Spray foam is the most convenient solution,
but also expensive. It would be more economical to use foam as only a glue that holds EPS
blocks together and seals gaps.
The exterior of the insulation may be smoothed and painted as is done with Monolithic’s
domes. Otherwise, a thin concrete shell can be built with either the laminated technique.
The mass of the shell consists of 51.8 tons of concrete and 4.5 tons of steel.
4.7.3 Results
Displacement under dead load is less than 2 mm for the structure. For seismic load cases,
the horizontal acceleration causes one side to lift slightly and the other to slump, as shown in
Fig. 4.2a. The maximum uplift for the 0.9D + E cases is 3 mm.
The highest stresses occur in the ribs, as they resist the shell’s tendency is to further push
out the walls. There are still surface stresses in the shell, especially at the top near the column
as shown in Fig. 4.2b. Thickening the ribs would reduce this tension. Window areas, especially
at the “face”, would also benefit from additional minor reinforcement. These openings are
generally not problematic because of the ribs.
Seismic load combinations govern in general. Wind loads only affect the large windows,
where the net force from pressure is resisted by the shell. Most of the windows are centered on
bulge sections to most efficiently distribute these loads.
(a) Vertical displacement for 1.2D + Eeast (b) Surface σ3 for 1.4D
In general, a structure like this which is designed with aesthetics as the priority is relatively
expensive. There is considerable effort devoted to studying weak areas, designing reinforcement
and ultimately more material is needed as opposed to designing with loads as the priority. This
is in stark contrast to the Ballena case study, which has a shape optimized for resisting loads
and performs accordingly.
43
CHAPTER 4. STRUCTURAL ANALYSIS
4.8.3 Results
This structure is very efficient at distributing loads and minimizing bending stresses. The
catenary door opening and circular windows also minimize stress concentrations. There is under
1 mm deflection under any load combination in any direction. The maximum displacement sum
is shown in Fig. 4.3a, where the shell is more flexible around the biggest windows. There are
no regions that need additional reinforcement. Crushing is also not a concern.
An example of stress distribution for a seismic load combination is shown in Fig. 4.3b.
Tensile stresses are around the openings but the rest of the structure remains very uniformly
stressed.
44
CHAPTER 4. STRUCTURAL ANALYSIS
(a) Displacement sum for 1.2D + Ewest (b) Surface σ3 for 0.9D + Enorth
45
Chapter 5
5.1 Overview
This chapter presents energy modeling of each of the case studies. The methodology and
software are focused on estimating the annual energy demand in two climates: Chicago, Illinois
and Miami, Florida. Energy demand is split into electricity, heating and cooling. Various design
options are explored, such as insulation thickness, thermal mass and openings, to minimize
energy demand.
The goal is a net-zero house, referring to a net annual electricity use of zero. The house is
assumed to be connected to the grid and is able to sell surplus electricity from on-site generation
and likewise buy from the grid to satisfy shortages. The focus of this analysis is only on reducing
energy demand of the building. The electricity demand is provided for later design of on-site
renewable energy sources such as solar photovoltaics (PV) or small wind turbines. Heating and
cooling systems for each case study are conceptual and sized according to need. An estimate of
a solar PV installation is done for each case study, since it is usually the easiest to implement
in an urban environment.
Another goal of the energy system is to eliminate the direct use of fossil fuels. Natural
gas is widely used for heating in the U.S., currently being an economical solution. To this
extent, ground source heat pumps (GSHP) are incorporated into each case study to cover
space heating, space cooling and hot water. Although higher in capital costs, GSHPs allow for
complete electrification of the energy system, use of on-site generation and long-term savings.
District water heating and cooling are not considered because they’re generally not available
for single family houses in the U.S.
According to the Energy Information Agency (EIA), the annual average residential electricity
use was 9,240 kWh in Illinois and 13,572 kWh in Florida in 2009 [62]. Gas for heating is not
included. These values are a useful baseline for comparing proposed designs.
Finally, the PassivHaus certification requirements for specific annual energy demand are
≤15 kWh/m2 for heating and ≤15 kWh/m2 for cooling [37]. Heat loads are to be ≤10 W/m2 .
These are useful benchmarks of excellent performance, even if PassivHaus certification isn’t
specifically the design focus.
5.2 Methodology
EnergyPlus 7.0 is the simulation framework used in this project. This is an open-source, ma-
ture code developed by the U.S. Department of Energy. It calculates electricity, heating and
46
CHAPTER 5. ENERGY SYSTEM DESIGN & ANALYSIS
cooling demand based on a building’s temperature setpoints, shape, local climate, solar heating
gains, light requirements, indoor equipment heat gains and others. A transient heat balance is
calculated in hourly time steps for an entire year.
DesignBuilder 3.0 is the application used to provide a graphical interface to EnergyPlus.
DesignBuilder is where the geometry is modeled, zones defined with activity levels, material
properties assigned to the geometry and the simulation solved. Due to the program’s limitations
in importing CAD geometry and simple modeling tools, there is considerable simplification of
the curved surfaces of the Ballena and Calabaza case studies. Simple blocks are modeled to
reasonably approximate the true volume, surface area, glazing and building shape.
Each case study geometry is re-created in DesignBuilder, complete with internal room par-
titions. Ferrocement thickness is the same as presented in Chapter 4, while the insulation
thickness is varied to reach PassivHaus performance. Other components are assumed to be en-
ergy efficient, such as triple-pane windows and insulated core doors. In the case of the retrofit,
a base energy demand is also established by assuming conventional wooden construction.
Solar PV sizing is done using PVSYST 5.5. A fixed tilt angle is assumed, optimized for
annual yield in each city. Electrical demand drives the system’s size, as PVSYST calculates
the expected hourly irradiation for Chicago and Miami for a sample year. The goal of the PV
analysis is to estimate annual yield per unit area.
47
CHAPTER 5. ENERGY SYSTEM DESIGN & ANALYSIS
5.3.2 Infiltration
Infiltration refers to the air tightness of a building, usually expressed as air changes per hour at a
pressure of 50 Pascals (ACH50 ). To determine infiltration on an existing building, a special door
is installed that contains a fan. All other openings in the house are closed. ACH50 is calculated
based on interior volume and how much air escapes through gaps in the building once 50 Pascals
is achieved. These uncontrolled losses lead to higher heating and cooling demand.
Natural infiltration for the energy model can be obtained from ACH50 . Conversion factors
1
vary by region. In this analysis, a value of 20 is used, which is a common approximation.
Therefore, the infiltration in the energy model is:
ACH50
ACHnat =
20
An existing home may be assumed to have an ACH50 of 5, which is a requirement for an
EnergyStar 2.0 rating. This value is used in the energy model of the retrofit before the outer
shell is built. PassivHaus requires ACH50 of 0.6. The energy model uses this value for all three
ferrocement structures.
QH
COP =
W
where QH is the heat discharged into the building and W is the electricity needed to run
the heat pump. Heat pumps work more efficiently for lower temperatures, which is why radiant
floor heating is the best choice for the Chicago climate rather than radiators.
EnergyStar requires a minimum average Energy Efficiency Ratio (EER) of 14.1. The EER
is similar to COP in heat pump cooling and is defined as:
QC [BTU]
EER =
W [Wh]
where QC is the heat removed from the building. Cooling COP is more convenient to work
with:
48
CHAPTER 5. ENERGY SYSTEM DESIGN & ANALYSIS
5.3.8 Openings
Openings are a major area of thermal losses and need to be energy-efficient for the whole
building to perform well. Windows are assumed to be triple-pane, low emissivity, with a U-
value of 0.99 W/m2 -K. External doors are assumed to have a U-value of 1.5 W/m2 -K. It’s
important to purchase windows and doors which provide an accurate installed U-value, which
includes thermal bridging at the frame.
Shading is important for hot climates to limit solar gains and avoid over-heating. Outside
shades are assumed in the model, which are closed once the window receives over 200 W/m2 of
solar irradiation. Overhangs may also fulfill this function.
The exterior doors present a thermal loss that can be reduced with a small room dedicated to
the entrance. This way, air exchange with the outside is more limited. An entrance is modeled
in the Calabaza case study.
5.3.9 Lighting
LED lighting is assumed in the building, with an energy intensity of 3.3 W/m2 -lux. Target
luminance for each zone is listed in Appendix B.3. Exterior night-time lighting is also included,
for a total of 50 W.
49
CHAPTER 5. ENERGY SYSTEM DESIGN & ANALYSIS
5.3.10 Solar PV
To illustrate the expected monthly output of a PV system, an example 20 m2 , 3.6 kW nominal
system is setup in PVSYST. The monthly electricity output for Chicago and Miami is shown
in Fig. 5.1.
EP V = G · A · P R · η
where G is the solar irradiation on the plane, A is the collector area, P R is the performance
ratio which incorporates thermal, inverter and other losses and η is the module conversion
efficiency.
It’s assumed that the modules are arranged in a fixed tilt, optimized for annual yield. The
tilt is 24° for Miami and 31° for Chicago. The modules selected are 300 W nominal, with 18.5%
efficiency. PR is 80% for Chicago and 75% for Miami. The difference in PR is due to the larger
thermal losses in hot climates. Snow cover is assumed for three winter months in Chicago,
removing all output.
For estimation purposes, it’s assumed that the system scales linearly with area. Based on
the sample system here, a system could be expected to produce 193 kWh/m2 -yr in Chicago and
282 kWh/m2 -yr in Miami.
50
CHAPTER 5. ENERGY SYSTEM DESIGN & ANALYSIS
which means that less XPS needs to be added. Thermal bridging of the soft wooden studs
and fiberglass batt insulation is not directly supported by EnergyPlus. In this case, bridging is
ignored because it would only increase the effective U-value by less than 10%.
The retrofit baseline also includes lighting and appliance upgrades that reduce electricity
use and heat gains. The original building is also analyzed, where no upgrades have been done
at all. A summary of the retrofit stages is in Tab. 5.1.
51
CHAPTER 5. ENERGY SYSTEM DESIGN & ANALYSIS
In terms of value for investment in energy savings, heat pumps and efficient appliances
provide a bigger benefit than additional insulation. Since the existing fiberglass is left in place,
the newly installed insulation can be thinner than the other case studies. It’s enough to install
75 mm to achieve PassivHaus’s goals of 15 kWh/m2 -yr for both heating and cooling.
In the best-case scenario, a solar PV would need to be 35 m2 in Chicago and 24 m2 in
Miami to cover electrical demand under a net-zero scheme. If solar water heating could cover
all demand, it would save 12 m2 in Chicago and 8 m2 in Miami. Solar DHW savings are the
same for all case studies.
Figure 5.3: Case 2 energy model, 1st and 2nd story zones
52
CHAPTER 5. ENERGY SYSTEM DESIGN & ANALYSIS
to volume as compared to the other cases, improving heat recovery is more important than
insulation. Investment should go towards window shading, more efficient heat recovery and
more efficient appliances.
In the best-case scenario, a solar PV would need to be 73 m2 in Chicago and 52 m2 in Miami
to cover electrical demand under a net-zero scheme.
53
CHAPTER 5. ENERGY SYSTEM DESIGN & ANALYSIS
In Chicago, the extra insulation results in significant reductions in heat demand, although
the difference in total electricity is reduced once DHW and equipment loads are considered.
Specific heating demand is under the PassivHaus limit of 15 kWh/m2 -yr.
In the hot Miami climate, increasing insulation shows much smaller gains. Instead, cooling
demand can be reduced by increasing window shading and decreasing the ventilation rate from
0.4 ACH to 0.3 ACH. Specific cooling demand is about 15 kWh/m2 -yr. Reducing internal gains
in the form of equipment and lighting would also be effective in saving on cooling, as these
gains dominate once window shading is in place. Current gains corresponding to lighting and
equipment are an average of 500 W for the household.
54
CHAPTER 5. ENERGY SYSTEM DESIGN & ANALYSIS
55
Chapter 6
CONCLUSIONS
56
Bibliography
[3] Am-cor Inc. Ferrocement Homes & Building Kits. http://www.am-cor.com/. [Accessed
2013-03-05].
[4] American Concrete Institute. ACI Educational Bulletin E2-00: Reinforcement for
Concrete - Materials and Applications, 2006.
[5] American Concrete Institute Committee 318. ACI 318-11: Building Code
Requirements for Structural Concrete, 2011.
[6] American Concrete Institute Committee 440. ACI 440.1R-06: Guide for the
Design and Construction of Structural Concrete Reinforced with FRP Bars , 2006.
[8] American Concrete Institute Committee 549. ACI 549.1R-93: Guide for the
Design, Construction, and Repair of Ferrocement, 1999.
[9] American Society of Civil Engineers. ASCE 7-05: Minimum Design Loads for
Buildings and Other Structures, 2006.
[12] BRE Global Ltd. BREEAM New Construction: Non-Domestic Buildings, 2011.
[13] British Standards Institute. British Standard 6399-2:1997. Loading for buildings -
Part 2: Code of practice for wind loads, 1997.
57
BIBLIOGRAPHY
[22] Feist, W., Peper, S., and Görg, M. CEPHEUS Final Technical Report . Tech. rep.,
July 2001. Part of the EU research project CEPHEUS - Cost Efficient Passive Houses as
European Standards.
[23] Fouad, F., Farrell, J., Heath, M., Shalaby, A., and Vichare, A. Behavior of
the MR Sandwich Panel in Flexure. Tech. rep., June 2009.
[26] Hartley & Brookes Associates. Boat Plans, Build Your Own Boat with Hartley
Boat Plans and Books. http://www.hartley-boats.com/. [Accessed 2013-03-07].
[29] International Code Council. International Residential Code for One- and
Two-Family Dwellings, 2012 edition. Available at
http://publicecodes.cyberregs.com/icod/irc/index.htm.
[32] Jacob, M., and O’Loughlin, S. Viking Foundations Report: Report on Insulation
Foundations. Tech. rep., December 2010. Report by Tanner Structural Designs, Ltd. for
the Viking Passive House.
[34] Kiffmeyer, D. Earthbag Building: The Tools, Tricks and Techniques (Natural Building
Series). New Society Publishers, June 2004.
[35] Kornher, S. FlyingConcrete - Structural and sculptural forms in light weight concrete.
http://www.flyingconcrete.com. [Accessed 2013-03-07].
58
BIBLIOGRAPHY
[36] Kosny, J., Kossecka, E., Desjarlais, A., and Christian, J. Dynamic Thermal
Performance and Energy Benefits of Using Massive Walls in Residential Buildings. Oak
Ridge National Labs
http://www.ornl.gov/sci/roofs+walls/research/detailed papers/dyn perf/. [Accessed
2013-05-27].
[37] Krapmeier, H., and Drössler, E. CEPHEUS: Living Comfort without Heating.
Springer-Verlag/Wien, 2001.
[42] Monolithic. Monolithic dome homes, schools, churches, storages, gyms and more.
http://www.monolithic.com. [Accessed 2013-03-01].
[45] Nordenson, G., Riley, T., and Candela, F. Seven Structual Engineers: The Felix
Candela Lectures. The Museum of Modern Art, July 2008.
[47] Pedreschi, R. Form, Force and Structure: A Brief History. Architectural Design 78, 2
(March 2008), 12–19.
[50] Reddick, H. W., and Miller, F. H. Advanced Mathematics for Engineers, 2nd ed.
John Wiley & Sons, 1950.
[51] Reynolds, M. Earthship: How to Build Your Own, Vol. 1, 1st ed. Solar Survival
Architecture, September 1990.
59
BIBLIOGRAPHY
[54] Smarter Building Systems. Basalt mesh reinforcement: Smarter Building Systems.
http://www.basalt-mesh-fiber.com/. [Accessed 2013-03-16].
[58] U.S. Coast Guard. Hull Structure Review. Tech. rep., 1976. [Accessed 2013-03-16].
[60] U.S. Department of Energy. Requirements for Residential Electric Resistance Water
Heaters. http://www1.eere.energy.gov/femp/technologies/eep electric waterheaters.html.
[Accessed 2013-06-01].
[62] U.S. Energy Information Administration. How much electricity does an American
home use? http://www.eia.gov/tools/faqs/faq.cfm?id=97&t=3. [Accessed 2013-05-27].
[63] U.S. Green Building Council. Green Building Design and Construction: LEED
Reference Guide for Green Building Design and Construction, 2009 edition.
[64] West, M. Thin Shell Concrete from Fabric Molds. Tech. rep., 2009. [Accessed
2013-03-18].
60
Appendix A
EXAMPLES OF FERROCEMENT
CONSTRUCTION
The purpose of this Appendix is to illustrate several notable ferrocement projects that have
been inspiring to the author. The examples are meant to show what is possible to build. The
descriptions are brief, with some extra sources included for further reading.
61
APPENDIX A. EXAMPLES OF FERROCEMENT CONSTRUCTION
Figure A.2: Monolithic domes are usually unimaginative and almost awkward as a house shape.
However, with large cutouts, the building takes on new character.
(Left) The large dome house in Pensacola Beach, Florida survived a direct hit by Hurricane
Ivan in 2004 and Hurricane Dennis in 2005.
(Right) The Garlock house in the mountains of Colorado also benefits aesthetically from large
cutouts in the dome.
Figure A.3: Vietnam’s Crazy House, designed by Dang Viet Nga as a personal project to create
a tree-like fairy tale house. It now includes 10 guestrooms and tours are offered to recoup the
money of construction. Nga is the daughter of a past Vietnamese president, which likely helped
with the building permits.
The official website is: http://www.crazyhouse.vn/
62
APPENDIX A. EXAMPLES OF FERROCEMENT CONSTRUCTION
Figure A.4: Candela’s hyperbolic paraboloid Los Manatiales restaurant in Xochimilco under
construction and the incredible open space inside. The formwork was straight wooden boards,
with wire reinforcement laboriously laid down on top before the concrete pour. The concrete
was only 4 cm thick.
63
Appendix B
B.1.1 Inputs
Seismic Design Category: D (maximum, California)
F = 1.1 for two-story buildings (§12.14.8.1)
Fa = 1.4 for soil sites (§12.14.8.1)
SS = 1.5, maximum short term acceleration per §12.14.8.1
R = 4, response modification coefficient per Table 12.14.1
B.1.2 Calculations
The calculation procedure is per Chapter 12 of ASCE 7-05. Equation numbers correspond to
the code.
The design spectral acceleration at short periods, SDS , from Equation 12.14.11 is:
2 2
SDS = Fa Ss = · 1.4 · 1.5 = 1.4
3 3
Equations 12.4.3 and 12.4.4 establish the earthquake loading, E, in horizontal and vertical
components:
E = Eh ± Ev
The vertical component is added in load combinations with 1.2D and subtracted in load
combinations with 0.9D. Per Equation 12.14.6:
Ev = 0.2SDS D = 0.28D
The base shear component per Equation 12.14.11:
64
APPENDIX B. DESIGN AND CONSTRUCTION DETAILS
Inputs
V = 76 m/s, design wind speed of 275 km/h
I = 1.0, Table 6-1, importance factor from Occupancy Category II
Kzt , Kz = 1, wind directionality and topography factors
Kd = 0.85, Table 6-4, wind directionality factor
G = 0.85, gust effect factor per §6.5.8.1
Roof angle = 30°
L/B = 8/15.5 = 0.52, floorplan width to length ratio
h/L = 3.65/4 = 0.91, mid-roof height to width ratio
Calculations
The velocity pressure, qz , is calculated from Equation 6-15:
p = qGCp − qi GCpi
where the internal pressure terms of qi drop out. The external pressure coefficient, Cp , is
based on surface type and wind direction. Positive values indicate wind pushing the surface and
negative values indicate suction. The relevant values are calculated from Table 6-6 and listed
in Table B.1. For wind parallel to the roof ridge, a constant value of -0.7 is used for Cp in the
analysis.
65
APPENDIX B. DESIGN AND CONSTRUCTION DETAILS
building shape coefficients can be found in British Standard 6399-2:1997 [13] and are shown in
Tab. B.3. The calculation of q is the same as above and the gust factor, G, is also 0.85.
In the FEA model, the pressures are interpolated for each element’s centroid location. The
transition from wall to roof is assumed to be beyond 45° from the horizontal.
Table B.2: Pressure distribution for dome roof [9, Tab. 6-7]
Angle θ from
incoming wind Cp Pressure [kPa]
0° +1.0 2.6
10° +0.9 2.3
20° +0.7 1.8
30° +0.4 1.0
40° 0 0
50° -0.4 -1.0
60° -0.8 -2.0
70° -1.1 -2.8
80° -1.05 -2.7
90° -0.85 -2.2
100° -0.65 -1.7
120° -0.35 -0.9
140° -0.3 -0.8
160° -0.3 -0.8
180° -0.3 -0.8
66
APPENDIX B. DESIGN AND CONSTRUCTION DETAILS
67
Appendix C
SOFTWARE
Kangaroo Physics engine that adds mass points to a mesh, which can be inflated, hung under
its weight or pulled like a fabric.
GhPython Adds Python scripting to Grasshopper. Used for ANSYS export and some geom-
etry building.
68
APPENDIX C. SOFTWARE
69
APPENDIX C. SOFTWARE
70